Pediatric Endocrinology — The Real Talk

Sex develops in four sequential layers. DSD is what happens when one of these arrows fails or gets bypassed. Most DSD anchors to the main broken arrow — a map, not rigid bins; several (NR5A1, WT1, POR, DHCR7) span more than one layer:

CHROMOSOME -> GONAD -> HORMONE -> END ORGAN

THE TWO FRAMING SENTENCES:

• 46,XY DSD = the male program FAILED TO COMPLETE -> undervirilization.
• 46,XX DSD = the female program GOT HIJACKED -> virilization (or the gonad got swapped) — but that's the NEWBORN-ambiguity story (usually androgen excess). Other 46,XX DSD is gonadal or anatomic, surfacing later as primary amenorrhea (ovarian dysgenesis, MRKH).

46,XY DSD breaks down as G-H-A (Gonad / Hormone / Action):

46,XX DSD has three buckets: fetal androgens, rogue testicular tissue, or gonadal/anatomic.

SEX CHROMOSOME DSDs are the third bucket, separate from XY / XX because the lesion is at LAYER 1:

• 45,X / variants -> Turner (female, short, gonadal dysgenesis).
• 47,XXY -> Klinefelter (male, primary hypogonadism, tall, learning).
• 45,X/46,XY mosaic -> mixed gonadal dysgenesis (one streak + one dysgenetic testis is classic; ambiguous; ALWAYS gonadectomy on the intra-abdominal dysgenetic gonad for tumor risk).
• 46,XX/46,XY -> chimerism, often ovotesticular.
• 47,XYY, 47,XXX -> usually no DSD.

THE DSD STAGING SCALES — the four scores you'll see quoted, and where each one applies:

• PRADER (1-5): degree of VIRILIZATION of 46,XX external genitalia. 1 = mild clitoromegaly; 3 = single urogenital-sinus opening + phallus; 5 = fully male-appearing with a penile urethra. This is how the virilized 46,XX congenital adrenal hyperplasia (CAH) newborn is graded.
• QUIGLEY (1-7): degree of UNDERVIRILIZATION in a 46,XY with androgen insensitivity. 1 = normal male; 2-5 = increasing hypospadias / micropenis / ambiguity (the PAIS range); 6-7 = female external genitalia with sparse-to-absent pubic hair (CAIS).
• EMS (External Masculinization Score, 0-12): sums scrotal fusion, phallus size, urethral position, and each gonad's location into one number. LOW = more undervirilized; used to quantify and track a 46,XY DSD.
• EGS (External Genitalia Score, 0-12): the updated EMS (Ahmed 2020), rescaled to apply to BOTH sexes and to normal newborns, so it reads as a continuous reference across the whole spectrum.

Shorthand: PRADER scores virilization UP (46,XX), QUIGLEY scores it DOWN (46,XY); EMS / EGS put a single NUMBER on the 46,XY exam.

BEDSIDE HEURISTIC for the ambiguous-genitalia newborn — the three questions that get you 80% there before genetics:

Q1. ARE THE GONADS PALPABLE?

• Yes (inguinal palpable) -> testicular tissue exists -> 46,XY DSD (dysgenesis, hormone synth, AIS, 5α-RD) or 46,XX testicular DSD.
• No (none palpable) -> most likely VIRILIZED 46,XX -> CAH first. Send 17-OHP, electrolytes, cortisol, testosterone, karyotype URGENTLY.

Q2. ARE MÜLLERIAN STRUCTURES PRESENT (pelvic US)?

• Yes + 46,XY -> AMH didn't work -> gonadal dysgenesis or persistent Müllerian duct syndrome (PMDS).
• No + 46,XY -> Sertoli was intact -> androgen synth or AR problem.

Q3. HORMONAL TRIAD (testosterone, DHT, AMH; minipuberty 1-3 mo or hCG stim if outside window):

• Low T, low DHT, low AMH -> testis failed (dysgenesis, regression).
• Low T, low DHT, normal/high AMH -> steroidogenesis block.
• Normal/high T, low DHT -> 5α-reductase deficiency.
• Normal/high T, normal/high DHT, undervirilized -> AIS.
• Normal T/DHT + undescended / Müllerian present -> PMDS (serum AMH LOW if the lesion is AMH itself, NORMAL/HIGH if AMHR2).

hCG STIMULATION TEST interpretation grid (cut-offs per Güran 2023):

PROTOCOL: short hCG = 1500 IU/m² IM daily x 3 days; measure T, DHT, androstenedione, 17-OHP at baseline and 24 h after last dose. Extended hCG = 1500 IU twice weekly x 3 weeks (more sensitive + therapeutic trial for micropenis). Always pair with baseline AMH + inhibin B (don't need stimulation).

The two scenarios where the test is most decisive: 1. 46,XY newborn with bilateral non-palpable gonads + small phallus. If hCG yields T and AMH/inhibin B detectable -> testes exist, locate them. If hCG yields nothing AND AMH/inhibin B undetectable -> anorchia, skip laparoscopy. 2. Micropenis + small palpable testes. Brisk T response = CHH (good prognosis for gonadotropin therapy). Blunted = primary testicular pathology.

Mayer-Rokitansky-Küster-Hauser syndrome (MRKH) — MÜLLERIAN AGENESIS. The 46,XX counterpart of complete androgen insensitivity syndrome (CAIS) in the "primary amenorrhea with blind vagina" differential.

• 46,XX woman, normal ovaries, normal external genitalia, normal pubertal development — but ABSENT uterus + upper two-thirds of vagina. Incidence ~1 in 4,500.
• TYPE I (isolated) vs TYPE II (MURCS: Müllerian + Renal + Cervicothoracic Somite). 30-50% of MRKH is type II.
• Embryology: Müllerian (paramesonephric) ducts form at 5-6 wk and fuse caudally to form uterus, cervix, upper vagina. Lower vagina is urogenital-sinus derived and intact. In females the program is default-on; failure of the transcription network (WNT4, WNT9B, LHX1, HNF1B, GREB1L) = MRKH.
• Key genes: * WNT4 — MRKH + hyperandrogenism + renal anomalies. The only MRKH with a biochemical signal (WNT4 normally represses theca-like androgen production). * LHX1, WNT9B — Müllerian/renal mesoderm regulators. * HNF1B (TCF2) — MRKH + renal cysts +/- MODY5. * 17q12 microdeletion (covers HNF1B + LHX1) — do CMA especially with renal anomalies, syndromic features, or family history.
• Workup: karyotype FIRST (rule out CAIS), CMA, renal US, spine imaging, audiology, echo.

MRKH vs CAIS — the differential at presentation:

Bedside: primary amenorrhea + normal breasts + NORMAL pubic hair + short blind vagina = MRKH until karyotype. Same anatomy + SPARSE pubic hair = CAIS.

Management: vaginal dilators (Frank's method) first-line (85-90% success). Fertility = oocyte retrieval + surrogacy historically; UTERINE TRANSPLANTATION now established in select centers with live births. Counseling is identity-loaded — address explicitly.

STRUCTURAL AMBIGUITY — THE NON-HORMONAL MIMIC. Not every atypical newborn genital exam is endocrine. A persistent cloaca — failure of the urorectal septum to split the cloaca (gestational wk 6-7), so the urethra, vagina, and rectum exit through ONE common channel and a SINGLE perineal orifice — shows a prominent genital tubercle + fused labia that mimic PRADER virilization. The genitalia are androgen-INNOCENT: this is plumbing, not hormones, and it occurs essentially ONLY in 46,XX females with normal ovaries (~1 in 50,000).

THE DISCRIMINATOR from a virilized 46,XX: find the uterus and count the orifices.

• Androgen excess (CAH) builds a urogenital SINUS (urethra + vagina share an outlet) but leaves a SINGLE NORMAL midline uterus (AMH never fired) and a SEPARATE, normally-sited anus.
• Cloaca gives MÜLLERIAN DUPLICATION (uterus didelphys, often with a longitudinal vaginal septum; ~30-60%), often a hydrocolpos (~30%, compressing the bladder trigone -> hydronephrosis), and the RECTUM draining into the common channel.
• Kill point: CAH NEVER builds a cloaca or a second uterus. A didelphys, or one perineal hole with the rectum in it, is structural — the 17-OHP you (correctly) still send to exclude CAH comes back normal because the lesion was anatomic all along.

Common-channel length is the prognosis lever: <3 cm = simpler repair, better continence; >3 cm = complex, ~78% need clean intermittent catheterization. Renal anomalies ride along in up to 90% (lifetime chronic kidney disease ~50%) — always image kidneys + spine (tethered cord, sacral ratio).

THE 46,XY CURVEBALL: a retained uterus here is NOT automatically PMDS. PMDS (AMH / AMHR2) is an AMH-only lesion -> a NORMALLY virilized male with a SINGLE normal uterus, found incidentally at hernia repair; it cannot produce an ambiguous phallus OR a didelphys. An ambiguous phallus WITH Müllerian structures in a 46,XY needs Sertoli (AMH) AND Leydig (androgen) to fail together = GONADAL DYSGENESIS, which then carries germ-cell tumor risk (§6). The cloaca + didelphys stay structural either way — the karyotype re-reads the GONAD, never the malformation.

MÜLLERIAN DUCTS FAIL THREE WAYS — the framework already names two; the cloaca case exposes the missing FUSION lane:

The shared steroidogenic map — one trunk (cholesterol), three product arms, and the congenital adrenal hyperplasia (CAH) enzyme blocks that divert the flow:

StAR shuttles cholesterol from the outer to the inner mitochondrial membrane so P450scc can cleave the side chain. StAR is THE rate-limiting acute step of steroidogenesis. Lipoid CAH = StAR LoF.

Two-hit model is the whole story. Hit 1: ~14% basal steroidogenesis without StAR keeps the cell limping along. Hit 2: ACTH keeps coming, cholesterol esters accumulate as toxic lipid droplets, the cell dies. So 46,XY infants are completely female at birth (testis active in utero, both hits already happened). 46,XX is born phenotypically female with adrenal failure but the ovary is quiet in utero — it gets some partial puberty before the second hit destroys it. There are 46,XX women with StAR who ovulated with stimulation.

Non-classic lipoid CAH = hypomorphic StAR with isolated cortisol deficiency, no DSD. Mimics FGD. Don't miss it.

CYP17A1 does two reactions: 17-alpha-hydroxylation and 17,20-lyase. Complete loss = no cortisol (but corticosterone partially substitutes, no Addison crisis), no sex steroids, AND mineralocorticoid hypertension from 11-deoxycorticosterone (DOC). The teaching triad: 46,XY DSD + no breast at puberty + HTN with hypokalemic alkalosis = 17-alpha-OH deficiency.

Isolated 17,20-lyase deficiency = the lyase reaction selectively dies while 17-alpha-OH still works. Cortisol preserved. No HTN. Three causes matter:

• CYP17A1 redox surface mutations (POR/b5 interface).
• CYB5A LoF — ALSO has methemoglobinemia because b5 reduces met-Hb in red cells. The "blue baby with 46,XY DSD" is CYB5A.
• P450 oxidoreductase (POR) LoF — broader pattern hitting 17-OH, 21-OH, and aromatase simultaneously, plus Antley-Bixler skeletal phenotype. 46,XX can virilize in utero by the backdoor (alt) androgen pathway.

BACKDOOR (ALTERNATIVE) ANDROGEN PATHWAY — how a fetus makes DHT WITHOUT a testosterone intermediate, and the mechanism this compendium keeps invoking (POR, 21-OH, 46,XX virilization):

• FRONTDOOR (classic): 17OH-progesterone -> (17,20-lyase) androstenedione -> testosterone -> (SRD5A2) DHT.
• BACKDOOR: when 17OH-progesterone PILES UP behind a block (21-OH or POR deficiency, or the fetal adrenal itself), it is 5α- and 3α-reduced FIRST (SRD5A1, AKR1C2/4), then finished by CYP17 lyase + 17βHSD/AKR1C3 to androsterone -> DHT — BYPASSING testosterone and androstenedione.
• WHY IT MATTERS: (1) it explains 46,XX VIRILIZATION in 21-OH and POR CAH — backed-up 17OH-prog drives fetal DHT and ambiguous genitalia WITHOUT a measurable testosterone rise; (2) DEFECTS in the backdoor enzymes themselves (AKR1C2/4, SRD5A1) give 46,XY UNDER-virilization despite an intact frontdoor, because the fetus needs BOTH routes. The 5α-reduced urinary steroid metabolome (GC-MS) is the tell.

SF1 (NR5A1) is the master steroidogenic and gonadal transcription factor. DAX1 (NR0B1) is its orphan-receptor antagonist. They oppose each other in a dose-sensitive way.

DAX1 LoF in 46,XY = X-linked adrenal hypoplasia congenita (AHC) plus hypogonadotropic hypogonadism. Adrenal crisis in infancy, gonadotropin failure unmasking at expected puberty.

DSS = the OPPOSITE — Xp21 duplication including NR0B1 in 46,XY produces ovotestis or female phenotype despite intact SRY. Too much DAX1 antagonizes SF1/SOX9 and silences testis program.

SF1 mutations span the entire 46,XY DSD spectrum from complete gonadal dysgenesis to isolated male-factor infertility. Adrenal failure variable. Oligogenic, poor genotype-phenotype correlation.

A child presenting with both adrenal insufficiency AND a DSD has a small, high-stakes differential. These are syndromes where one gene or one embryologic event hits both axes.

Myelodysplasia, infection, restriction of growth, adrenal hypoplasia, genital, enteropathy (MIRAGE) SYNDROME (SAMD9 gain-of-function, sporadic de novo): Myelodysplasia + Infections + Restriction of growth + Adrenal hypoplasia + Genital phenotype (46,XY undervirilization, sometimes complete sex reversal) + Enteropathy. SAMD9 GoF inhibits cell proliferation across multiple lineages. Bone marrow failure progresses to monosomy 7 / MDS in many cases. Severe IUGR. Variable adrenal failure. 46,XY DSD ranges from female external genitalia to ambiguous.

The MONOSOMY 7 "RESCUE" looks paradoxical but it is clonal selection, not a random consequence. SAMD9 sits on 7q21. The GoF allele actively suppresses cell division. Any marrow cell that happens to LOSE the chromosome 7 carrying the mutant allele (either by full -7 monosomy or by 7q deletion encompassing SAMD9) escapes growth arrest and out- proliferates its restrained siblings. Over time the marrow becomes clonally dominated by -7 / 7q- cells. So you see the strange combination: the hematologic phenotype IMPROVES (the patient is no longer transfusion- dependent) WHILE the karyotype now shows monosomy 7, which would ordinarily flag pre-MDS. The same logic applies to MIRAGE patients who acquire UPD7q with a second-site loss-of-function SAMD9 mutation on the remaining allele — another way to silence the bad allele. The teaching point: in MIRAGE, monosomy 7 is sometimes adaptive rather than malignant, and MDS workup needs SAMD9 context before jumping to transplant.

IMAGE SYNDROME (CDKN1C gain-of-function, MATERNALLY inherited): Intrauterine growth restriction + Metaphyseal dysplasia + Adrenal hypoplasia congenita + Genital anomalies (46,XY DSD). CDKN1C is paternally imprinted — only the maternal allele is normally expressed. Activating mutations on the maternal allele cause growth arrest. Father transmits silently (his allele is silenced by imprinting). Mother transmits disease. Severe IUGR plus AHC plus 46,XY undervirilization plus skeletal dysplasia.

IMAGEi: IMAGE + Immunodeficiency. CDKN1C also affects T-cell development; some mutations add this layer.

SERKAL SYNDROME (WNT4 loss-of-function, AR): SEx Reversal in 46,XX (gonad develops as testis-like) + Kidney anomalies (renal agenesis) + Adrenal anomalies + Lung agenesis. Very rare. WNT4 is critical for ovarian determination; loss derepresses the testis pathway in a 46,XX gonad.

SMITH-LEMLI-OPITZ (DHCR7 LoF, AR): cholesterol synthesis failure at the 7-dehydrocholesterol reductase step. 46,XY DSD (undervirilization from substrate-limited steroidogenesis) + adrenal insufficiency + multi-organ malformations + dysmorphic face (microcephaly, ptosis, anteverted nostrils, 2-3 toe syndactyly). Elevated plasma 7-dehydrocholesterol is diagnostic.

PALLISTER-HALL (GLI3 truncating mutation -> a constitutive REPRESSOR, NOT simple LoF — the distinction is the teaching point: plain GLI3 haploinsufficiency gives Greig cephalopolysyndactyly instead): hypothalamic hamartoma + central pituitary dysfunction (including ACTH deficiency -> secondary AI) + polydactyly (especially central/mesoaxial) + DSD variable. Sonic hedgehog pathway.

STAR / CYP11A1 (lipoid CAH): 46,XY phenotype female at birth + complete adrenal failure (covered in §2).

P450 oxidoreductase (POR) DEFICIENCY: backdoor androgen pathway 46,XX virilization in utero + 46,XY undervirilization + variable adrenal symptoms + Antley-Bixler skeletal phenotype.

And the classical CAHs that straddle this combo: 21-OH congenital adrenal hyperplasia (CAH) (46,XX virilization + salt loss) and 11b-OH CAH (46,XX virilization + HTN).

Clinical move when AI + DSD coexist: karyotype + microarray + a targeted panel covering SF1, DAX1, StAR, CYP11A1, SAMD9, CDKN1C, WNT4, DHCR7, GLI3, plus the CAH biochem (17-OHP, 11-deoxycortisol, renin/aldo, electrolytes). The differential narrows fast.

ADRENAL + GONADAL COMBINED DISORDERS — where the two axes fail together, because shared steroidogenic enzymes or shared transcription factors hit both glands:

TWO RULES: 1. Steroidogenic-enzyme defects -> HIGH LH/FSH (no sex steroid feedback). 2. DAX-1 (NR0B1) breaks rule 1: adrenal failure + LOW gonadotropins. Same factor governs hypothalamic GnRH neurons AND adrenal steroidogenesis. Neonatal salt-wasting AI in a boy who later fails puberty with low LH/FSH = sequence NR0B1.

RECOGNITION HEURISTIC:

• Salt-wasting + 46,XY undervirilization -> StAR / CYP11A1 / HSD3B2.
• Hypertension + 46,XY undervirilization -> CYP17A1.
• Skeletal + ambiguous + maternal virilization in pregnancy -> POR.
• Salt-wasting + low gonadotropins later -> DAX-1.
• Adrenal AI + 46,XY DSD + normal pituitary -> SF-1 (NR5A1).

Two cells, two gonadotropins, two products. Reading both is what makes the DSD workup informative.

AMH. TGF-β family. Sertoli from ~7 wk gestation. Regresses Müllerian ducts via AMHR2. After that, the serum level reads SERTOLI MASS + IMMATURITY. High through fetus, infancy, childhood; collapses at puberty.

Inhibin B. α-βB heterodimer. Sertoli (and granulosa). FSH-stimulated; suppresses pituitary FSH selectively. Reads Sertoli function and spermatogenesis. Inhibin A (α-βA) is the female luteal-phase / placental marker — negligible in males. (Granulosa-cell tumors are marked by inhibin B + AMH, not inhibin A.)

AMH paradox at puberty — the mechanism behind the most useful pediatric marker. Immature Sertoli has NO functional AR. AMH secretion is FSH-driven, T-blind, through fetus and childhood. At puberty, intratesticular T surges, Sertoli expresses AR, AR represses AMH transcription. So at pubertal transition AMH FALLS while T RISES, and inhibin B does the OPPOSITE of AMH (rises with active spermatogenesis). Two hormones, same cell, opposite directions.

Corollary: AMH stays HIGH in complete androgen insensitivity syndrome (CAIS) / PAIS — Sertoli can't sense T to shut it off.

FIVE REFINEMENTS to "FSH makes sperm, LH makes T":

1. FSH alone doesn't make sperm. Need FSH + very high intratesticular T (10-100x serum) from neighbor Leydig acting paracrine on Sertoli. For FUTURE fertility the modern pediatric sequence PRIMES with FSH FIRST (to expand the Sertoli / germ-cell pool), THEN adds hCG for the intratesticular-T milieu (Abaci 2024) — the reverse of the older adult hCG-first induction. (An adult wanting sperm NOW: hCG-first to build testis size + T, then add FSH, still applies.)

2. Virilization splits T vs DHT.

• T -> Wolffian (epididymis, vas, seminal vesicles), voice, muscle, libido, growth spurt.
• DHT (via 5α-RD) -> external genitalia masculinization (fetal), prostate, beard, terminal body / facial hair, acne. 5α-RD def = normal T + no DHT -> female externals at birth, then virilizes at puberty when T pushes through AR directly.

3. Descent has two phases.

• Transabdominal (8-15 wk): INSL3 (Leydig PEPTIDE) -> RXFP2 on gubernaculum. NOT testosterone.
• Inguinoscrotal (25-35 wk): androgen-dependent. Bilateral intra-abdominal testes = severe Leydig / INSL3 problem, not mild central HH.

4. AMH and inhibin B speak Sertoli only. Healthy = Sertoli pool fine, fertility plausible. Low = red flag. Fertility ALSO needs Leydig, tract, post-pubertal drive.

5. Don't replace with T if fertility matters. Exogenous T suppresses gonadotropins, shrinks testes, kills spermatogenesis. Use hCG (+/- FSH) instead.

DIAGNOSTIC USES:

• Bilateral non-palpable testes + detectable AMH = testicular tissue exists. Undetectable AMH + undetectable inhibin B in 46,XY = anorchia (skip laparoscopy — but only if no Müllerian structures / Y material; the same pattern can be severe dysgenesis with a gonad worth removing).
• Klinefelter: AMH and inhibin B fall BEFORE T. Sertoli leads the way, Leydig follows.
• Sertoli cell tumor: high AMH + high inhibin.
• Ovarian reserve / granulosa-cell tumor: AMH = reserve marker, inhibin = tumor marker.

SERTOLI-CELL-ONLY SYNDROME (SCO; old: Del Castillo): seminiferous tubules lined by Sertoli cells with NO germ cells -> non-obstructive azoospermia. Lab is the framework in one line: NORMAL T (Leydig intact) + HIGH FSH + LOW inhibin B, small firm testes, normal virilization. Largely an adult-infertility histology — the peds value is the upstream, preventable causes:

• Y MICRODELETION (AZFa/b/c) — the genetic anchor.
• Klinefelter (47,XXY) — tubules burn out to SCO architecture.
• Prolonged bilateral CRYPTORCHIDISM — the orchiopexy-by-age-1 driver.
• CHEMO / RADIATION (alkylators) — the pre-treatment cryopreservation driver.
• Post-pubertal mumps ORCHITIS; idiopathic.

The move is upstream: timely orchiopexy, pre-chemo cryopreservation, family Y-microdeletion testing (MICRO-TESE + ICSI is the downstream option).

All three present as 46,XY undervirilization with female or ambiguous external at birth, NO Müllerian structures (Sertoli AMH worked), palpable or inguinal testes. They diverge on a single axis: WHERE in the androgen pathway is signal lost?

LH -> Leydig -> [steroidogenesis] -> T -> 5α-RD -> DHT -> AR -> response

Three disorders, three break-points on that one chain:

• 17β-HSD3 deficiency breaks [steroidogenesis -> T] — CAN'T MAKE T.
• 5α-RD2 deficiency breaks [T -> DHT] — CAN'T MAKE DHT.
• AIS breaks [T / DHT -> AR -> response] — CAN'T HEAR T OR DHT.

That single block rewrites the biochemistry, the puberty, the management:

WHY EACH VIRILIZES (OR DOESN'T) AT PUBERTY:

17β-HSD3. The block is testicular and partially BYPASSABLE. At puberty, LH drives massive Δ4-androstenedione production. Peripheral tissues express OTHER 17β-HSD isoforms (notably AKR1C3 / type 5) in fat, skin, liver that convert Δ4-A -> T. So serum T climbs into the male range despite the testicular block, AR is intact, and virilization proceeds — sometimes dramatically. This is why a 46,XY child raised female with 17β-HSD3 will deepen voice, grow phallus, develop muscle at puberty.

5α-RD2. T synthesis is INTACT. AR works. Two routes drive virilization despite missing DHT:

• Pubertal T concentrations are high enough to push some signal through AR directly (AR has lower affinity for T than DHT, but concentration overcomes it).
• 5α-reductase TYPE 1 (different gene, SRD5A1) is expressed in liver and skin and INCREASES at puberty; it produces some DHT extragenitally.

The trade-off: DHT-specific phenotypes still suffer. Beard, temporal recession, acne, prostate growth don't develop normally. Voice change, muscle, libido, phallic growth do.

Complete androgen insensitivity syndrome (CAIS). AR doesn't work in any tissue. T and DHT both rise at puberty but produce NO androgen-mediated change. Meanwhile T -> E2 aromatization happens normally, and E2 acts UNOPPOSED on intact ERα. Result: full breast development, female fat, female habitus, no menses (no uterus since AMH-driven Müllerian regression worked), and SCANT pubic/axillary hair — the diagnostic giveaway on exam.

CAIS vs SWYER — the other classic phenotypic-female + 46,XY pair:

The single most useful exam: PELVIC ULTRASOUND.

• Uterus PRESENT in 46,XY phenotypic female -> Swyer (or partial GD). Plan urgent gonadectomy.
• Uterus ABSENT in 46,XY phenotypic female -> CAIS. Locate testes.
• Uterus ABSENT in 46,XX phenotypic female with primary amenorrhea -> Mayer-Rokitansky-Küster-Hauser syndrome (MRKH) (covered in §1).

The two-sentence shorthand:

• CAIS = breasts but sparse-to-absent pubic/axillary hair. No AR action means E2 rules unopposed.
• Swyer = nothing without HRT. No gonad, no estrogen. Uterus is fine.

A Swyer patient CAN carry a pregnancy with donor oocytes + IVF. A CAIS patient cannot (no uterus). Mention this early in counseling.

The WT1 CAVEAT in 46,XY gonadal dysgenesis: when the driver is WT1 the KIDNEY is the other organ on the line, so always image kidneys + screen urine. Denys-Drash (WT1 missense, exon 8/9) = diffuse mesangial sclerosis + Wilms tumor + dysgenesis; Frasier (WT1 intron-9 +KTS splice defect) = FSGS / steroid-resistant nephrotic syndrome + high gonadoblastoma risk. Don't just gonadectomize a Swyer-pattern XY without checking the kidneys.

Gonadal germ-cell tumor (GGCT) risk is the consensus stratification that actually drives the gonadectomy decision [Guerrero-Fernandez 2022]. Two things set the risk: how much Y material (the GBY locus / TSPY) sits in an undervirilized or dysgenetic gonad, and gonadal position (intra-abdominal >> inguinal > scrotal).

Two PRECURSOR lesions, two trajectories [Guerrero-Fernandez 2022]:

• GONADOBLASTOMA (GB) — in a granulosa/FOXL2-supported dysgenetic gonad; ~50% progress to invasive disease (dysgerminoma or other GGCT).
• GCNIS (germ-cell neoplasia in situ — the old CIS / ITGCN) — Sertoli/SOX9-supported; ~100% predestined to malignancy, mean onset 14-44y. Stain OCT3/4, c-KIT, PLAP; TSPY marks the Y-derived clone.

IF GONADS ARE RETAINED (CAIS, selected scrotal testes) the surveillance package is [Guerrero-Fernandez 2022]: inguinal/scrotal orchidopexy to bring any abdominal gonad somewhere you can palpate it; self-examination plus YEARLY gonadal ULTRASOUND from puberty; YEARLY AFP, beta-hCG, LDH from puberty. A single biopsy can MISS a focal precursor — sample heterogeneous gonads (ovotestes) in multiple sites, and treat a negative biopsy as never fully clearing the gonad. No imaging/marker package reliably EXCLUDES a premalignant lesion, so retain-vs-remove stays a shared decision.

Partial androgen insensitivity syndrome (PAIS) sits between CAIS and AIS / MAIS on the same AR-defect spectrum. The Quigley scale grades undervirilization 1-7. Management is the hardest decision in DSD because phenotype can present anywhere from "undervirilized male with hypospadias + gyne" to "ambiguous" to "virilized female with clitoromegaly". Multidisciplinary case-by-case.

Three findings, three different fetal windows. Knowing which window was hit narrows the differential before any lab returns.

EMBRYOLOGICAL WINDOWS:

• Genital MASCULINIZATION (urethral fold fusion, glans formation, scrotal fusion): weeks 8-14, hCG -> Leydig T -> DHT -> AR signaling in genital skin. Failure here = HYPOSPADIAS (formation defect).
• PHALLIC GROWTH (penile elongation of an already-formed organ): week 14 onward, depends on fetal HPG axis taking over from placental hCG, plus GH/IGF-1 in late gestation, plus minipuberty 1-6 mo postnatal. Failure here = MICROPENIS (growth defect of normally-formed penis).
• TESTICULAR DESCENT in two phases: * Transabdominal (weeks 8-15): INSL3 -> RXFP2 on gubernaculum. NOT testosterone. Bilateral intra-abdominal testes points HARD at INSL3 / RXFP2 / severe Leydig pathway, not at "mild central HH". * Inguinoscrotal (weeks 25-35): androgen-dependent.

So the question becomes: was Leydig + androgen-action intact in the FIRST half of gestation, the SECOND half, both, or neither?

HYPOSPADIAS DIFFERENTIAL = "what perturbed androgen synthesis or action during weeks 8-14?" -> DSD-spectrum etiologies dominate.

MICROPENIS DIFFERENTIAL = "what failed in the second-half HPG drive or the GH axis?" -> CHH and CPHD (combined pituitary hormone deficiency) dominate. Severity proportional.

When BOTH are present + cryptorchidism, the lesion spans both windows and the differential becomes severe-DSD.

THE WORKUP SPLITS BY PHENOTYPE:

ISOLATED MICROPENIS WITH NORMAL URETHRA AND DESCENDED TESTES = pituitary workup, NOT DSD workup:

• Karyotype (still — don't skip).
• Pituitary panel: cortisol, GLUCOSE, free T4 / TSH, prolactin, insulin-like growth factor (IGF-1).
• Minipuberty hormones (1-3 mo): LH, FSH, T, AMH, inhibin B.
• Pituitary MRI — ectopic posterior pituitary, hypoplastic anterior, absent septum pellucidum, optic nerve hypoplasia.
• Eye exam for SOD.
• Smell evaluation later in childhood for Kallmann.
• Genetics: CHH/Kallmann panel (ANOS1, FGFR1, PROK2/PROKR2, CHD7, GNRHR, KISS1R, TAC3/TACR3); CPHD genes (HESX1, PROP1, POU1F1, LHX3/4, SOX2/3); methylation for Prader-Willi if hypotonia / feeding issues.

KEY POINT: the urgent danger is HYPOGLYCEMIA from cortisol+GH deficiency, not the phallus. Don't get distracted.

ISOLATED HYPOSPADIAS (no other findings):

• DISTAL (glanular, coronal, subcoronal) in a phenotypically normal- appearing male: endocrine yield ~5-10%, surgical referral usually suffices. But still verify both testes descended, scrotum fused, phallus normal-sized before declaring "isolated".
• PROXIMAL (penoscrotal, perineal) hypospadias, or hypospadias with bifid scrotum / chordee / undescended testis / suspected Müllerian: this is DSD UNTIL PROVEN OTHERWISE. Endocrine yield 30-50%. Karyotype URGENTLY + 17-OHP + electrolytes + cortisol + ACTH (rule out CAH) + hCG stim (T, DHT, androstenedione) + AMH + inhibin B + pelvic/abdominal imaging for Müllerian + DSD gene panel (AR, SRD5A2, HSD17B3, NR5A1, WT1, MAP3K1, DHH, GATA4/FOG2, MAMLD1, MID1, DHCR7, NR0B1, CYP17A1, HSD3B2, StAR, CYP11A1, POR, DHX37).
• Renal US (especially if WT1 on the table — Denys-Drash / Frasier).
• Cholesterol + 7-DHC for Smith-Lemli-Opitz.

MICROPENIS + BILATERAL CRYPTORCHIDISM: This triad is the STRONGEST endocrine signal in newborn genital exam. Both formation and growth went wrong, descent likely failed too. Treat as SEVERE DSD: karyotype, full biochem, imaging, genetics from day 1, multidisciplinary involvement.

TREATMENT-MEANINGFUL DECISIONS:

• CHH with cryptorchidism in infancy: GROWING CASE for gonadotropin therapy during minipuberty window (recombinant FSH + hCG, or GnRH pump). Restores phallic growth, descends testes, expands the Sertoli pool, may improve future fertility. Physiologically more attractive than testosterone, which only addresses the phallus.
• Isolated phallic growth without DSD ambiguity: LOW-DOSE T trial (25 mg IM every 4 weeks x 3) gets normal stretched length, simplest empirical approach.
• CPHD with micropenis: GH replacement contributes meaningfully to phallic growth, alongside cortisol + thyroid.
• Vanishing testes: skip laparoscopy if AMH + inhibin B undetectable AND hCG fails to elicit T response AND no Müllerian structures / Y material. If Müllerian or Y material are present, the same labs may be severe dysgenesis — locate the gonad (tumor risk).

SHORTHAND:

• Hypospadias = androgen problem in weeks 8-14. Differential is DSD-spectrum.
• Micropenis = HPG / GH problem after week 14, or postnatal. Differential is pituitary-spectrum.
• Both = sustained androgen-axis problem spanning both windows. Differential is severe DSD.

PEARL: severity of hypospadias predicts the endocrine yield. Distal = 5-10%; proximal = 30-50%. Don't assume "small testes" means cryptorchid — orchidometer or US to measure. Small but descended testis + small phallus + normal urethra is still primarily a CHH / CPHD picture, not a DSD.

Complete androgen insensitivity syndrome (CAIS) is a 46,XY person born phenotypically female because their androgen receptor is dead. Testes are intact and work fine. Testosterone is normal or high. LH is high (no AR feedback at the hypothalamus). AMH from Sertoli cells regresses the Mullerian ducts, so no uterus. The lower vagina forms from urogenital sinus but it is blind-ending. At puberty T gets aromatized to estradiol and breasts develop with no pubic/axillary hair because that hair is androgen-dependent. They're tall — from Y-chromosome growth-gene dosage plus androgen-independent factors, not androgen action per se — but epiphyses fuse normally (estrogen does that, not testosterone).

Partial androgen insensitivity syndrome (PAIS) is the middle of the spectrum. AR works partly. Phenotype scales with residual function. Quigley 3-5 (mid-scale on the 1-7 androgen-insensitivity grading; the DSD staging scales are in §1) is where the chaos lives. Gynecomastia is common because estrogen-mediated breast development happens unopposed by a weak AR.

The 46,XY pubertal differential you cannot screw up: in PAIS/CAIS the AR is broken so puberty does NOT rescue. In 5-alpha-reductase 2 deficiency the AR is fine but T cannot become DHT; at puberty, T gets so high it hits the AR directly and the patient virilizes (muscle, voice, phallic growth) with sparse beard. In 17-beta-HSD3 deficiency the testis cannot make T from androstenedione; at puberty peripheral conversion fills the gap, the intact AR responds, virilization is striking, gynecomastia is common because androstenedione also aromatizes to estrogen.

The quick triad: PAIS = gynecomastia + high LH/T ratio. 5aRD2 = minimal gyneco, low DHT, T/DHT >10. 17bHSD3 = gyneco + low T + high A/T ratio.

The MILD end (MAIS): a normal-appearing male with gynecomastia, high-normal T with high-normal LH, and impaired spermatogenesis / infertility — the AR works, just weakly. And the AR-gene curveball: Kennedy disease (spinobulbar muscular atrophy) is a CAG-repeat expansion in AR -> adult-onset lower-motor-neuron degeneration WITH mild androgen insensitivity (gynecomastia, reduced fertility) — the one AR disorder that is a polyglutamine neurodegeneration.

Breast tissue grows when ESTROGEN acts on functional ERα in the breast. Androgens directly INHIBIT breast development at the ER level. So three questions decide the outcome:

1. Is there an androgen SUBSTRATE? (testis or adrenal) 2. Is AROMATASE available to convert it to E2? (testis, adipose, placenta, etc.) 3. Are AR and ER BOTH FUNCTIONAL? Because if AR works, breast is suppressed; if AR doesn't, unopposed E2 wins.

SHORTHAND TO KEEP:

• complete androgen insensitivity syndrome (CAIS) = breasts but no hair below the neck. No AR means E2 rules unopposed.
• Swyer = nothing without HRT. No gonad, no estrogen.
• Aromatase deficiency = no breasts in either sex. The factory is broken.
• Mayer-Rokitansky-Küster-Hauser syndrome (MRKH) = breasts and hair, just no uterus. Ovaries fine.

AROMATASE DEFICIENCY: TALL STATURE + OSTEOPOROSIS DESPITE TESTOSTERONE. This is the single most important teaching point in reproductive endocrinology:

IT IS ESTROGEN, NOT TESTOSTERONE, THAT CLOSES GROWTH PLATES AND MAINTAINS BONE IN BOTH SEXES.

The discovery came from male patients with CYP19A1 LoF (and separately ESR1 LoF): normal/high T, virilization, but CONTINUED LINEAR GROWTH into adulthood, UNFUSED EPIPHYSES on radiographs, and SEVERE OSTEOPOROSIS.

WHY TALL:

• Epiphyseal fusion requires E2 acting on ERα in growth plate chondrocytes.
• T alone CANNOT do this — it must be AROMATIZED to E2.
• No aromatase (or no ERα) -> growth plates never fuse -> linear growth continues into the 20s and 30s -> eunuchoidal proportions (long limbs, arm span > height).

WHY OSTEOPOROSIS:

• Estrogen is the DOMINANT regulator of bone in BOTH sexes: * Suppresses receptor activator of NF-κB ligand (RANKL) -> restrains osteoclast resorption. * Promotes osteoblast and osteocyte survival. * Maintains bone formation/resorption coupling.
• T supports bone partly via direct AR action, BUT a major share of T's bone effect is via LOCAL AROMATIZATION TO E2 in osteoblasts.
• Without aromatase: T can't be locally converted, RANKL unrestrained, bone resorption accelerates -> low BMD despite normal/high serum T.

TREATMENT IN MALE AROMATASE-DEFICIENT PATIENTS IS, PARADOXICALLY, LOW-DOSE ESTROGEN. It closes growth plates, restores BMD, normalizes lipids, improves insulin sensitivity. Same is true in ESR1 mutation — except E2 doesn't work because the receptor is broken; those patients have very limited options.

USEFUL COROLLARY: in estrogen-resistant male patients with severe osteoporosis, you can have VERY HIGH serum E2 and still bone disease — because the receptor doesn't sense it. The hormone level alone doesn't tell you anything; the ACTION does.

THE MIRROR — AROMATASE EXCESS SYNDROME (CYP19A1 gain / promoter rearrangement, AD): too MUCH peripheral aromatization -> prepubertal GYNECOMASTIA in boys (early breast + advanced bone age in girls), ADVANCED bone age, and SHORT final height from premature epiphyseal fusion — the photographic negative of aromatase deficiency, and proof the same enzyme sets plate closure in both directions. Treat with an aromatase inhibitor.

ESTROGEN-AND-BONE SYMMETRY POINTS:

• Postmenopausal osteoporosis: same mechanism, accelerated bone resorption from loss of estrogen-mediated RANKL suppression.
• Klinefelter osteoporosis (despite male karyotype): low T + low T-to-E2 conversion + hypogonadal axis.
• POI in girls: same mechanism, hypoestrogenic bone loss before peak bone mass is reached -> highest fracture risk later.
• The reason estradiol is the bone-modifying hormone clinically: bisphosphonates / denosumab work by blocking the SAME osteoclast pathway estrogen suppresses physiologically.

ADRENAL vs GONADAL STEROIDOGENESIS — the framework that makes every congenital adrenal hyperplasia (CAH) and DSD pathway interpretable. Both glands share the same trunk: cholesterol -> StAR (mitochondrial import) -> CYP11A1 (side- chain cleavage) -> pregnenolone. From there, what enzymes the tissue expresses decides the product:

Reticularis differs from fasciculata even though both are ACTH-driven: b5 boosts 17,20-lyase, low 3β-HSD2 traps steroids on the Δ5 side, SULT2A1 makes DHEAS (long half-life). This zone WAKES at adrenarche (6-8 y) and supplies the prepubertal adrenal androgen signal.

Two-cell, two-gonadotropin ovary: theca makes androgen, granulosa aromatizes. Neither cell makes E2 alone.

SEX ASYMMETRY OF ADRENAL ANDROGENS — the single most clinically useful corollary:

So adrenal hyperandrogenism conspicuously virilizes a girl, barely shows in a boy. CAH virilizes 46,XX dramatically; in 46,XY the testicular T pool already dominates so the same enzymatic block reads out differently. Premature adrenarche is the commonest "first virilizing event" in girls — and unremarkable in boys.

CYP21A2 LoF -> cortisol low, ACTH up, 17-OHP up, androgen shunt up, aldosterone variable. The CYP21A2 locus is next to a non-functional pseudogene CYP21A1P; high gene conversion rate explains ~95% of disease alleles being recurrent.

Phenotypes by residual activity:

• Salt-wasting (<1%): 46,XX virilized at birth (Prader 2-5), 46,XY normal genitalia. Salt crisis 1-3 wks (vomiting, hyponatremia, hyperkalemia, hypoglycemia, shock).
• Simple-virilizing (1-2%): 46,XX virilized, 46,XY presents later with precocious pseudopuberty + advanced bone age.
• Non-classic (20-50%): premature pubarche, hirsutism, oligomenorrhea in adolescence.

The 46,XX SW infant whose virilization is missed at birth is the ICU disaster: a "boy" with bilateral cryptorchidism turns out to be a 46,XX female with shock and hyperkalemia.

Acute crisis:

• IM hydrocortisone (25 mg <1y, 50 mg 1-5y, 100 mg >5y) if no IV available, then 50-100 mg/m² IV bolus.
• 0.9% saline 20 mL/kg bolus, repeat.
• D10 to address hypoglycemia.
• Avoid hypotonic fluids until sodium normalizes.

Chronic:

• Hydrocortisone 8-10 mg/m²/day PO TID. Lower is better (avoid iatrogenic Cushing, growth failure). Short-acting HC in growing children; save prednisolone/dexamethasone for after final height.
• Fludrocortisone 0.05-0.2 mg morning, titrate to renin. ALL classic (salt-wasting AND simple-virilizing) need it; SV is a spectrum.
• NaCl 1-2 g/day in infants until solids.

TREATMENT GOALS — aim for CONTROL, not normalization (driving the markers to normal = over-treatment):

• 17-OHP HIGH-NORMAL to mildly elevated (a fully normal 17-OHP means too much glucocorticoid); androstenedione / testosterone normal-for-age; renin mid-normal (the mineralocorticoid-adequacy check).
• GROWTH VELOCITY + BONE AGE + weight are the clinical readout of the under- vs over-treatment balance — the single best long-term gauge.
• Stress-dose for illness/surgery (§12); emergency IM HC kit.

CRINECERFONT (FDA-approved, Sarafoglou NEJM 2024) is the first non-glucocorticoid axis-modifier. It's a CRF1 receptor antagonist that blocks ACTH drive at the pituitary. In the pediatric CAHtalyst trial it let glucocorticoid dose drop ~18% while maintaining androgen control. NOT a substitute for hydrocortisone, does NOT cover stress.

TART — TESTICULAR ADRENAL REST TUMORS. Ectopic adrenal cortex trapped in the testis at descent, dormant unless chronic ACTH excess makes them proliferate.

• Setting: poorly-controlled CAH (21-OHD by far most common, prevalence 30-94% in adolescent/adult CAH males), 11β-OHD, 3β-HSD2, untreated Addison, Nelson syndrome.
• Behavior: ACTH-responsive (shrink with glucocorticoid intensification); secrete the patient's ACTH-driven adrenal steroids — 11-OXYGENATED ANDROGENS (11-OH-androstenedione, 11-ketotestosterone: active, gonadotropin-independent) PLUS, in 21-OHD, the disease markers 17-OHP and 21-deoxycortisol. They do NOT make true testicular T. (Note: 21-deoxycortisol and 11-deoxycortisol are CAH precursor markers, NOT 11-oxygenated androgens — different classes.)
• Location: bilateral ~70%, at the rete testis / hilum.
• Clinical bite: compress seminiferous tubules -> obstructive azoospermia + infertility in CAH adults. Long-standing compression is IRREVERSIBLE. Surveillance scrotal US from adolescence in CAH boys, especially with poor control.

TART vs LEYDIG CELL TUMOR:

Discrimination matters: Leydig tumor = orchiectomy. TART = INTENSIFY GLUCOCORTICOID first; testis-sparing enucleation only for refractory cases. Female counterpart OARTs (ovarian adrenal rest tumors) exist but are vanishingly rare.

2026 ENDO-ERN on 46,XX surgery: delayed/individualized, MDT, truthful disclosure, surgical decisions can wait for patient participation. Emergent obstructed urogenital sinus excepted.

DIAGNOSTIC CUTOFFS for 21-OH CAH — this is where people fudge numbers and miss non-classical disease.

Baseline morning 17-OHP at 8 AM (the only time it makes biological sense to measure; afternoon levels are unreliable):

ACTH stim test (250 mcg cosyntropin IV, measure 17-OHP at 60 min) is the gold standard, especially for non-classical:

SW vs SV is NOT distinguished by 17-OHP. Both have similarly stratospheric values. The distinction is the MINERALOCORTICOID axis:

So the workup move at the bedside: a 46,XX virilized newborn or a shocked salt-losing infant gets electrolytes + renin + aldosterone + 17-OHP all at once. SW is identified by the renin/aldo axis, not by 17-OHP alone.

Newborn screening: dried blood spot at 48-72 h. Cutoffs MUST be stratified by gestational age and birth weight because preterm infants have physiologically higher 17-OHP. Pre-stratification false-positive rates in preterms were 30-50%, mostly transient stress-related elevation. Most programs now use tiered cutoffs (e.g., <1500 g, <34 wk, 34-37 wk, term) and re-test in the LBW / preterm group rather than calling them positive. Two-tier screening with LC-MS/MS confirmation — adding 21-deoxycortisol and the (17-OHP + 21-deoxy) / cortisol ratio — dramatically cuts false positives further. False NEGATIVES happen with antenatal or perinatal glucocorticoid exposure (suppresses 17-OHP transiently).

Congenital adrenal hyperplasia (CAH) genetics is the part most people skip and then get burned by. 21-OH deficiency is the rare endocrine disease where the PSEUDOGENE is the engine of disease. You can't talk CAH genetics without RCCX.

The module sits at chr 6p21.3 inside MHC class III, usually BIMODULAR (two tandem copies). The diagram is the disease in one picture: the functional CYP21A2 sits one column from a near-identical pseudogene (CYP21A1P, ~98% identity, ~30 kb upstream), and that homology is the engine. Two mechanisms drive disease:

GENE CONVERSION is the dominant mechanism (~75% of disease alleles). Non-reciprocal transfer of mutated sequence from CYP21A1P (pseudogene) into CYP21A2 (functional). The pseudogene already carries inactivating variants because it's not under selection; conversion lands them in the functional gene.

LARGE DELETIONS account for most of the rest (~20-25%). Unequal crossover during meiosis between the pseudogene and the functional gene deletes CYP21A2 entirely (along with C4 in some cases).

RCCX COPY NUMBER — the same unequal crossover also changes the module COUNT: ~70-80% of chromosome-6 copies are BIMODULAR (the diagram), but MONO-, TRI-, and QUADRIMODULAR arrangements occur — one chromosome loses a module, its partner gains one. THE GENOTYPING TRAP: a chromosome can end up carrying a DUPLICATED CYP21A2 — a normal copy IN CIS with a mutated one — which is NOT pathogenic (the normal copy still makes enzyme) but looks like a disease allele on sequencing alone. A CYP21A2 variant with PRESERVED gene dosage on MLPA may be a benign duplication, not CAH — copy-number / phasing is what settles it (Genetic Steroid Disorders, 2e 2023). Crossover at C4 moves only the pseudogene (no CAH); it is crossover at CYP21 or TNX that removes the active gene.

The sharpest case of that trap is Q318X (p.Gln318*; HGVS p.Gln319Ter): a null, salt-wasting stop codon as a LONE allele — but it preferentially rides IN CIS on a DUPLICATED functional CYP21A2 (a founder haplotype, HLA-B*50-linked) in a large share of carriers, where the intact copy keeps 21-OH activity and the allele is NON-pathogenic. Sequencing alone cannot tell the two apart, so a solitary Q318X OVER-CALLS carriers and patients — reflex to MLPA / copy-number before you counsel or start cascade testing (EMQN). The variant stays pathogenic on its own; it is only benign when a second functional copy sits beside it.

The clinical consequence: the same set of recurrent pathogenic variants accounts for ~95% of disease alleles. You see the same genotypes over and over.

The GENOTYPE-PHENOTYPE RULE: the LESS SEVERE allele determines the phenotype, because you only need one allele making enzyme to set the residual-activity ceiling. So:

• Two severe alleles (e.g., I2G/I2G, deletion/I2G) -> salt-wasting.
• One severe + one moderate (e.g., I2G/I172N) -> simple-virilizing.
• One severe + one mild (e.g., deletion/V281L) -> non-classic.
• Two mild alleles -> non-classic.

Carrier frequency is high. ~1 in 50-60 in the general population carries a classical CAH allele. Non-classical alleles (V281L, P30L) are even more frequent — ~1 in 5-15 in some populations (Ashkenazi Jews, Yupik, Hispanic).

C4 connection worth knowing:

• C4 deletions on the RCCX module are associated with SLE susceptibility (~50% of classical CAH patients carry a C4 deletion). The CAH-X / TNXB story — 21-OH CAH + hypermobile Ehlers-Danlos — is detailed below.

Genetic testing approach: targeted CYP21A2 sequencing covering the recurrent variants catches ~95% of cases. LONG-RANGE PCR is needed to distinguish CYP21A2 from CYP21A1P (standard PCR primers can amplify the pseudogene and miss real disease alleles). MLPA for deletions. Whole-gene sequencing for unusual cases.

CAH-X TANDEM DELETION ARCHITECTURE — RCCX module has a fragile gene-order (deep intron of CYP21A2 partially overlaps with start of TNXB). Large deletions sometimes take out BOTH CYP21A2 AND adjacent TNXB -> 21-OH CAH + TNXB-deficient connective tissue disorder (hypermobile Ehlers-Danlos-like phenotype): joint hypermobility, soft skin, easy bruising, GI dysmotility, dysautonomia / POTS. Score the hypermobility with the standard Beighton score (9-point scale; ≥5 = generalized joint hypermobility, age-adjusted — ≥6 prepubertal, ≥4 if >50 yr). ~5-10% of classical 21-OH CAH cohorts carry the tandem deletion when specifically screened. Practical: in any 21-OH CAH child with unexplained joint pain, hypermobility, or chronic GI complaints -> test TNXB sequencing / MLPA for the CAH-X allele.

The COMMONEST cause of adrenal insufficiency overall is IATROGENIC — chronic exogenous glucocorticoid then abrupt withdrawal (a suppressed HPA axis). Among PRIMARY causes, CAH dominates in infancy while autoimmune Addison is the commonest ACQUIRED primary AI in older children.

Primary AI: cortisol low, ACTH high, aldo low/renin high, hyperpigmented.

DIAGNOSIS: morning cortisol >18 μg/dL (500 nmol/L) excludes AI, 250 ug cosyntropin stim (peak <18 μg/dL [500 nmol/L] = PAI); the low-dose 1 ug test is more sensitive for partial SECONDARY AI. ACTH high in PAI / low in SAI; renin high in PAI once mineralocorticoid is lost; 21-OH Ab confirms autoimmune Addison; very-long-chain fatty acids (VLCFA) mandatory in a boy.

In a boy with isolated primary AI, ALWAYS SEND VLCFA before labeling autoimmune. X-ALD (ABCD1 LoF) — defective peroxisomal VLCFA import, boys only. Two pictures: childhood cerebral adrenoleukodystrophy (ALD) age 4-10 (progressive demyelination, untreated fatal); adrenomyeloneuropathy AMN in adulthood. Addison can precede neurology by years, and there is NO genotype-phenotype correlation — so every X-ALD boy needs SERIAL BRAIN MRI surveillance to catch cerebral conversion in its treatable window: HSCT (or elivaldogene autotemcel gene therapy) only helps in EARLY / pre-symptomatic cerebral ALD; once it is neurologically advanced, that window is gone. Increasingly caught pre-symptomatically since X-ALD was added to RUSP in 2016 — many states now do VLCFA screening on the newborn DBS, so the boy with isolated AI later in childhood is becoming less common, replaced by the boy diagnosed at birth before any phenotype.

APS-1 (APECED, AIRE LoF, AR) — TRIAD IN TIME ORDER (the order is the clue): chronic mucocutaneous candidiasis (age 0-5, the first sign — neutralizing anti-IL-17/IL-22 autoantibodies knock out Th17 antifungal defense) -> hypoparathyroidism (age 5-15, anti-NALP5) -> Addison (age 10-20, anti-21-hydroxylase / anti-side-chain-cleavage). AIRE drives promiscuous self-antigen expression in medullary thymic epithelial cells for central tolerance; LoF collapses negative selection. Plus ECTODERMAL features (enamel hypoplasia, nail dystrophy, alopecia, vitiligo, keratoconjunctivitis) and accumulating autoimmunity (T1DM ~20%, POI from anti-side-chain-cleavage cross-reacting with ovarian steroidogenesis, pernicious anemia, autoimmune hepatitis). Ruxolitinib / baricitinib (JAK inhibitors) emerging for IFN-driven disease.

APS-2 (Schmidt, HLA-linked): Addison + autoimmune thyroid + T1DM.

FGD (familial glucocorticoid deficiency): isolated cortisol deficiency, MR axis preserved.

• FGD-1 MC2R (ACTH receptor).
• FGD-2 MRAP (MC2R accessory protein).
• FGD-3 StAR partial.
• FGD-4 NNT (mitochondrial NADPH).
• FGD-5 TXNRD2.

AHC (adrenal hypoplasia congenita) — the OTHER "primary AI in a boy" to put beside X-ALD: DAX1 / NR0B1 (X-linked) = primary AI in infancy PLUS, the discriminator, HYPOGONADOTROPIC hypogonadism that surfaces at expected puberty; SF1 / NR5A1 is the autosomal counterpart. And Wolman disease / LAL deficiency (LIPA): infantile primary AI with ADRENAL CALCIFICATION + hepatosplenomegaly + malabsorption — the classic "bilateral adrenal calcification + AI in infancy".

Triple-A / Allgrove (AAAS, ALADIN): achalasia + alacrima + Addison. Alacrima from infancy is the earliest clue.

CHROUSOS SYNDROME (primary GENERALIZED GLUCOCORTICOID RESISTANCE, NR3C1 / GR LoF, AD with variable penetrance, also rare AR forms). Mechanistically the opposite of every story so far — the adrenal works fine, cortisol is HIGH, ACTH is HIGH, but the glucocorticoid receptor doesn't translate signal into action. The body keeps pushing ACTH up trying to overcome the unresponsive GR. The lab pattern mimics Cushing, the phenotype DOES NOT.

The four clinical pillars: 1. HIGH cortisol (UFC, serum, salivary) + HIGH ACTH that fails to suppress on LD-DST. 2. NO Cushingoid phenotype — the discriminator. Normal growth, no moon face, no striae, no central obesity, no muscle wasting. The GR-mediated effects of cortisol are blunted. 3. HYPERTENSION + HYPOKALEMIA. The cortisol spillover overwhelms 11β-HSD2 and illegitimately activates MR ("apparent mineralocorticoid excess" mechanism, but driven endogenously by cortisol excess rather than HSD11B2 deficiency). Suppressed renin and aldo. 4. HYPERANDROGENISM in females. High ACTH co-stimulates adrenal androgen production (DHEAS, androstenedione, testosterone) -> hirsutism, acne, oligomenorrhea, virilization in girls; precocious adrenarche or oligo-azoospermia in boys.

Distinguish from Cushing by the missing phenotype (no growth failure, no central obesity, no striae) and the resistant pattern of the LD-DST (no suppression at low dose AND no suppression at high dose, because the GR is structurally unresponsive — "neither suppressed by low nor high dex" with no clinical Cushingoid traits).

Treatment is mechanistically clever: HIGH-DOSE DEXAMETHASONE (1-3 mg/day in adults, weight-banded in peds). Dex has lower MR cross- reactivity than cortisol and can saturate the partially-functional mutant GR to suppress ACTH. Suppressing ACTH collapses both the cortisol output AND the adrenal androgen surge, which fixes both the HTN/hypokalemia and the virilization. ADD an MR ANTAGONIST (spironolactone, eplerenone) if HTN persists. Hyperandrogenism not responding to dex alone may need a peripheral anti-androgen.

The point: cortisol HIGH + ACTH HIGH + NO Cushing phenotype + HTN + hirsutism = think Chrousos, sequence NR3C1.

STRESS DOSING (drill this into families):

• Maintenance: HC 8-12 mg/m²/day PO TID.
• Minor (low fever, mild URI no vomiting): 2x maintenance 24-48h.
• Moderate (fever ≥38.5, vomit once, minor procedure): 3x maintenance.
• Severe (sepsis, persistent vomiting, surgery under GA, major trauma): 5-10x maintenance, IV/IM, admit.
• Vomiting -> IM hydrocortisone (age-banded 25/50/100 mg) then ER.
• Major surgery: HC 50-100 mg/m² IV at induction, then continuous.

Every family: emergency IM kit + written plan + medical alert + school plan + annual injection training.

DHEAS >800 IN A GIRL — the adrenal-tumor workup pulled by the lab result alone. DHEAS is essentially adrenal-only (zona reticularis + SULT2A1; ovary doesn't make it). Half-life 16-20 h vs DHEA's 30 min makes a single morning value reliable. DHEAS >700-800 mcg/dL flags adrenal pathology — particularly tumor — and changes the workup from "PCOS panel" to "rule out ACC".

PAPSS2 DEFICIENCY — the "low DHEAS + hyperandrogenism" mirror trap (NOT to confuse with PAPP-A2, which is the GH/IGF-axis disorder in §38). PAPSS2 (3'-phosphoadenosine-5'-phosphosulfate synthase 2) makes PAPS, the universal sulfate donor. SULT2A1 in the zona reticularis needs PAPS to convert DHEA -> DHEAS. Lose PAPSS2 and the sulfation step fails:

• Reticularis still makes DHEA in normal/high amounts.
• DHEA can't be parked as DHEAS (which would be metabolically inert with a long half-life).
• Free DHEA spills peripherally and gets converted to androstenedione and testosterone by 3β-HSD and 17β-HSD.
• Net = HYPERANDROGENISM with LOW or inappropriately normal DHEAS.
• A second axis: PAPSS2 also sulfates chondroitin / heparan-sulfate GAGs in cartilage, so deficiency also causes BRACHYOLMIA — a spondyloepimetaphyseal dysplasia with short trunk, short stature, and platyspondyly.

Clinical phenotype in adolescent girls: premature pubarche, hirsutism, acne, oligomenorrhea, PCOS-like picture — PLUS the skeletal signal (short stature with short trunk, mild platyspondyly on lateral spine x-ray). The lab discriminator from PCOS / PMOS: DHEA HIGH, DHEAS LOW or low-normal, T mildly elevated, androstenedione elevated. The DHEA to DHEAS RATIO is the diagnostic giveaway.

Treatment: anti-androgens (spironolactone), OCP for cycle control, no specific PAPS replacement available. Bone surveillance for the brachyolmia.

Workup logic:

• TEMPO: rapid virilization (weeks-months) -> ACC strongly.
• Cushingoid signs + virilization + HTN + growth deceleration -> mixed ACC.
• Family history: TP53 spectrum cancers (Li-Fraumeni), or hemihypertrophy/macroglossia/abdominal wall defect (BWS).
• Biochemistry: total + free T, A, 17-OHP (basal + ACTH-stim if intermediate), DHEA alongside DHEAS, 11-deoxycortisol, 8 AM cortisol + ACTH, overnight 1-mg DST. LC-MS/MS steroid metabolomics is highly sensitive for ACC (chaotic mixed pattern).
• Imaging: ADRENAL MRI (preferred over CT in peds). Pelvic US to evaluate ovaries. If a CT is the study you have, read the HOUNSFIELD UNITS (HU): water = 0, air = -1000, fat negative, soft tissue ~+20-70. An unenhanced attenuation ≤10 HU = LIPID-RICH -> benign adenoma (high specificity). >10 HU is indeterminate -> washout CT (benign adenomas wash out fast: absolute >60% / relative >40%) or MRI chemical-shift. The ACC flags are the OPPOSITE: high HU (>10, often >>20), heterogeneous, slow washout, >4 cm, calcification/necrosis. Pheo is also >10 HU and avidly enhancing — don't call it benign on density alone; myelolipoma gives away MACROSCOPIC FAT (markedly negative HU).
• Genetics: TP53 in any pediatric ACC; methylation studies for BWS if overgrowth.

This is where people get burned. They are completely different proteins in completely different parts of the body.

11-beta-hydroxylase (CYP11B1) is an adrenal cortex steroidogenic enzyme. LoF = congenital adrenal hyperplasia (CAH), blocks 11-deoxycortisol -> cortisol AND DOC -> corticosterone. ACTH-driven 11-deoxycorticosterone (DOC) excess causes HTN with hypokalemic alkalosis. Adrenal androgens shunt up, so 46,XX virilizes at birth. Treatment: glucocorticoid to suppress ACTH and shut off DOC.

11-beta-HSD2 (HSD11B2) is in the kidney distal nephron. Its job is to oxidize cortisol to cortisone locally, so cortisol cannot occupy MR. Cortisol binds MR with the same affinity as aldosterone but its circulating level is 100-1000x higher, so without 11bHSD2 cortisol jams MR open. That is AME — severe HTN with hypokalemic alkalosis, low renin, LOW aldosterone (volume expanded). Urinary tetrahydrocortisol / tetrahydrocortisone ratio is elevated. Licorice phenocopies. Treatment is MR antagonism (spironolactone, eplerenone, amiloride). Dexamethasone is no longer first-line but low-dose dex still appears as an adjunct in some refractory cases.

Geller MR S810L is a different angle on the same receptor: progesterone becomes a partial agonist at the mutant MR. Pregnancy decompensates with severe HTN as progesterone surges 100-fold. Spironolactone is ALSO a partial agonist at the mutant — contraindicated. Use amiloride.

JG-cell renin release has three triggers: low afferent stretch (baroreceptor), low distal NaCl (macula densa), and BETA-1 ADRENERGIC. NOT alpha-1. This matters because beta-blockers lower renin precisely because they antagonize beta-1.

Renin -> AT-I -> ACE -> AT-II -> AT1R: vasoconstriction, sympathetic facilitation, aldo synthesis, ADH release, vascular trophic effects. Aldosterone in distal nephron drives ENaC/ROMK/Na-K-ATPase. Na in, K out, H out.

THE ALDOSTERONE-TO-RENIN RATIO (ARR) is the screening test for PRIMARY aldosteronism (the aldo-excess HTN that is NOT renin-driven). Logic: in primary aldosteronism aldosterone is HIGH while renin is SUPPRESSED (the volume expansion shuts renin off), so the RATIO climbs even when aldo looks "normal." A HIGH ARR (high aldo + low/suppressed renin) = screen positive -> confirm with a salt-loading / saline-suppression test, then localize (adrenal imaging +/- adrenal venous sampling). Read it against the two mirrors:

• HIGH aldo + LOW renin + HIGH ARR = PRIMARY aldosteronism (Conn adenoma, bilateral hyperplasia, GRA/FH).
• LOW aldo + LOW renin (low ARR by low numerator) = the NON-aldosterone mineralocorticoid HTNs — Liddle, AME, Geller, Gordon/PHA2 (the grid below).
• HIGH aldo + HIGH renin = SECONDARY (renovascular, diuretic, the salt-wasting states) — renin is driving aldo, so the ratio stays low.

Pitfalls before you trust an ARR: correct hypokalemia first (low K suppresses aldo -> false-negative), and remember most antihypertensives skew it (MR antagonists + ACEi/ARB RAISE renin -> false-negative; beta-blockers LOWER renin -> false-positive). Conn syndrome is under-diagnosed in adolescents — screen a hypertensive teen with spontaneous or easily-provoked hypokalemia.

LIDDLE = ENaC GoF. Mutations in the PY motif of SCNN1A/B/G prevent Nedd4-2 from ubiquitinating and internalizing ENaC. The channel sits at the apical membrane forever. HTN with low K, low renin, LOW aldosterone. Spironolactone fails because the lesion is downstream of MR. Use amiloride or triamterene.

GELLER = MR S810L. See above. Suppressed aldo. Spiro contraindicated.

GORDON (PHA2): the one HYPERKALEMIC mineralocorticoid HTN — described in full with its salt-wasting PHA-1 sibling in the PHA family below.

Glucocorticoid-remediable aldosteronism (GRA) / FH-I = chimeric CYP11B1-CYP11B2 from unequal 8q crossover. The chimera has the ACTH-responsive promoter of 11B1 fused to the coding sequence of aldosterone synthase. Aldosterone is now ACTH-driven and made in the zona fasciculata. Hybrid steroids (18-OH-cortisol, 18-oxo-cortisol) on mass spec. Treat with low-dose dexamethasone or amiloride/eplerenone.

AME = HSD11B2 LoF. Already covered.

THE TWO HYPERTENSIVE CAHs both pile up DOC (-> low-renin HTN + hypokalemia) but split on SEX STEROIDS — that split is the discriminator:

Same HTN entry point, OPPOSITE gonadal readout: 11β VIRILIZES (46,XX), 17α UNDER-virilizes BOTH sexes. (Contrast the salt-WASTING CAHs — 21-OH, lipoid — where the block starves the mineralocorticoid arm instead.)

PSEUDOHYPOALDOSTERONISM (PHA) — the distal nephron behaves as if there is NO aldosterone, so EVERY PHA runs HYPERKALEMIC with metabolic acidosis. The fork is BLOOD PRESSURE: PHA-1 can't hold salt (salt-wasting, LOW / normal BP), PHA-2 hoards it (HYPERTENSION). Take them in order.

PHA-1 = TRUE aldosterone resistance — renin AND aldosterone both run HIGH (the axis is screaming; the distal nephron is deaf). It comes in two forms that are almost diseases of different organs, and which one you have decides whether you tell the family "this self-resolves" or "this is lifelong":

PHA-1 RENAL (AD, NR3C2 / MR loss-of-function):

• One bad MR allele is enough — the aldosterone signal that reaches the distal nephron is sub-threshold.
• RENAL ONLY: sweat, saliva, colon, airway are normal because ENaC itself is intact and other MR-using tissues compensate.
• Neonatal salt wasting, hyperkalemia, metabolic acidosis, poor weight gain.
• SELF-RESOLVES across the first 1-3 years as the kidney matures and recruits other Na transporters; many adults are asymptomatic carriers. Treat with oral salt during infancy only — no lifelong therapy.

PHA-1 SYSTEMIC (AR, SCNN1A / B / G biallelic LoF) = the mirror of Liddle: loss-of-function in ENaC itself, not the receptor upstream. ENaC won't open, so the kidney can't reabsorb sodium AND the same ENaC sitting in salivary, sweat, colonic and respiratory epithelia makes the salt wasting MULTI-ORGAN: severe neonatal salt wasting, recurrent respiratory infections (impaired airway surface fluid clearance, a pseudo-CF picture), high sweat chloride and salty kisses, hypercalciuria with stones later. Sky-high aldosterone and renin (the kidney is screaming at MR, but the door downstream won't open). LIFELONG, no escape — the defect is in the channel itself, not a developmentally compensable receptor. Treat with high-dose sodium (often grams of NaCl/day), potassium-binding resins or dietary K restriction, indomethacin or thiazide adjuncts; fludrocortisone DOES NOT WORK (it would in the renal form, where the channel is intact).

Short version: AD/renal PHA-1 = receptor problem, kidney only, self-resolves; AR/systemic PHA-1 = channel problem, every wet surface, lifelong.

PHA-2 = GORDON SYNDROME — the MIRROR of PHA-1 on blood pressure (and the anti-Gitelman). Mutations in WNK1 / WNK4 / KLHL3 / CUL3 all converge on an OVERACTIVE NCC: the DCT grabs Na before it can reach ENaC, so volume expands (HYPERTENSION) while distal K and H secretion fall (the same hyperkalemia + metabolic acidosis as PHA-1) — the one HYPERKALEMIC mineralocorticoid HTN, with SUPPRESSED renin (volume-expanded). Low-dose THIAZIDE blocks the overactive NCC and is both the treatment and the confirmation.

PRIMARY ALDOSTERONISM IN PEDS is rare but not zero. Sporadic aldosterone-producing adenoma (APA) with somatic KCNJ5 mutations described in adolescents. Familial forms: FH-I = GRA (above), FH-II (CLCN2, AD), FH-III (KCNJ5 germline, severe early-onset), FH-IV (CACNA1H); plus CACNA1D — somatic in APAs and germline in PASNA (primary aldosteronism + seizures + neurologic abnormalities), the early-onset calcium-channel form. Workup: confirm aldo:renin ratio elevated, salt-loading or captopril challenge, adrenal CT/MRI, adrenal venous sampling for lateralization in selected cases. Conn syndrome is not just an adult disease.

Two distinct vocabularies. "Cushing syndrome" is the umbrella for any glucocorticoid excess. "Cushing disease" is specifically a pituitary ACTH-secreting adenoma. Don't conflate them.

The pediatric red flag is the one adult patients don't have: WEIGHT GAIN WITH GROWTH FAILURE. Adult Cushing presents with muscle wasting, skin thinning, classic striae. Kids present with crossing height percentiles downward while gaining weight. If the height curve is falling and the weight curve is climbing, Cushing is on the differential until proven otherwise. Striae, hirsutism, pubertal arrest, hypertension, glucose intolerance, easy bruising fill in the picture.

Workup is two steps, and the ORDER is the rule: first prove the hypercortisolism is real (step 1), THEN ask where it comes from (step 2). Measuring ACTH before step 1 is positive is a trap — the ACTH split only means something once hypercortisolism is confirmed.

Step 1 — confirm hypercortisolism. No single result is definitive; you want at least TWO different tests abnormal. The screens and the cutoffs worth memorising:

A "positive" screen is failure to suppress, or a value above the Cushing threshold; the late-night salivary, UFC, and overnight 1 mg dex are the usual first-line trio.

LOSS OF THE DIURNAL NADIR is the earliest change — the late-night cortisol stops falling before the mean rises. The late-night salivary and midnight serum cortisol above ARE that rhythm test; the older paired 08:00 + 23:00 serum cortisol (an awake late-night value that fails to fall well below the morning level) reads the same physiology. Unstimulated 8 AM ACTH is NOT a confirmatory test — it is the step-2 splitter, so hold it until step 1 is positive.

Pseudo-Cushing is the trap — depression, anorexia, severe obesity, alcohol, poorly-controlled diabetes can all push UFC up. The dex-CRH test, response to treating the underlying condition, and ABSENCE of the classic phenotype (growth failure, striae) help distinguish.

Step 2 — localize. ONE axis runs the whole differential: is the cortisol ACTH-DEPENDENT or not. Orient on this mirror, then nail it down with the cutoffs below:

ACTH SUPPRESSED (<10) -> adrenal source (ACTH-independent):

ACTH NORMAL OR HIGH -> ACTH-dependent:

PITUITARY (CENTRAL) vs ECTOPIC ACTH — the labs split them cleanly:

IPSS is the gold-standard localization test when pituitary MRI is non-diagnostic in ACTH-dependent Cushing. Catheters in both petrosal sinuses sample ACTH simultaneously with a peripheral draw; CRH is given as a provocation. A gradient confirms central source AND lateralizes the adenoma (left vs right pituitary).

PARADOXICAL DEX RESPONSE — the high-yield PPNAD twist. On the 2-day Liddle test (Grant Liddle, not the syndrome: low-dose then high-dose dex q6h), normal subjects suppress and Cushing disease partially suppresses on the high dose. But in PRIMARY PIGMENTED NODULAR ADRENOCORTICAL DISEASE (PPNAD, part of Carney complex via PRKAR1A), urinary cortisol metabolites PARADOXICALLY INCREASE on high-dose dex — the autonomous micronodules read dexamethasone as an agonist (aberrant GR / PKA signaling). A paradoxical RISE = think PPNAD / Carney first; pair with the phenotype (cardiac myxoma, lentigines on lips and conjunctiva, LCCSCT in boys, schwannomas, acromegaly) and order PRKAR1A.

ABERRANT RECEPTORS IN BMAH / AIMAH — the high-yield twist for ACTH-independent Cushing with bilateral big nodular adrenals. ACTH is suppressed, yet cortisol is high. What's driving the cortisol? The adrenal cells aberrantly express GPCRs that don't belong there, and those receptors hijack the steroidogenic machinery instead of the silent MC2R. Each receptor gives a stereotyped clinical pattern that you can elicit at the bedside:

• GIP RECEPTOR (gastric inhibitory peptide) -> FOOD-DEPENDENT Cushing. Cortisol rises after every meal because GIP rises post-prandially. Test: morning fasting cortisol normal, post-prandial cortisol high.
• V1 VASOPRESSIN RECEPTOR -> POSTURAL Cushing. Cortisol rises on standing because AVP rises with orthostasis. Test: supine vs upright cortisol.
• LH / hCG RECEPTOR -> pregnancy- and menopause-onset Cushing in females (pregnancy hCG; post-menopausal LH surge). New-onset in late reproductive years is the clinical flag.
• BETA-ADRENERGIC RECEPTOR -> catecholamine-dependent Cushing. Stress / exercise / isoproterenol provokes cortisol. Propranolol paradoxically suppresses.
• 5-HT4, ANGIOTENSIN II, GLP-1 RECEPTOR variants described.

The clinical move: in any ACTH-independent BMAH, run a structured provocation panel (overnight fast morning cortisol, mixed meal, upright posture, GnRH stim, metoclopramide for 5-HT4, ACTH stim). A positive provocation lets you suppress cortisol with the matching antagonist (somatostatin analog for GIP, V1 antagonists, beta-blocker, GnRHa) and can BUY TIME or avoid bilateral adrenalectomy. The diagnosis also dictates surveillance for the secondary axis (LH/hCG-driven BMAH needs gonadal source workup).

Treatment workflow for hypercortisolism, separated from the diagnostic differential in §16 for length. Cross-ref §16 for DHEAS workup and the BMAH aberrant-receptor differential.

Treatment by source:

• Pituitary Cushing -> transsphenoidal surgery at a high-volume pediatric center. Cure 70-85%. Recurrence 10-20% at 10 years.
• Adrenal adenoma -> unilateral adrenalectomy. Cure usually complete.
• ACC -> radical resection + mitotane + cytotoxic chemotherapy (cisplatin / etoposide / doxorubicin). Stage and completeness drive prognosis.
• Bilateral disease (PPNAD, MMAD) -> bilateral adrenalectomy + lifelong glucocorticoid + mineralocorticoid replacement.
• Medical therapy — group by mechanism, not alphabet: * STEROIDOGENESIS INHIBITORS (block adrenal cortisol synthesis):
• Ketoconazole (off-label, hepatotoxicity, QT, multiple cytochrome interactions). Historic workhorse.
• LEVOKETOCONAZOLE (FDA-approved): the cis-enantiomer of ketoconazole, cleaner liver and adrenal profile, less androgen suppression.
• Metyrapone: blocks 11β-hydroxylase, fast onset (hours), useful for acute pre-op cortisol drop, accumulates 11-deoxycortisol.
• OSILODROSTAT (FDA-approved): potent 11β-hydroxylase inhibitor, oral BID. Most effective single-agent today. Side effect signal: accumulation of steroid precursors and 11-deoxycorticosterone (DOC) -> hypokalemia and hypertension; aldosterone synthase also blocked at higher doses.
• Mitotane: adrenolytic (kills cortex), slow onset, mostly used in ACC. Adrenal insufficiency is the endpoint, not the side effect.
• ETOMIDATE: the only IV option — a sub-hypnotic ICU infusion that rapidly blocks 11β-hydroxylase; the rescue for severe / ectopic hypercortisolism (or a cortisol crisis) when oral therapy isn't possible. * GR ANTAGONIST — BLOCKS CORTISOL AT THE RECEPTOR:
• MIFEPRISTONE (FDA-approved, hyperglycemia indication): cortisol labs become uninterpretable (the GR is blocked, ACTH and cortisol both rise, you cannot measure cortisol to titrate — titrate by clinical effect and glucose). Endometrial hyperplasia in females. Adrenal crisis can occur with cortisol levels that look "fine".
• RELACORILANT: next-generation selective GR antagonist, no PR cross-reactivity (so no endometrial issue). NOT FDA-approved for hypercortisolism — the Cushing program did not clear the FDA's effectiveness bar (positive GRACE notwithstanding). * SOMATOSTATIN / DOPAMINE AGONISTS (pituitary source only):
• PASIREOTIDE (SSTR5 agonist, multi-receptor analog): tumor-directed in Cushing disease. Big hyperglycemia signal (suppresses incretin), counsel and screen.
• Cabergoline: adjunct for Cushing disease with residual disease after TSS; modest efficacy.

After cure, the suppressed normal HPA needs months to years to recover — universal post-cure AI requiring glucocorticoid replacement and stress dosing until ACTH stim demonstrates recovery.

Five layers, one axis: anatomy -> biochemistry -> genetics -> imaging -> prognosis.

The adrenal medulla is bathed in cortisol-rich portal venous blood from the cortex. Cortisol induces phenylethanolamine N-methyltransferase (PNMT), which converts norepinephrine to epinephrine. So adrenal pheo expresses PNMT, secretes epi and metanephrine. Extra-adrenal sympathetic PGL has no cortisol portal bath, no PNMT, secretes only norepi and normetanephrine. That is why anatomy predicts biochemistry.

PNMT is the cortisol-gated step: only the cortisol-bathed adrenal medulla makes the jump to epinephrine, so adrenal pheo is adrenergic (epi / metanephrine) while extra-adrenal PGL stays noradrenergic (norepi / normetanephrine). You assay the O-methylated metanephrines, not the catecholamines, because tumors leak them continuously.

Eisenhofer clusters organize PPGL by which signaling axis is broken — the cluster predicts location, catecholamine profile, and risk:

Pediatric PPGL is ~80% germline. Genetic testing on every proband. Two SDH-linked syndromic clusters to flag in a young patient: Carney triad (PGL + gastric GIST + pulmonary chondroma; non-germline SDHC promoter hypermethylation, young females) and the Carney-Stratakis dyad (PGL + GIST, germline SDHx).

Diagnose with plasma free or urine fractionated metanephrines (not catecholamines themselves, because tumors continuously O-methylate stored catecholamines and metanephrines are stable). 3-methoxytyramine flags dopamine-secreting cluster-1 tumors.

Imaging: MRI for anatomy (T2 "lightbulb" sign), 68Ga-DOTATATE PET as the first-line functional study now (better than MIBG for SDHx and extra-adrenal disease). MIBG retained for therapy planning. FDG-PET for metastatic SDHB.

Pre-op: ALPHA before BETA. Phenoxybenzamine 10-14 days plus salt and fluid loading (phenoxybenzamine is unavailable in Turkey — use a selective alpha-1 blocker, doxazosin or prazosin). Beta-blockade only after alpha is established, otherwise unopposed alpha-1 -> hypertensive crisis. Calcium channel blocker as adjunct.

Adequacy of blockade is the booking criterion, not the calendar. The classic Roizen criteria call pre-op alpha-blockade adequate when there is no BP above 160/90 in the 24 h before surgery, no more than ~5 ventricular ectopics/min, and — counter-intuitively — a TOLERATED mild orthostatic drop (the proof the alpha is on, so do NOT back off the dose for it). Modern targets run tighter: seated BP <130/80, and in children below the 90th percentile for age/height with heart rate in the 10th-90th percentile. For a bulky, metastatic, or blockade-resistant tumor add metyrosine (tyrosine- hydroxylase inhibitor), which cuts catecholamine synthesis at the source. The pre-op vascular bed is contracted and intravascularly DRY — salt and volume loading is what prevents the post-resection collapse. INTRA-OP catecholamine surges (tumor handling, vein ligation) are met with SHORT-ACTING, seconds-off agents ONLY — phentolamine, sodium nitroprusside, or nicardipine for the pressure spike, magnesium as an adjunct, esmolol for the tachyarrhythmia — never a long-acting drug onto a tumor about to lose its blood supply. Post-resection: hypotension once the catecholamine drive vanishes, plus rebound hyperinsulinemic hypoglycemia (watch glucose 24-48 h) — the perioperative cascade is §84.

Three circuits.

FGF23/Klotho/NaPi: osteocytes secrete FGF23 in response to phosphate, 1,25-D, and PTH. FGF23 binds FGFR1c + Klotho on the proximal tubule, internalizes NaPi-IIa and NaPi-IIc, suppresses CYP27B1, induces CYP24A1. Net: phosphate out in urine, 1,25-D drops. This is the fingerprint of every FGF23-driven hypophosphatemic rickets.

Vitamin D activation: skin 7-DHC + UVB -> D3 -> liver CYP2R1 -> 25-D -> proximal tubule CYP27B1 -> 1,25-D -> nuclear VDR. Catabolism via CYP24A1 and (in some pathology) CYP3A4.

CaSR / PTH / Gs-alpha: CaSR on parathyroid chief cells inversely controls PTH, and on thick ascending limb inhibits paracellular calcium reabsorption.

DECODE THE NAMES FIRST — every mineral acronym renames to its MECHANISM, and the look-alikes stop biting once they sit side by side. Each links to its home section.

FGF23 / phosphate-wasting rickets — the trio differs ONLY by inheritance, all with LOW/NORMAL 1,25-D -> CALCITRIOL HELPS (§20; HHRH in §21 is the HIGH-1,25-D odd one out):

• XLH (PHEX) = XL FGF23 excess -> phosphate loss.
• ADHR (cleavage-resistant FGF23) = AD FGF23 excess -> phosphate loss.
• ARHR (DMP1 / ENPP1) = AR FGF23 excess -> phosphate loss.
• HHRH (SLC34A3 / NaPi-IIc) = NaPi-IIc phosphate loss -> HIGH active vit D + hypercalciuria (the FGF23-INDEPENDENT one -> phosphate ONLY, never calcitriol).

CaSR set-point — loss and gain are MIRROR IMAGES (§23):

• FHH (CASR het LoF) = CaSR activity LOSS -> hypocalciuric hypercalcemia.
• NSHPT (CASR biallelic LoF) = SEVERE CaSR activity loss -> hypocalciuric hypercalcemia (newborn emergency).
• ADH (CASR GoF) = CaSR activity GAIN -> hypercalciuric hypocalcemia (the exact mirror of FHH).

Idiopathic infantile hypercalcemia — both end in HIGH active vit D (§21):

• IIH type 1 (CYP24A1) = active-vit-D CATABOLISM dysfunction -> hypercalciuric hypercalcemia.
• IIH type 2 (SLC34A1 / NaPi-IIa) = NaPi-IIa phosphate loss -> hypercalciuric hypercalcemia.

LOOK-ALIKES that bite: IIH (two I's, CALCIUM) is NOT IHH (hypogonadotropic hypogonadism / Indian Hedgehog); ADHR (PHOSPHATE, FGF23) is NOT ADH (CALCIUM, CaSR); HHRH = HYPER-calciuria, FHH = HYPO-calciuric.

CALCIOPENIC vs FGF23-MEDIATED HYPOPHOSPHATEMIC vs HEREDITARY HYPOPHOSPHATEMIC RICKETS WITH HYPERCALCIURIA (HHRH) — THE UNIFIED LAB COMPARISON:

THE MECHANISTIC LOGIC:

CALCIOPENIC RICKETS starts with INSUFFICIENT CALCIUM reaching the mineralization front — either substrate (dietary Ca, vitamin D substrate, 1α-hydroxylation) or end-organ response (VDR). The parathyroids sense the ionized Ca drop and crank up PTH. Secondary hyperparathyroidism then does TWO THINGS that shape the labs: it MOBILIZES BONE Ca (so serum Ca often looks low-normal rather than overtly low — a frequent trap) and DUMPS PHOSPHATE in the urine via NaPi-IIa downregulation. So phosphate is low here too, but it's a DOWNSTREAM CONSEQUENCE of PTH, not the primary lesion. ALP is highest of any group because the mineralization defect is compounded by PTH-driven bone turnover.

The 1,25D SPLIT inside the calciopenic group is the discriminator:

• LOW 1,25D in 1α-hydroxylase deficiency (VDDR1A / CYP27B1) or substrate deficiency (nutritional VDD).
• PARADOXICALLY ELEVATED 1,25D in VDDR2A/B because the receptor (or post-receptor signaling) can't shut off the upstream drive.

FGF23-MEDIATED HYPOPHOSPHATEMIC RICKETS is MECHANISTICALLY INVERTED. FGF23 (or a phenocopy of FGF23 excess via PHEX / DMP1 / ENPP1 loss) does TWO THINGS at the proximal tubule:

• Internalizes NaPi-IIa/IIc -> RENAL PHOSPHATE WASTING.
• SUPPRESSES CYP27B1 while INDUCING CYP24A1 -> 1,25D fails to rise.

That second effect is the DIAGNOSTIC ANCHOR. Hypophosphatemia SHOULD drive 1,25D up; here it doesn't. So you get low P with inappropriately normal or low 1,25D, normal Ca, and consequently NORMAL PTH (no secondary HPT because Ca homeostasis is intact). Intestinal Ca absorption isn't impaired, so urine Ca is normal or low.

HHRH isolates the RENAL-PHOSPHATE-WASTING mechanism WITHOUT FGF23. SLC34A3 LoF knocks out NaPi-IIc directly. Because 1α-hydroxylase isn't suppressed, low P does what it's SUPPOSED to do — DRIVE 1,25D UP. Elevated 1,25D then drives INTESTINAL Ca HYPERABSORPTION -> HYPERCALCIURIA, suppressed PTH, and NEPHROCALCINOSIS RISK. This is the ONE HYPOPHOSPHATEMIC RICKETS YOU DO NOT WANT TO TREAT WITH CALCITRIOL.

QUICK BEDSIDE HEURISTIC:

• HIGH PTH + LOW Ca + LOW P -> CALCIOPENIC. Split on 25(OH)D and 1,25D.
• NORMAL PTH + NORMAL Ca + LOW P -> RENAL PHOSPHATE WASTING. Split on FGF23 and urine Ca: * HIGH FGF23 -> XLH / ADHR / ARHR / TIO. * LOW FGF23 + HYPERCALCIURIA -> HHRH.

NEONATAL HYPOCALCEMIA splits by TIMING. EARLY (<72 h) = the post-natal PTH surge lags: prematurity, infant of a diabetic mother, perinatal asphyxia / stress, and the one to CATCH — MATERNAL HYPERPARATHYROIDISM / hypercalcemia suppressing the fetal parathyroids. LATE (>72 h to weeks) = high phosphate load (cow's-milk / unmodified formula), HYPOMAGNESEMIA, maternal vitamin-D deficiency, or congenital hypoparathyroidism / 22q11 (§23). Early is usually transient (support with calcium +/- calcitriol); persistent or late hypocalcemia earns the PTH / magnesium / vitamin-D / 22q11 workup — and ALWAYS check and replace MAGNESIUM, since hypomagnesemia both blocks PTH release and causes PTH resistance.

XLH (PHEX) is the COMMONEST inherited rickets / hereditary hypophosphatemia. The whole hypophosphatemic-rickets family maps onto one axis — FGF23 raised (most of them), or the renal phosphate transporter lost outright (the FGF23-independent escape, §21):

X-linked hypophosphatemia (XLH) = PHEX LoF, X-linked dominant. PHEX is an osteocyte/osteoblast endopeptidase; loss raises intact FGF23 indirectly, plus there are direct skeletal effects via ASARM peptide. Phenotype: bowing once weight-bearing begins, disproportionate short stature, dental abscesses, craniosynostosis, enthesopathy in adulthood, high ALP.

Old conventional therapy: oral phosphate plus calcitriol, multiple daily doses. Costs: GI intolerance, hypercalciuria, nephrocalcinosis, tertiary hyperparathyroidism, incomplete rickets healing.

Burosumab is anti-FGF23 monoclonal. Now first-line in pediatric XLH with active rickets per 2025 guidance. 0.8 mg/kg SC q2 weeks, titrated to fasting phosphate in the lower half of the age-specific reference range. DO NOT combine with phosphate plus calcitriol.

The FGF23-excess family (mechanism-grouped, autosomal recessive forms ARHR1-3 + the others):

Autosomal dominant hypophosphatemic rickets (ADHR) = FGF23 GoF (R176/R179 furin cleavage site). Iron deficiency upregulates FGF23 transcription, so always check iron. Treating iron often normalizes phosphate.

The IRON EXCEPTION worth knowing: iron DEFICIENCY raises FGF23 production but normally raises its CLEAVAGE in step, so INTACT FGF23 (and phosphate) hold steady. FERRIC CARBOXYMALTOSE (FCM) is the trap — it specifically BLOCKS FGF23 cleavage, so intact FGF23 climbs within ~24 h of the infusion -> renal phosphate wasting -> hypophosphatemic OSTEOMALACIA (far less with iron isomaltoside / other formulations; Wolf 2020 JAMA). An acquired, drug-induced phenocopy of ADHR — check phosphate in anyone getting repeated FCM, especially with bone pain.

ARHR1 = DMP1 LoF. DMP1 is an osteocyte protein that normally suppresses FGF23 transcription; loss derepresses FGF23 -> XLH-like phenotype. Treat like XLH; burosumab off-label.

ARHR2 = ENPP1 LoF. THIS ONE IS DIFFERENT. ENPP1 makes inorganic pyrophosphate (PPi), the master inhibitor of ectopic mineralization. Severe end: GACI (generalized arterial calcification of infancy) — lethal vascular and visceral calcification in the first 6 months from coronary and aortic mineralization. Survivors evolve into hypophosphatemic rickets later. Burosumab in ENPP1 is theoretically contraindicated — raising phosphate and 1,25-D in a PPi-deficient patient could accelerate the very calcification you need to prevent. Off-label use exists in case reports without uniform disaster, but stay outside it until INZ-701 (ENPP1-Fc fusion ERT) is available — that's the rational therapy.

ARHR3 = FAM20C LoF (RAINE SYNDROME). Originally described as a lethal generalized osteosclerotic dysplasia with characteristic facies (exophthalmos, midface hypoplasia, depressed nasal bridge, gum hyperplasia), but milder survivors are now recognized with adult-onset hypophosphatemic rickets and intracranial calcifications. FAM20C is a Golgi kinase that phosphorylates secreted proteins involved in mineralization, including FGF23, DMP1, MEPE, and osteopontin. Loss of FAM20C means FGF23 is not properly phosphorylated at the furin recognition site, so it accumulates as intact (uncleaved) and active. The paradox: broad mineralization failure produces OSTEOSCLEROSIS in some bones AND phosphate wasting from FGF23 excess simultaneously. The other FGF23 disorders give pure undermineralization; Raine is the one with sclerosis. Burosumab use is reported in case reports but carries the same mineralization-related concern as ENPP1.

TIO (tumor-induced osteomalacia): acquired FGF23 excess from a small phosphaturic mesenchymal tumor. Find it with 68Ga-DOTATATE PET-CT. Resection is curative.

McCune-Albright FD with hypophosphatemia: FGF23 from FD lesions.

The whole family signature: phosphate low, 1,25-D inappropriately low/normal (the FGF23 suppression), urine calcium normal (this is the discriminator from FGF23-INDEPENDENT wasters like HHRH where 1,25-D and urine calcium are both high).

THE LOW-FGF23 MIRROR — hyperphosphatemic familial tumoral calcinosis (HFTC), the opposite end of the same axis. Everything above is FGF23 EXCESS -> phosphate WASTING; HFTC is FGF23 DEFICIENCY or resistance -> phosphate RETENTION. Three genes, one result: GALNT3 (normally O-GLYCOSYLATES FGF23 to SHIELD it from cleavage — lose it and FGF23 is over-cleaved), FGF23 itself (LoF), or KLOTHO (KL, the coreceptor — FGF23 can no longer signal). Intact FGF23 collapses or goes unheard -> HIGH phosphate + HIGH/normal 1,25-D -> painful periarticular / ectopic CALCIFIED masses (and sometimes diaphysitis — the hyperostosis- hyperphosphatemia variant). So the FGF23 cleavage switch runs both ways: CAN'T cleave (ADHR, FCM, Raine) -> too much FGF23 -> wasting; OVER-cleave / can't make / can't hear it (GALNT3, FGF23, KL) -> too little -> retention (Jaschke 2021 Bone Res).

Hereditary hypophosphatemic rickets with hypercalciuria (HHRH) = SLC34A3 (NaPi-IIc) LoF. Phosphate is lost in urine but FGF23 is NOT driving this. So CYP27B1 is not suppressed. The proximal tubule senses low phosphate and upregulates 1-alpha-hydroxylase. 1,25-D RISES. Elevated 1,25-D drives intestinal calcium hyperabsorption, suppresses PTH, dumps calcium in the urine. Stones and nephrocalcinosis follow.

This is the COMPENSATION-OF-COMPENSATION pattern in action. The loudest biochemical signal (sky-high 1,25-D plus hypercalciuria) is the second compensation, not the primary lesion. The actual broken node is SLC34A3 (NaPi-IIc). Trace from the loudest deviation back to the broken node — the same logic applies to any endocrine cascade where the eye is drawn to the most dramatic value (think GRA aldosterone, or ACTH-driven DOC in 11-beta-OH CAH).

The XLH-vs-HHRH discriminator (high-yield): in both, P is low and TmP/GFR is low. But in X-linked hypophosphatemia (XLH) the 1,25-D is low/normal and urine calcium is normal. In HHRH 1,25-D is HIGH and urine calcium is HIGH. PTH is suppressed in HHRH, normal-to-high in XLH.

Treatment of HHRH: phosphate alone, multiple daily doses. Active vitamin D is AVOIDED as default (1,25-D is already high). The "never give vitamin D" line is too absolute — in severe refractory disease with persistent rickets or secondary hyperparathyroidism, low-dose calcitriol with strict urinary calcium monitoring may be necessary. Adjuncts: fluconazole (inhibits 1-alpha-OH) and GH for short stature.

Dent disease (CLCN5) is the differential — proximal tubulopathy with LMW proteinuria, hypercalciuria, nephrocalcinosis, progressive renal failure. Check urine beta-2-microglobulin or retinol-binding protein.

And idiopathic infantile hypercalcemia (IIH) is the FGF23-independent flip side of this story. Two genes do it, with one common thread: the routine vitamin D drops you give every infant are what unmasks the disease.

CYP24A1 LoF (AR) is the catabolic side. CYP24A1 normally 24-hydroxylates 1,25-D and 25-D to inactive forms. Lose the catabolism and active vitamin D piles up. Phenotype: hypercalcemia, hypercalciuria, suppressed PTH, nephrocalcinosis — often a healthy infant who DETERIORATES after the prophylactic 400-800 IU vitamin D supplement at 2-4 months. The diagnostic tell is a high 25-D : 24,25-D ratio. Stop the vitamin D, low-calcium diet, hydrate, sometimes fluconazole (off-label, inhibits 1-alpha-OH).

SLC34A1 LoF (AR) is the TRANSPORT side — IIH type 2 (the partner of IIH type 1 = the CYP24A1 catabolism defect just above; both end at high 1,25-D, by different routes). The trap: it is a DIFFERENT gene from HHRH — HHRH is SLC34A3 (NaPi-IIc), this is SLC34A1 (NaPi-IIa). Both are sodium-phosphate cotransporters whose LoF wastes phosphate -> hypophosphatemia -> 1-alpha-hydroxylase revs up -> 1,25-D RISES. They DIVERGE in which phenotype wins: SLC34A3 (HHRH) presents as RICKETS, while SLC34A1 presents as the HYPERCALCEMIA picture — the high 1,25-D drives gut calcium hyperabsorption -> hypercalcemia + hypercalciuria + nephrocalcinosis, PTH suppressed (Schlingmann, JASN 2015). ("Calcium phenotype" = exactly that: hypercalcemia + hypercalciuria + nephrocalcinosis, the same endpoint as the CYP24A1 / IIH type 1 cousin.)

Williams syndrome (the 7q11.23 microdeletion — see §23) overlaps with this picture biochemically in infancy. Don't routinely supplement vitamin D in a Williams baby.

NUTRITIONAL VITAMIN D DEFICIENCY RICKETS unfolds in THREE BIOCHEMICAL PHASES. Knowing the phase tells you whether the labs look "normal-Ca" or "low-Ca + low-P" — and patients can present in any phase, so the labs by themselves don't rule out rickets. The phases are a moving picture, not a single snapshot:

THE LOGIC OF THE PHASES (and how each one presents):

• PHASE I: 25(OH)D depot is depleted. 1,25D production drops, intestinal Ca absorption falls, serum Ca starts to dip, and the parathyroids begin to rise but haven't fully compensated yet. Often ASYMPTOMATIC, but the early-presentation pattern is hypocalcemic seizures or tetany in infants and adolescents; bone changes are minimal.
• PHASE II: secondary HPT kicks in. PTH mobilizes bone Ca (so serum Ca normalizes) AND dumps phosphate in the urine (so serum P falls). This is the LONGEST and most commonly-presenting phase, and the bone paradoxically gets WORSE — PTH-driven turnover unmasked on already-undermineralized bone gives the CLASSIC RICKETS PICTURE on X-ray (frayed metaphyses, cupping, rachitic rosary, bowing). Trap: serum Ca looks "normal," so rickets is easily MISSED here.
• PHASE III: bone reserves exhausted, PTH can no longer maintain serum Ca, so both Ca and P are now low — compensation has FAILED. Severe disease: hypocalcemic seizures, tetany, severe rachitic deformity, cardiomyopathy in infants; ALP is highest.

This phasing is one reason why "the patient with rickets has normal calcium" trips clinicians — they're looking at Phase II. Always combine Ca with P + PTH + ALP + 25(OH)D for the phase to come into focus. Treatment in all phases is REPLETION: cholecalciferol 2000-6000 IU/day for 6-12 weeks (or single-dose stoss 100,000-600,000 IU per local protocol) plus oral calcium 30-75 mg/kg/day until labs and radiographs normalize. Calcitriol is unnecessary in nutritional VDD — once 25-D repletes, the patient makes their own 1,25-D.

The vitamin D-dependent rickets (VDDR) family is the GENETIC mimics that look like nutritional VDD but DON'T respond to plain D repletion. Each one breaks — or, for VDDR3, over-runs — a single step of the activation -> action -> catabolism path:

Reverse-lookup from the labs: LOW 25-D = VDDR1B or VDDR3; normal/high 25-D with LOW 1,25-D = VDDR1A; normal 25-D with VERY HIGH 1,25-D (± alopecia) = VDDR2A or VDDR2B.

CaSR is a class C GPCR. Loss-of-function raises the set-point at which PTH is suppressed.

• Heterozygous LoF = FHH (familial hypocalciuric hypercalcemia) — the COMMONEST inherited hypercalcemia. Mild hypercalcemia, normal-to-mildly-elevated PTH, LOW urine calcium (Ca/Cr ratio <0.01). The discriminator from primary hyperparathyroidism. Do not parathyroidectomize.
• GNA11 LoF = FHH2. AP2S1 LoF = FHH3 (more symptomatic).
• Homozygous LoF = NSHPT (neonatal severe hyperparathyroidism). Critical infant, urgent total parathyroidectomy.

Gain-of-function lowers the set-point.

• Autosomal dominant hypocalcemia type 1 (ADH1) = activating CaSR — the set-point is reset DOWN, so PTH is switched off at a low calcium. Lab signature: Ca LOW, P HIGH, PTH inappropriately normal-to-low, 1,25-D LOW / inappropriately low-normal (no PTH drive to renal 1-alpha-hydroxylase), 25-D NORMAL (nutritional, not PTH/CaSR-controlled), urine Ca HIGH; +/- hypomagnesemia (the activated CaSR also wastes Mg in the thick ascending limb). The renaming "hypercalciuric hypocalcemia type 1" captures the trap: treating with calcitriol pushes already-elevated urine calcium higher -> nephrocalcinosis. Aim for the LOW end of normal calcium, not eucalcemia. Calcilytics (encaleret) are in trials.
• ADH2 = activating GNA11.

• DiGeorge (22q11.2 del, TBX1) = conotruncal cardiac, cleft palate, T-cell ID, characteristic face. Hypopara variable. Get 22q11.2 FISH on any hypocalcemic neonate with a murmur or cleft.
• Sanjad-Sakati (TBCE, AR) = HRD = Hypoparathyroidism + Retardation + Dysmorphism. IUGR, microcephaly, deep-set eyes, beaked nose, micrognathia, severe intellectual disability. Middle Eastern.
• Kenny-Caffey type 1 (TBCE, AR) = allelic to Sanjad-Sakati, so it SHARES the intellectual disability; medullary stenosis of long bones is the radiograph.
• Kenny-Caffey type 2 (FAM111A, AD) = same skeletal phenotype (medullary stenosis) but NORMAL cognition — that is the type-1-vs-type-2 split.
• Barakat (GATA3, AD) = HDR = Hypoparathyroidism + Deafness + Renal dysplasia.
• autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED) / APS-1 (AIRE LoF, AR) = triad of mucocutaneous candidiasis (by age 5) + hypoparathyroidism (age 10) + Addison's (age 15). Plus ectodermal features. AIRE drives "promiscuous" expression of self-antigens in medullary thymic epithelial cells for central tolerance. Loss = collapse of thymic negative selection. Ruxolitinib (JAK inhibitor) emerging for IFN-driven autoimmunity.
• THE REVERSIBLE one you must not miss: severe HYPOMAGNESEMIA causes FUNCTIONAL hypoparathyroidism — low Mg blocks BOTH PTH secretion AND PTH action at the receptor, so you get hypocalcemia with LOW / inappropriately-normal PTH that will NOT correct until Mg is replaced. The genetic form is TRPM6 LoF = "hypomagnesemia with secondary hypocalcemia" (HSH): an intestinal/renal Mg-transport defect presenting in infancy with hypocalcemic + hypomagnesemic SEIZURES — treat with magnesium, not calcium alone.

So the mnemonic that finally sticks: Sanjad-Sakati = R (retardation), Barakat = D (deafness + renal dysplasia).

DiGeorge / 22q11.2 is more than the hypopara line in the tree above. TBX1 haploinsufficiency wrecks every pharyngeal arch derivative: conotruncal cardiac (tetralogy, truncus, IAA-B, VSD), cleft palate or velopharyngeal insufficiency, T-cell immunodeficiency from thymic hypoplasia, characteristic facies (low-set ears, short philtrum, downturned mouth), 30% renal anomalies, short stature in 20-30%.

The hypopara is the trap — it's LATE-ONSET and INTERMITTENT in many patients. A normal neonatal calcium does NOT rule it out. The school- age kid with new hypocalcemic seizures and a known 22q11 microdeletion is the classic story. Lifelong calcium + PTH screening. Add a 25-fold lifetime schizophrenia risk and high rates of ADHD and autism-spectrum to the surveillance load.

WILLIAMS SYNDROME (7q11.23 microdeletion) shows up here because of the hypercalcemia, but the syndrome itself is multisystem and you'll meet it in clinic for other reasons. The face is "elfin" (broad forehead, periorbital fullness, stellate iris, full lips). The personality is overfriendly and verbally fluent against weaker visuospatial skills with mild-moderate ID. Cardiac: supravalvular aortic stenosis from ELN haploinsufficiency, peripheral pulmonary stenosis, lifelong HTN risk.

The Williams endocrine pack:

• Infantile hypercalcemia in ~15%, vitamin-D-triggered (same picture as IIH from §21). Avoid routine D drops in a Williams baby.
• Hypothyroidism in ~30%, often subclinical, often hemiagenesis on US.
• Glucose intolerance in ~75% of teens and adults.
• Short stature in ~70%.
• Early / advanced puberty is common (menarche ~1-1.5 yr early); true CENTRAL precocious puberty in a minority (~15-20% of girls), not half.

Surveillance: annual BP plus renal artery screen, TSH yearly, OGTT from adolescence, lipid panel, ophtho (stellate iris is benign but the ocular workup catches other findings).

22q11.2 FISH or microarray on any hypocalcemic neonate with a heart murmur or cleft palate. Williams FISH on any infant with hypercalcemia + supravalvular AS + elfin face.

GNAS encodes Gsα, the transducer that couples the PTH/PTHrP receptor (and every other Gs-coupled receptor) to cAMP. Break Gsα signalling and you get pseudohypoparathyroidism (PHP). What makes the family confusing is that GNAS is an IMPRINTED COMPLEX LOCUS — so WHICH parent's allele is hit, and HOW, sets the phenotype.

THE IMPRINTING IN ONE IDEA: Gsα has no DMR at its own promoter, so it runs BIALLELICALLY in most tissues — but in a few (proximal renal tubule, thyroid, gonad, pituitary somatotrope) the PATERNAL Gsα is silenced, leaving those tissues on the MATERNAL copy alone. That paternal silencing is set in cis by the locus's other transcripts, each off its own differentially methylated promoter (DMR). Along the locus, 5'->3': NESP55, GNAS-AS1, XLαs, A/B, then the Gsα exons.

THE CONTROL KNOB is the A/B DMR: normally the MATERNAL A/B is methylated (silent) and the PATERNAL A/B is unmethylated (active, suppressing paternal Gsα where it counts). STX16 is the upstream cis-element that ESTABLISHES the maternal A/B imprint during female gametogenesis — delete it and that imprint fails (the pure-imprinting form of PHP, below).

THE WHOLE FAMILY IS THREE WAYS TO BREAK ONE AXIS, by parent of origin:

• MATERNAL Gsα CODING LoF -> imprinted tissues lose their only working copy -> MULTIHORMONE RESISTANCE + AHO (= PHP1A; = PHP1C when in vitro Gsα activity looks normal but coupling is broken).
• PATERNAL Gsα CODING LoF -> imprinted tissues already ran on maternal Gsα, so NO hormone resistance — but biallelic tissues (bone / cartilage, fat) still lose a copy -> AHO ALONE (= PPHP), or severe deep heterotopic ossification (= POH).
• MATERNAL A/B LOSS OF METHYLATION -> the maternal allele now behaves paternal and suppresses Gsα in the imprinted tissues -> PTH (± mild TSH) resistance WITHOUT AHO (= PHP1B). No coding mutation at all.

PHP2 sits OUTSIDE GNAS: the block is DOWNSTREAM of cAMP, and the commonest cause is plain vitamin-D deficiency phenocopying PHP. Treatment of the PTH-resistant forms is active vitamin D + calcium, WITHOUT the hypercalciuria worry of true hypoparathyroidism — PTH is high, so the kidney is still reabsorbing calcium.

The EuroPHP network (Thiele 2016, endorsed by the Mantovani 2018 consensus) folded all of this into the iPPSD umbrella (numbering below); the classical PHP names still rule clinically.

THE TYPES:

LAB SIGNATURE across types with PTH resistance (1A, 1B, 1C, 2):

• Ca LOW, P HIGH, PTH HIGH — the hypoparathyroid biochemical pattern with ELEVATED rather than SUPPRESSED PTH.
• 1,25(OH)2D LOW or inappropriately normal — proximal tubule Gsα failure blocks PTH-driven CYP27B1 induction.
• Magnesium NORMAL (distinguishes from hypomagnesemic functional hypoparathyroidism).
• ELLSWORTH-HOWARD TEST (mostly historical): infused PTH FAILS to raise urinary cAMP and phosphate in PHP1; raises cAMP but NOT phosphate in PHP2. Useful conceptually even if rarely performed.

HOW TO DISCRIMINATE AT THE BENCH:

• AHO present + multihormone resistance + LOW erythrocyte Gsα -> PHP1A. Sequence GNAS coding exons.
• AHO present + multihormone resistance + NORMAL Gsα activity -> PHP1C. Same GNAS sequencing; mutation will be in the receptor- interaction domain (often exon 13). Many centers collapse 1C into 1A given identical management.
• NO AHO + ISOLATED PTH resistance (± mild TSH elevation) + normal Gsα -> PHP1B. Go to methylation analysis at GNAS DMRs: loss of methylation at GNAS A/B:TSS-DMR is the hallmark; a broader multi-DMR defect occurs in BOTH sporadic PHP1B and familial NESP55/AS deletions.
• AHO WITHOUT hormone resistance + family history with maternal- transmission PHP1A -> PPHP. Same GNAS mutation, paternal allele.
• Severe progressive heterotopic ossification beyond skin / SC tissue, often dermatomal -> progressive osseous heteroplasia (POH). Paternal GNAS.
• Isolated PTH resistance with normal cAMP response but blunted phosphaturia -> PHP2. ALWAYS RULE OUT VITAMIN D DEFICIENCY FIRST — it's the commonest phenocopy and reverses with repletion.

STX16 AND AD-PHP1B — THE IMPRINT-ESTABLISHMENT LESION:

The classic finding is a ~3 kb microdeletion of STX16 exons 4-6, located ~220 kb centromeric to GNAS on 20q13. STX16 itself is NOT imprinted and isn't part of GNAS — but the deleted region contains a CIS-ACTING ELEMENT REQUIRED to establish or maintain methylation at the GNAS A/B:TSS-DMR DURING FEMALE GAMETOGENESIS.

WHY "AD" BUT MATERNAL-ONLY EXPRESSION: The deletion segregates as autosomal dominant — transmitted to 50% of offspring of either sex carrier. But the PHP1B phenotype only appears when the deletion is INHERITED FROM THE MOTHER:

• MATERNAL transmission -> maternal GNAS A/B-DMR in the offspring FAILS to acquire its normal methylation imprint -> loss of A/B silencing -> BIALLELIC A/B transcription suppresses Gsα expression in imprinted tissues (proximal tubule, thyroid) -> PTH ± mild TSH resistance.
• PATERNAL transmission -> SILENT CARRIER. Paternal A/B is normally unmethylated; paternal Gsα is normally silenced in imprinted tissues. The deletion has no phenotypic consequence; carriers recognized only by pedigree analysis.

So pedigrees look like CLASSICAL IMPRINTED-GENE DISORDERS: the mutation tracks autosomal dominantly through both sexes, but affected individuals ONLY appear in offspring of FEMALE carriers. Mirrors the PHP1A / PPHP parent-of-origin logic, just ONE REGULATORY LAYER UP — at imprint establishment rather than at the coding sequence.

METHYLATION SIGNATURE — discriminates familial AD-PHP1B from sporadic:

If a methylation panel shows isolated A/B loss of methylation, screen STX16 for the deletion and offer cascade testing. A BROAD multi-DMR defect is usually sporadic (low recurrence) — BUT exclude a familial NESP55 / AS-region deletion first, because those are AD with high recurrence. AD inheritance means siblings, daughters of female carriers, and grandchildren via the female line are all at risk; paternal-line carriers are silent but can transmit.

THE iPPSD CLASSIFICATION (EuroPHP network, Thiele 2016; endorsed by the Mantovani 2018 consensus) reframes the family by the COMMON LESION — inactivating PTH/PTHrP cAMP signaling — and NUMBERS BY THE BROKEN NODE on the receptor -> Gsα -> cAMP -> PKA -> PDE cascade, not by eponym:

• iPPSD1 = PTH1R (the receptor itself) — Blomstrand lethal + Eiken chondrodysplasia (§25).
• iPPSD2 = GNAS Gsα CODING loss — PHP1A, PHP1C, PPHP, AND POH (one Gsα defect, split by parent-of-origin).
• iPPSD3 = GNAS METHYLATION / imprinting defect — PHP1B.
• iPPSD4 = PRKAR1A — acrodysostosis type 1 (WITH hormone resistance).
• iPPSD5 = PDE4D — acrodysostosis type 2 (usually NO hormone resistance).
• iPPSD6 = PDE3A — autosomal-dominant HYPERTENSION with brachydactyly (HTNB / Bilginturan).
• iPPSDx = the full clinical picture with no molecular defect found yet.

The common slips this corrects: PRKAR1A is iPPSD4 (not 1) and PDE4D is iPPSD5 (not 4); and POH is just iPPSD2 (a Gsα disorder), NOT its own number.

The convergence is the point: four eponymic labels (PHP1A, PHP1C, PPHP, POH) collapse into one molecular node, iPPSD2 (Gsα coding), and PHP2 falls OUT of the scheme entirely — its defect is downstream of Gsα.

iPPSD2 puts PHP1A and PPHP in ONE molecular bucket (same GNAS coding lesion) but MANAGEMENT DIVERGES by parent of origin — maternal transmission + multihormone resistance (PHP1A) vs paternal transmission without resistance (PPHP / POH). Recurrence is 50% per pregnancy; which phenotype the child expresses depends on which parent carries. Grouping them keeps genetic counseling and at-risk-relative screening coherent, and flags PPHP cases that may progress to POH:

TWO CLINICAL TRAPS: 1. TSH elevation can precede PTH resistance by years in PHP1A — congenital/neonatal SUBCLINICAL HYPOTHYROIDISM on newborn screening with later-onset hypocalcemia is a classic trajectory. Brachydactyly often isn't apparent until ~age 4. 2. GH DEFICIENCY IS UNDERRECOGNIZED in PHP1A (GHRH resistance). Worth screening if growth velocity is poor; replacement is effective.

ARCHIBALD SIGN — the classic AHO bedside finding, the quick test for the brachydactyly type E hand (§43). Have the patient make a fist. Normal: knuckles 2-3-4-5 form a smooth ascending curve. In AHO (PHP1A or PPHP): the 4TH METACARPAL is SHORT, so when the fist is made, the 4th knuckle DIMPLES IN (or is replaced by a dimple) instead of bulging out. The 5th metacarpal is often also short. Confirmed on hand X-ray (metacarpal sign — line drawn through heads of 4th and 5th metacarpals crosses through or distal to head of 3rd; in normal hands the line passes distal to all three).

WHEN THE AHO HAND IS THERE BUT GNAS IS NEGATIVE — run the differential off the hand. Brachydactyly type E is the most objective AHO feature; the ONE feature fairly specific to GNAS is ectopic / subcutaneous ossification — short stature, obesity, round face, and ID are all shared with the mimics. So with GNAS clean, let the BDE pattern + the extra features pick the gene — this differential lands a molecular diagnosis in about HALF of GNAS-negative AHO, TRPS1 the single biggest yield:

• TRPS1 (tricho-rhino-phalangeal syndrome) — the classic mimic: sparse, slow-growing scalp hair + thin lateral brows (tricho-), bulbous / pear-shaped nose + long philtrum + thin upper lip (rhino-), and a brachydactyly that involves the PHALANGES with CONE-SHAPED epiphyses ("outcarving") (-phalangeal). TRPS2 = Langer-Giedion = a contiguous 8q24 deletion of TRPS1 + EXT1, adding multiple exostoses + ID.
• ACRODYSOSTOSIS (PRKAR1A = iPPSD4, PDE4D = iPPSD5): the brachydactyly + cone-shaped epiphyses appear EARLY in childhood, unlike the slowly-emerging brachydactyly of PHP. Hormone resistance is a WEAK discriminator — present in ~76% of iPPSD4 but also ~27% of iPPSD5 — so don't lean on it.
• PTHLH (the PTHrP ligand): BDE with ADVANCED bone age -> early epiphyseal closure, so the short stature may only declare in late childhood (final height ends up below target).

The full Bell-type landscape (incl. 2q37 / HDAC4 and Turner) is in §43.

Mirror mutations on the PTH/PTHrP receptor.

JANSEN metaphyseal chondrodysplasia = PTH1R activating mutation (H223R most common, also T410P, I458R). Autosomal dominant, often de novo. The receptor signals constitutively without PTH or PTHrP ligand. Biochemistry looks like primary hyperparathyroidism with one critical exception: PTH and PTHrP are both NORMAL or SUPPRESSED — the receptor is doing the work alone. Chronic hypercalcemia, hypophosphatemia, hyperphosphaturia. The skeletal phenotype is the disease: PTHrP via PTH1R normally delays hypertrophic differentiation of growth-plate chondrocytes. Constitutive activation arrests them in the proliferative zone — severe metaphyseal cupping and fraying, "kissing" femoral heads, severe short-limbed dwarfism, characteristic frontal bossing + hypertelorism + micrognathia.

BLOMSTRAND lethal chondrodysplasia = PTH1R loss-of-function biallelic. Autosomal recessive. Lethal in utero or shortly after birth. Mirror of Jansen: without the PTHrP signal at the growth plate, chondrocytes mature TOO FAST. Accelerated bone age in utero, prematurely fused growth plates by 20 weeks gestation, profoundly increased skeletal density on prenatal imaging. Plus breast aplasia and tooth aplasia (PTHrP also drives mammary and dental bud development). And because PTH cannot reach its receptor either, Blomstrand is simultaneously a state of complete PTH RESISTANCE — functional hypoparathyroidism with HYPOCALCEMIA, HYPERPHOSPHATEMIA, and inappropriately HIGH PTH, plus reactive PARATHYROID HYPERPLASIA (the glands ramping uselessly against a dead receptor). It is NOT "just bone": the lethal hyperdense skeleton is the PTHrP half, the mineral derangement is the PTH half of losing one shared receptor. (The maternal-fetal calcium handoff also fails — fetal PTH cannot rescue across the placenta.)

The teaching pair: PTHrP via PTH1R is the brake on chondrocyte maturation. Lose the brake (Blomstrand) -> premature maturation, hyperdense fetal skeleton, lethal. Lock the brake on (Jansen) -> arrested chondrocytes, soft metaphyseal dysplasia, hypercalcemia.

EIKEN syndrome rounds out the spectrum — a rare NON-lethal homozygous PTH1R mutation that, crucially, does NOT sit cleanly on the Jansen-to- Blomstrand axis: it is PLEIOTROPIC. In the growth-plate chondrocytes it behaves GAIN-of-function-LIKE — retarded, DELAYED ossification with growth plates that stay open long past closure and severely hypomineralized epiphyses (the same DIRECTION as Jansen, the OPPOSITE of Blomstrand's premature ossification — which is exactly why a simple "partial LoF" label is wrong). But for PTH-driven mineral handling the same mutation reads LOSS-of-function: characterized cases show overt PTH RESISTANCE — HYPOCALCEMIA, HYPERPHOSPHATEMIA, and inappropriately HIGH PTH, a pseudohypoparathyroidism-like biochemistry (Eiken JBMR 2024). So "calcium and PTH normal" is a trap — the calcium axis is the loss-of-function half. Clinically: survival to adulthood, short stature, abnormal hand/foot shape, delayed dentition, ON TOP of the hypocalcemic PTH-resistance labs.

So read PTH1R on TWO axes, not one dose ladder. GROWTH-PLATE brake (PTHrP): brake locked ON (Jansen GoF — and Eiken's GoF-LIKE chondrocyte effect) -> delayed maturation, open plates; brake OFF (Blomstrand LoF) -> premature ossification, hyperdense lethal skeleton. CALCIUM (PTH): GoF (Jansen) -> HYPERcalcemia; LoF (Blomstrand AND Eiken) -> PTH resistance with HYPOcalcemia + hyperphosphatemia + high PTH. Jansen and Blomstrand are clean mirrors on BOTH axes; Eiken is the pleiotropic outlier — Jansen-like growth plate, Blomstrand-like calcium. One receptor, two jobs (the PTHrP brake + PTH calcium-handling), and only Eiken splits them.

OI is classically COL1A1/COL1A2 dominant. Sillence types I-IV was the original (1979) framework; type V (IFITM5) was added later — the recurrent c.-14C>T variant in the IFITM5 5' UTR (creates a novel start codon) giving hyperplastic callus after fracture, radiodense metaphyseal bands, and interosseous-membrane calcification of the forearm. Distinct from type IV.

2024 dyadic nosology replaces the endless type-numbering with phenotype-plus-gene pairing. Non-COL1 genes you should know: SERPINF1 (old type VI), CRTAP / P3H1 / PPIB (prolyl 3-hydroxylation complex, severe AR), WNT1, SP7, BMP1, PLOD2/FKBP10 (Bruck syndrome).

IV bisphosphonates + rodding + setrusumab (anti-sclerostin, trials) are the management (bisphosphonate mechanism + the full anti-resorptive ladder are in §27).

Hypophosphatasia (ALPL LoF) is the disorder you must not give bisphosphonates to. TNSALP normally hydrolyzes inorganic pyrophosphate (PPi) at the mineralizing front. PPi accumulates, PLP and PEA accumulate, mineralization fails. Phenotypic spectrum:

• Perinatal lethal.
• Infantile: craniosynostosis, hypercalcemia (paradoxical from PPi releasing skeletal calcium), nephrocalcinosis, B6-responsive seizures (PLP cannot cross BBB without dephosphorylation).
• Childhood: premature loss of deciduous teeth with intact roots is the pathognomonic finding. Rickets-like radiographs with normal Ca/P, low ALP.
• Adult: stress fractures, dental disease, CPPD pseudogout.

THE COMMONEST MISS: "low ALP" only reads as low against an AGE- AND SEX-SPECIFIC range — infants and children normally run ALP 2-3x the adult level (growing bone), so a value that looks "normal for an adult" can be pathologically LOW for a child, and HPP gets missed because the lab flags only the adult range. Persistently low ALP + the picture -> measure the substrates that back up behind TNSALP: PLP (pyridoxal-5'-phosphate, the best test), PEA, and urinary PPi.

Bisphosphonates are stable PPi analogs. Giving them to an HPP patient is adding poison to a poison-overload state. Atypical femoral fractures reported in HPP misdiagnosed as osteoporosis. Asfotase alfa (recombinant TNSALP with deca-aspartate bone-targeting tail) is the ERT.

The two PRIMARY skeletal-fragility disorders — OI and the bisphosphonate-trap mimic hypophosphatasia — live in §26; rickets / osteomalacia (a MINERALIZATION defect, not fragility per se) is the §22 story. This section is the OTHER bucket plus the shared toolkit: a structurally normal skeleton made fragile by a SYSTEMIC disease or its treatment — SECONDARY osteoporosis — the 2019 ISCD way to diagnose it, how to read the DXA you order to chase it, and the anti-resorptive drug ladder (which serves primary OI too).

DIAGNOSIS is clinical + radiographic, NEVER DXA alone — this is the commonest error, and the 2019 ISCD reframe exists precisely because of it: a non-traumatic VERTEBRAL compression fracture is diagnostic BY ITSELF, at ANY Z-score (you can crush a vertebra with a normal DXA, and requiring both would miss it). So diagnose pediatric osteoporosis on EITHER:

• one or more non-traumatic vertebral compression fractures (any Z-score), OR
• a clinically significant fracture history (≥2 long-bone fractures by age 10, OR ≥3 by 19) AND BMD Z-score ≤ -2.0.

"Low BMD for age" (Z ≤ -2.0) on its own is a risk FLAG, not the disease. Grade the vertebral fracture by Genant height loss — grade 1 (20-25%), grade 2 (25-40%), grade 3 (>40%); anything grade 1 or worse counts.

The whole thing turns on the ENERGY of the injury — define it before you count any fracture:

• LOW-ENERGY (fragility) fracture = a break from force that would NOT snap a healthy bone: a fall from STANDING HEIGHT or less, a minor stumble or twist, or no identifiable trauma at all. THESE are the ones that count toward the diagnosis.
• HIGH-ENERGY fracture = a motor-vehicle crash, a fall from a HEIGHT (off a wall, a bike at speed, a trampoline), a hard sports collision — a force that would break a normal bone too. These do NOT count; they say nothing about bone quality.
• And not every site counts even at low energy: VERTEBRAL and LONG-BONE fractures are the currency. Isolated skull, facial, finger and toe fractures are excluded.

So the question at the bedside is never "did it fracture" but "did it fracture from something that SHOULDN'T have broken bone" — a low-energy long-bone or any non-traumatic vertebral fracture is the signal.

THE MECHANISTIC DIFFERENTIAL — secondary fragility comes from three failures, and most chronic-disease kids stack more than one:

SILENT vertebral fractures are the most under-diagnosed pediatric finding — missed without dedicated imaging. Get a baseline LATERAL-SPINE film or DXA vertebral-fracture assessment (DXA-VFA) in every high-risk child: glucocorticoid-treated, leukemia on the path to transplant, OI, severe CF, JIA on chronic steroids. The single highest-yield trigger is GLUCOCORTICOID exposure — screen anyone on ≥0.16 mg/kg/day prednisolone-equivalent for ≥3 months — because the DXA Z-score can read normal while the spine is already crushed.

READING THE DXA — the pediatric pitfalls that make adult habits wrong:

• Z-SCORE, never T-score — the T-score compares to peak ADULT mass and is meaningless in a growing skeleton. Report against age / sex-matched Z.
• SIZE ARTIFACT — DXA reports AREAL density (g/cm², a 2-D projection), so it UNDER-reads a small / short child's bone and OVER-reads a tall child's. The short kid (GHD, Turner, chronic disease) gets a falsely-low Z that is partly just small bones. The 2019 ISCD fix: ADJUST — height-for-age Z-score, or bone mineral apparent density (BMAD) for the spine, and adjust for BONE AGE if puberty is delayed but height is preserved. An unadjusted Z in a short child overstates the deficit.
• WHICH SITES — lumbar spine + total-body-less-head (TBLH) from ~age 3-5 to 15; add hip only from ~16, once the growing proximal femur is reproducible (hip is unreliable in growing bone). For the NON-AMBULANT / contractured child (CP, DMD) use the LATERAL DISTAL FEMUR — positioning is feasible and it samples the bone that actually fractures; forearm is the fallback if the child can't hold still.
• SERIAL SCANS — same scanner, and judge change against the least significant change (LSC), not any drift. The child's bone is also GROWING between scans, so a rising areal BMD can be SIZE rather than real density gain — re-adjust before calling a treatment response.

THE PEDIATRIC ADVANTAGE — vertebral RESHAPING. Unlike an adult, a growing child can spontaneously REMODEL a crushed vertebra back toward normal height — a real "window of opportunity" present in the PRE / EARLY-pubertal years and largely gone by mid-to- late puberty (in childhood-leukemia cohorts ~75% reshape, but ~25%, mostly the older kids, keep a permanent deformity). This flips the treat-vs-observe call:

• TRANSIENT, treatable driver (inflammatory disease into remission, steroids stopped, nutrition restored) + meaningful residual growth -> the skeleton can repair ITSELF; OBSERVE, optimize calcium / vitamin D / weight-bearing, re-image.
• PERMANENT or relentless driver (ongoing high-dose steroids, OI, progressive neuromuscular disease) OR mid-to-late puberty (window closing) -> TREAT; and in a few very-high-risk groups (SMA, severe CP, where a single fracture can cost ambulation) treat PRE-EMPTIVELY, before the formal criteria are met.

TREATMENT — the drug ladder (it treats primary OI too), and the mechanisms that matter:

• BISPHOSPHONATES (IV pamidronate or zoledronate; pamidronate has the most pediatric experience) for symptomatic vertebral fractures, ongoing fragility fractures, or a heavy glucocorticoid burden. MECHANISM: stable pyrophosphate analogs that bind avidly to hydroxyapatite at resorption surfaces; the osteoclast endocytoses them mid-resorption. NITROGEN-containing bisphosphonates then inhibit farnesyl pyrophosphate synthase in the MEVALONATE pathway -> no prenylation of the small GTPases the osteoclast needs for its ruffled border and migration -> it cannot resorb and undergoes apoptosis (the older non-nitrogen ones form a toxic ATP analog instead). Pediatric tells: zebra lines (dense metaphyseal bands, one per cycle, evenly spaced) and possible metaphyseal under-modeling. Check 25-OH-D (>20-30 ng/mL) and normocalcemia BEFORE dosing (acute-phase reaction + hypocalcemia risk; a first-dose flu-like reaction is common). Jaw osteonecrosis has NOT been reported in children. Anti-resorptives slow the remodeling that clears damaged bone, but starting ~1-2 weeks after a fracture still allows strong callus.
• DENOSUMAB (anti-RANKL monoclonal, SC) blocks osteoclast formation and survival. The pediatric catch: it is NOT retained in bone, so on withdrawal there is a REBOUND surge of resorption — expect on-dose hypocalcemia and between-dose / withdrawal REBOUND HYPERCALCEMIA (treat the rebound with a bisphosphonate). Useful where bisphosphonates fail or are contraindicated, but the rebound makes it hard to stop.
• ROMOSOZUMAB (anti-sclerostin, SC) releases WNT signaling -> BOTH anabolic AND anti-resorptive; carries an adult cardiovascular (MI) signal and essentially no pediatric experience.
• TERIPARATIDE (recombinant PTH 1-34, anabolic) is CONTRAINDICATED in children — the rat osteosarcoma signal rules it out while the growth plates are open.
• ADJUNCTS, not monotherapy: calcium to the age RDA (~700 mg under 3 yr, ~1000 mg at 4-8, ~1300 mg at 9-18; more with malabsorption or glucocorticoids, though calcium alone barely moves BMD), vitamin D repletion, and high-impact weight-bearing activity in the prepubertal / early-pubertal window (the mechanostat; whole-body vibration for the immobile). Always treat the underlying disease.

Pediatric hyperparathyroidism is rare but reviewers love the framework. The trick is to anchor each form to its DRIVER — the question "why is PTH high?" splits the differential cleanly.

PRIMARY HYPERPARATHYROIDISM (PRIMARY = AUTONOMOUS): The parathyroid is the problem. It oversecretes PTH independent of calcium. Bloodwork: HIGH PTH, HIGH calcium, LOW phosphate (PTH drives phosphaturia), HIGH or normal-high 1,25(OH)2D (PTH activates 1α- hydroxylase), HIGH urine calcium (filtered load wins over PTH's distal reabsorption). The high-urine-calcium piece distinguishes 1°HPT from FHH where urine calcium is LOW. In pediatrics, 1°HPT is almost always SYNDROMIC — send a panel:

• NSHPT (CASR homozygous LoF) — neonatal severe; urgent total parathyroidectomy.
• MEN1 (MEN1 / menin) — usually all-four-gland hyperplasia in teens; the most common pediatric cause of 1°HPT.
• MEN2A (RET) — 15-30% develop 1°HPT, usually later than medullary thyroid carcinoma (MTC).
• HPT-JT (CDC73 / parafibromin LoF) — jaw fibroma + parathyroid CARCINOMA risk + renal cysts/Wilms.
• FIHP (familial isolated HPT) — various genes including GCM2 GoF.
• Sporadic adenoma — exists in peds but rare; image, localize, resect.

Treat: parathyroidectomy. Calcimimetic (cinacalcet) as bridge in inoperable disease or while awaiting surgery in NSHPT.

SECONDARY HYPERPARATHYROIDISM (SECONDARY = COMPENSATORY): The parathyroid is RESPONDING APPROPRIATELY to a low-calcium / high-phosphate / low-vitamin-D signal. Bloodwork: HIGH PTH, LOW or NORMAL calcium, variable phosphate (HIGH in CKD, LOW in vit-D deficiency / rickets / GI malabsorption). Drivers in pediatrics:

• CHRONIC KIDNEY DISEASE — phosphate retention, low 1α-hydroxylation, reduced calcium absorption, FGF23 elevation. The pediatric CKD-MBD syndrome.
• VITAMIN D DEFICIENCY — nutritional or vitamin D-dependent rickets (VDDR) variants.
• INTESTINAL MALABSORPTION — celiac, IBD, CF, post-resection.
• HYPERCALCIURIC LOSS / Bartter / loop diuretics chronically.

Treat: fix the upstream driver. Calcium + vitamin D for nutritional; phosphate binders + calcitriol + paricalcitol for CKD; address GI malabsorption.

TERTIARY HYPERPARATHYROIDISM (TERTIARY = AUTONOMY AFTER LONG COMPENSATION): Long-standing 2°HPT eventually transitions to AUTONOMOUS hypersecretion — the parathyroid glands hyperplastic from years of overdrive lose their setpoint regulation and now overshoot. Bloodwork: HIGH PTH + HIGH calcium (the giveaway — compensatory 2°HPT has high PTH with low/normal calcium; once calcium climbs, it's 3°HPT). Almost always in the setting of long-standing CKD, especially post-transplant when the new kidney corrects phosphate / 1,25D but the hyperplastic glands keep firing. Other settings: prolonged untreated X-linked hypophosphatemia (XLH) on phosphate without burosumab (the historic complication), prolonged vit-D-deficient rickets undertreated for years. Treat: parathyroidectomy (often subtotal 3.5-gland or total + autotransplant); cinacalcet as bridge.

ECTOPIC PTH-LIKE OR PTHrP-DRIVEN HYPERCALCEMIA — separate entity, PTH itself is SUPPRESSED. PTHrP from malignancy (humoral hypercalcemia of malignancy, rare in peds but described in rhabdoid, Wilms, neuroblastoma). PTH-secreting NETs essentially nonexistent in peds.

The quick framework:

BIOCHEMICAL PHENOTYPES OF PRIMARY HPT (orthogonal to the genetic classification above):

• CLASSICAL PHPT — high Ca, HIGH (or inappropriately non-suppressed) PTH.
• NORMOHORMONAL PHPT — high Ca with PTH in the UPPER NORMAL range; PTH should be fully suppressed at that Ca, so it isn't truly normal.
• NORMOCALCEMIC PHPT — persistently elevated PTH with NORMAL Ca, after excluding vit D deficiency, low Ca intake, hypercalciuria / CKD, and drugs (thiazides, lithium, bisphosphonates, denosumab). A subset evolves to classical PHPT over years.

THE GERMLINE LESIONS CLUSTER INTO THREE BIOLOGICAL THEMES: 1. LOSS OF TUMOR SUPPRESSORS that constrain parathyroid proliferation — MEN1, CDC73, CDKN1B. These produce hyperplasia / adenomas with classical "high Ca + high PTH" picture and HIGH RECURRENCE after subtotal surgery (especially MEN1). 2. GAIN-OF-FUNCTION in growth signaling — RET in MEN2A. PHPT here is INCIDENTAL and MILDER; the disease is driven by C-cell and chromaffin pathology. 3. SET-POINT DISORDERS of the calcium-sensing axis — CASR, GNA11, AP2S1. Heterozygous LoF raises the set point -> FHH (mild hypercalcemia, hypocalciuria, NOT surgical). Biallelic LoF abolishes feedback -> NSHPT, a true emergency. The same axis with activating mutations gives ADH (autosomal dominant hypocalcemia) — the MIRROR IMAGE, worth remembering when you see "hypoparathyroidism with hypercalciuria".

PEDIATRIC-SPECIFIC PRACTICAL POINTS:

• ANY CHILD OR ADOLESCENT WITH PHPT GETS GERMLINE TESTING. Sporadic disease in this age group is the exception, not the rule. Panel should include at minimum MEN1, CDC73, RET, CASR, CDKN1B, GCM2, AP2S1, GNA11.
• ALWAYS CALCULATE FECa (or 24-h Ca/Cr clearance ratio) BEFORE SURGERY. <0.01 is the FHH cutoff; missing this leads to unnecessary parathyroidectomy in a disease that doesn't need it.
• In neonates with severe hypercalcemia, get PARENTAL Ca and CASR status urgently — distinguishes NSHPT (biallelic) from the milder TRANSIENT neonatal hyperparathyroidism of a HETEROZYGOUS infant, which is worst when the allele is paternally-derived or de-novo in a NORMOCALCEMIC mother (the fetus reads normal maternal calcium as low) and is set-point-CONCORDANT, hence mild, when inherited from an FHH mother; an UNAFFECTED infant of an FHH mother instead gets transient HYPOcalcemia. §87
• CDC73-related disease has the HIGHEST CARCINOMA RISK and warrants en bloc resection of the affected gland with consideration of ipsilateral thymectomy, plus lifelong surveillance for jaw, renal, and uterine lesions.

OSTEITIS FIBROSA CYSTICA & THE BROWN TUMOR — the skeletal end-organ disease of sustained PTH excess, and form-agnostic (1°, 2°, or 3° all do it). Chronic high PTH drives relentless OSTEOCLASTIC resorption; marrow is replaced by PERITRABECULAR FIBROSIS, and focal resorption cavities fill with osteoclast-type GIANT CELLS, fibrovascular stroma, and old hemorrhage. The hemosiderin and vascularity of that repair tissue is what gives the lesion its BROWN color and its name — it is a "tumor" in name only, NON-NEOPLASTIC, a reparative mass rather than a growth.

• WHERE: jaw (mandible / maxilla), long bones, ribs, pelvis, clavicle, phalanges — expansile lytic lesions. The surrounding osteitis fibrosa carries the classic radiographic set: SUBPERIOSTEAL RESORPTION on the RADIAL side of the middle phalanges of the 2nd-3rd fingers (the most specific sign), a "salt-and-pepper" skull, distal-clavicle and phalangeal-tuft resorption, and loss of the dental LAMINA DURA. ALP runs HIGH (high bone turnover) — a clue the bone is involved.
• THE TRAP: the histology is giant-cell-rich and OVERLAPS a giant cell tumor of bone or a central giant cell granuloma of the jaw. So any giant-cell bone lesion — a jaw lesion in a child especially — earns a calcium + PTH BEFORE anyone calls it a primary giant cell tumor. A brown tumor is occult hyperparathyroidism until the labs say otherwise.
• PEDS CONTEXT: in the developed world the usual driver is now RENAL osteodystrophy (2°/3° HPT of CKD-MBD) or undertreated 3° (the historic phosphate-without-burosumab XLH, §20), rather than the florid 1°HPT that labs now catch early.
• TREATMENT = TREAT THE HYPERPARATHYROIDISM, not the lesion. Parathyroidectomy (1°/3°) or correction of the upstream driver (2°) normalizes PTH, and the brown tumors REMINERALIZE and regress. Direct bone surgery is reserved for pathological fracture, cord / nerve compression, or a lesion that fails to heal once PTH is fixed. The pearl: fix the gland, the bone heals itself.

HOW SEVERE / HOW URGENT (grades any hypercalcemia, whatever the PTH driver):

Mild and asymptomatic (or barely symptomatic — constipation) needs no emergency treatment. CHRONIC 12-14 is usually well tolerated — the same conservative measures as mild, still no emergency. What flips it to a code is Ca >14, OR a 12-14 that is RISING ACUTELY: that is where the sensorium drops and the patient deteriorates fast, and it earns urgent, aggressive therapy.

A sporadic solitary adenoma is the single most common cause of sporadic 1°HPT — one clonal gland oversecreting, three normal. But that is the ADULT framing. In CHILDREN the reframe is the whole point: pediatric 1°HPT is far more often GENETIC/SYNDROMIC than sporadic, so a single adenoma in a child is a diagnosis of EXCLUSION until germline testing is back. The adult thinks "adenoma until proven otherwise"; the pediatric attending thinks "syndrome until proven otherwise".

THE DRIVERS OF PEDIATRIC / YOUNG-ADULT PHPT:

• MEN1 (menin LoF) — the most common heritable cause. Biology is MULTIGLANDULAR HYPERPLASIA, NOT a solitary adenoma — the surgical trap, because focused single-gland removal guarantees recurrence.
• MEN2A (RET GoF) — 1°HPT in a minority, usually mild and LATE, overshadowed by MTC/pheo; the parathyroid is the afterthought, not the lead.
• CDC73 / HRPT2 (parafibromin LoF) -> HPT-JT — ossifying jaw fibromas, renal cysts/hamartomas/Wilms, uterine tumors, and the CAN'T-MISS: the highest parathyroid CARCINOMA risk of any germline lesion.
• GCM2 GoF — familial isolated HPT (FIHP), no syndromic stigmata.
• CASR-spectrum (FHH/NSHPT) is the set-point mimic that must be excluded, not operated — owned by §23.

CARCINOMA RED FLAGS (think CDC73): Markedly high calcium (often >14 mg/dL), PTH often >5x ULN (vs the modest elevation of a benign adenoma), a PALPABLE neck mass, and severe synchronous end-organ disease (bone + stone + renal). Any one should trigger en bloc resection planning, not enucleation.

LOCALIZATION — FOR THE SURGEON, NOT THE DIAGNOSIS: Neck ultrasound + sestamibi-SPECT/CT, +/- 4D-CT. KEY POINT: imaging localizes; it does NOT diagnose. PHPT is a BIOCHEMICAL diagnosis (high Ca + non-suppressed PTH) — a negative scan never overturns the biochemistry, and a "lesion" never makes the call on its own. And FHH must be excluded FIRST — 24h urine Ca / Ca:creatinine clearance ratio (<0.01 = FHH) — before any knife touches the neck.

FNA-PTH WASHOUT for the equivocal lesion: aspirate the suspect target, rinse the needle in saline, and assay PTH in the washout — a level far above serum (often many-fold) confirms PARATHYROID tissue (vs a thyroid nodule, lymph node, or fat). Earns its keep for the intrathyroidal or ectopic adenoma and the re-operative neck. Distinct from the INTRA-OP PTH drop below: washout LOCALIZES the gland pre/peri-op, ioPTH CONFIRMS it is out.

SURGERY:

• FOCUSED (minimally invasive) parathyroidectomy when a single adenoma is well-localized, guided by INTRA-OP PTH: >50% drop from baseline at 10 min = cure (the half-life of PTH is ~2-4 min, so a real drop is fast).
• BILATERAL neck exploration when imaging is negative or multiglandular disease is expected.
• MEN1 needs SUBTOTAL (3.5-gland) or TOTAL parathyroidectomy + heterotopic autotransplant (+ cervical thymectomy) — because the disease is MULTIGLANDULAR and recurs; a focused operation is the wrong operation here.

PEDIATRIC OPERATIVE THRESHOLD: Lower than adults. Adult guidelines tolerate watchful waiting for mild asymptomatic disease; children don't qualify — pediatric PHPT is usually SYMPTOMATIC and/or GENETIC, so essentially all warrant parathyroidectomy, done at a HIGH-VOLUME center. FIRST exclude FHH (CASR) — it mimics PHPT but does NOT benefit from parathyroidectomy.

HUNGRY BONE SYNDROME — the dangerous EARLY post-parathyroidectomy trap: once chronically high PTH is suddenly removed, the demineralized skeleton avidly re-uptakes calcium (plus phosphate + magnesium), driving PROFOUND, prolonged HYPOCALCEMIA within 1-4 days. Highest after long-standing high-turnover disease — 3° HPT / renal osteodystrophy, large adenomas, brown tumors, very high pre-op PTH/ALP. Pre-load and aggressively replace calcium + calcitriol (+ magnesium) peri-operatively and monitor for days. Distinguish from transient POST-SURGICAL HYPOPARATHYROIDISM (PTH LOW): in hungry bone PTH is appropriately non-suppressed (normal-to-high, rising as Ca falls) — the bone is simply swallowing the calcium.

The gland runs an assembly line: trap iodide, organify it onto thyroglobulin, couple the iodotyrosines into T4/T3, store it in colloid, then release and deiodinate to the active T3. Every congenital dyshormonogenesis is a named break in that line; the thionamides (methimazole, propylthiouracil) are a deliberate block at organification, and PTU additionally blocks the peripheral T4 -> T3 step.

Graves = TSI stimulating the TSH receptor. Diagnosis: suppressed TSH, high fT4 or fT3 (fT3 more sensitive for overt), TRAb (TSHRAb) positive. fT3 disproportionate elevation = synthesis disease. Pediatric medical + definitive treatment in detail in §34.

Hashitoxicosis = destructive Hashimoto phase. Preformed hormone leak. BIPHASIC (toxic -> hypo) — destruction releases stored hormone, then once the depot is empty the gland fails. Radioactive iodine uptake (RAIU) is LOW (the key discriminator from Graves). MMI is ineffective — this is not a synthesis problem. The TRIPHASIC pattern (toxic -> euthyroid -> hypo -> recovery) belongs to subacute / de Quervain thyroiditis and silent / postpartum thyroiditis, not to Hashitoxicosis — those gland do recover. The analogy to triphasic post-pit-surgery DI fits subacute, not Hashitoxicosis.

Pediatric thyrotoxicosis differential discriminators:

• Graves: TRAb+, RAIU high diffuse.
• Hashitoxicosis: TRAb-, TPO+/Tg+, RAIU low.
• Toxic adenoma: TRAb-, focal hot scan.
• TSH-oma: TRAb-, MRI shows adenoma, alpha-subunit/TSH >1.
• Exogenous (factitia): TRAb-, RAIU low, AND TG IS LOW. Only one with low Tg.
• Struma ovarii: ectopic functioning thyroid — LOW neck RAIU but HIGH Tg, with PELVIC uptake on whole-body scan (the low-RAIU-but-HIGH-Tg trap that separates it from factitia).
• hCG-mediated: very high hCG cross-stimulating the TSHR (molar pregnancy, germ-cell tumor) — TRAb-, transient; the adolescent-girl trap.
• McCune-Albright: GNAS R201 mosaic, multinodular autonomous gland.

NONTHYROIDAL ILLNESS (NTI, formerly "SICK EUTHYROID SYNDROME") — the adaptive thyroid response to systemic illness. Three phases that move from low T3 only -> low T3 plus low T4 -> recovery, with TSH following its own trajectory. Critical for the ICU consult question "do we replace?":

THE MECHANISTIC LOGIC:

• PHASE I (low T3 syndrome). Illness induces deiodinase shifts at the peripheral tissue level: TYPE 1 deiodinase (D1) DOWN -> less T4->T3 conversion in liver / kidney; TYPE 3 deiodinase (D3) UP -> increased T4 -> rT3 (reverse T3, inactive). Net: T3 drops, rT3 rises, T4 stays normal. TSH stays normal because the hypothalamus / pituitary haven't been pushed yet. This is ADAPTIVE — a "low metabolic demand" state, sparing the body during acute illness. DO NOT REPLACE.
• PHASE II (low T3, low T4 syndrome). With prolonged illness, central signaling is also affected: hypothalamic TRH falls (cytokines, cortisol, leptin), so TSH falls, so T4 production falls. T3 continues to be low from peripheral deiodinase pattern. T4-binding protein levels also drop (acute-phase response), so total T4 looks particularly low while FREE T4 may be less reduced. The discriminator from true central hypothyroidism is the clinical context (the patient is critically ill) and the rT3 (HIGH in NTI, LOW or normal in central hypo). DO NOT REPLACE empirically — trials of T3 or T4 in NTI have not improved outcomes and may worsen them.
• PHASE III (recovery). As the illness resolves, the hypothalamic / pituitary axis re-activates first while peripheral T4 is still low. TSH rises briefly to push T4 up. This TSH OVERSHOOT can be 5-20 mIU/L for 1-2 weeks before normalizing. The classic trap: TSH drawn during this window -> diagnostic confusion with newly-acquired primary hypothyroidism. The discriminator is the trajectory (resolves spontaneously over 1-2 weeks) and the rT3 (normalizing). Do NOT start levothyroxine on a recovering ICU patient with TSH 12 unless you've confirmed it stays high after 2-4 weeks of recovery.

NTI vs CENTRAL HYPOTHYROIDISM — the two settings most likely to be confused:

NTI in pediatrics matters especially in PICU (sepsis, post-cardiac surgery, severe trauma), in chronic illness (CKD, severe IBD, anorexia nervosa, oncology patients on chemotherapy), and in newborn screening (sick newborns can flag false-positive abnormal TFTs that resolve as they recover).

SICK EUTHYROID is the old name. NONTHYROIDAL ILLNESS / NTI is the current term — it doesn't presuppose euthyroidism (the patient is actually low-T3 hypo, just appropriately so) and aligns with the adaptive-physiology framing.

FOUR LEVELS at which a newborn ends up hypothyroid — name the level before you reach for a gene: (1) the gland was NEVER BUILT right = DYSGENESIS; (2) the gland is built but a STEP in hormone synthesis is broken = DYSHORMONOGENESIS; (3) the gland is fine but UNDER-DRIVEN from above = CENTRAL; (4) the gland is normal but STARVED of substrate = IODINE deficiency. Each gene below slots under one level and tells you exactly what is broken.

THE EPIDEMIOLOGY (anchors the differential before the genes):

• DYSGENESIS is the most common PERMANENT cause (~85% of permanent CH). Within it, ECTOPY (sublingual / lingual remnant) is ~2/3, athyreosis / agenesis next, hypoplasia the rest. Mostly SPORADIC — only a monogenic minority (genes below).
• DYSHORMONOGENESIS is ~10-15% in outbred populations but climbs to 20-25% where CONSANGUINITY is common (Turkey included). It is AR, so it RECURS in siblings — the heritable bulk of CH.
• IODINE DEFICIENCY is the most common TRANSIENT cause, and the most common cause of CH WORLDWIDE — the gland is substrate-starved, not broken.

DYSGENESIS — the FACTORY ITSELF was never built right: the gland is absent, ectopic (sublingual remnant), or hypoplastic. Mostly sporadic; the monogenic minority are DEVELOPMENTAL transcription factors, and the EXTRA-thyroidal features name the gene:

• TSHR LoF — the gland cannot HEAR the TSH "grow and make hormone" signal -> hypoplasia / a resistance spectrum (high TSH, a normally-sited small gland, normal Tg).
• NKX2-1 — master TF for thyroid + lung + basal ganglia -> BRAIN-LUNG-THYROID: choreoathetosis + neonatal RDS + CH.
• FOXE1 — thyroid-migration + palate/hair TF -> Bamforth-Lazarus: cleft palate + spiky hair + bifid epiglottis + athyreosis.
• PAX8 — thyroid + kidney TF -> CH + RENAL anomalies.
• GLIS3 — beta-cell + thyroid TF -> NEONATAL DIABETES + CH + glaucoma + polycystic kidneys.

DYSHORMONOGENESIS — the gland is BUILT and TSH-driven, but ONE STATION on the iodide-to-hormone assembly line is broken. Walk the line in order — get iodide IN, push it across into the colloid, lay down the scaffold, weld iodine onto it, then recycle the scraps — and each gene is one broken station. All AR (so they RECUR in siblings) and GOITROUS AS A RULE (TSH flogs the straining gland) — with NIS the exception (below). What is actually wrong:

• NIS / SLC5A5 — the IODIDE PUMP on the blood side is dead, so iodide never gets INTO the gland: the factory is starved of raw material. LOW radioactive iodine uptake (RAIU) that an iodide load does NOT rescue (salivary + gastric iodide are low too — NIS sits there as well). It is the ODD ONE OUT of the list: often NON-goitrous, and the only one whose uptake scan reads BLANK (nothing is trapped) so it MIMICS agenesis/athyreosis — ULTRASOUND is what proves a eutopic gland is in place. (Tracer uptake stays low, but pharmacologic high-dose iodide can still diffuse in passively and partly treat.)
• Pendrin / SLC26A4 — iodide is inside the cell but the APICAL DOOR into the colloid won't open, so it can't reach where hormone is assembled. The same transporter sits in the inner ear, so loss = Pendred: goiter + sensorineural deafness (Mondini cochlea / enlarged vestibular aqueduct). Partial organification defect, perchlorate discharge >15%.
• TG (thyroglobulin) — the SCAFFOLD PROTEIN that is supposed to be fed iodine is missing or malformed, so there is nothing to iodinate. LOW serum Tg with a LARGE goiter.
• TPO (thyroid peroxidase) — the ENZYME that welds iodine onto the scaffold is dead: iodide arrives but cannot be organified at all. TOTAL organification defect, perchlorate discharge >90%.
• DUOX2 / DUOXA2 — TPO's hydrogen-peroxide supply fails, so the welding torch has no fuel and organification stalls. Usually PARTIAL / TRANSIENT (a backup oxidase covers some) — often presents as neonatal hyperthyrotropinemia that resolves.
• IYD / DEHAL1 — the iodide RECYCLER is broken: spent iodine on MIT/DIT is dumped in the urine instead of being salvaged, so the gland runs an internal iodine LEAK — late-onset goiter that MIMICS dietary iodine deficiency.

CENTRAL CH — the gland is FINE but UNDER-DRIVEN; the signal from above is missing:

• TSHB — no bioactive TSH is made (isolated central hypothyroidism).
• TRHR — the thyrotroph cannot hear hypothalamic TRH (isolated).
• IGSF1 — X-linked central hypothyroidism + adult MACROORCHIDISM + low prolactin (mechanism below).

IGSF1 syndrome belongs here because it is the one CENTRAL-hypothyroidism gene with a striking testicular phenotype: IGSF1 (a pituitary membrane protein) loss drops thyrotroph TRH-receptor expression (-> central hypothyroidism) AND, by an incompletely understood mechanism, is associated with adult macroorchidism -> mild central hypothyroidism + delayed pubertal T rise + low prolactin + adult MACROORCHIDISM (30-45 mL), with fertility preserved. X-linked; carrier females may show only mild central hypothyroidism. The full big-testes differential lives in §63.

NEONATAL GOITER — etiology splits by maternal vs fetal source, and most are TRANSIENT. Two columns decide it: is the baby HYPER or hypo, and will it RESOLVE:

URINARY IODINE STATUS — WHO/UNICEF/ICCIDD population cutoffs (use spot urinary iodine concentration, μg/L):

• <20 = severe deficiency
• 20-49 = moderate deficiency
• 50-99 = mild deficiency
• 100-199 = adequate
• 200-299 = above requirements
• ≥300 = excessive (risk of iodine-induced hyperthyroidism in some, thyroiditis in others)

PREGNANCY targets are HIGHER: median 150-249 μg/L is adequate. In NEONATES with positive NBS for CH, urinary iodine helps distinguish transient maternal iodine excess (very high UI from povidone-iodine skin prep, contrast, amiodarone) — the gland is acutely blocked (Wolff-Chaikoff), TFTs normalize in weeks. The iodine-deficient transient CH on the opposite end resolves with maternal iodine repletion.

The clean call: serum free T4 BELOW the age reference WITH a high TSH = primary CH confirmed — start levothyroxine (LT4) IMMEDIATELY, do not wait. The hard part is the milder, TSH-only and grey-zone cases. Two converging frameworks (ESPE / ENDO-European, van Trotsenburg 2021; and the age-graded venous cutoffs) drive the decision.

THE SCREENING PIPELINE (ESPE / ENDO-European):

• Heel-prick (filter-paper) TSH ≥40 mIU/L -> draw the confirmatory serum sample and START LT4 WITHOUT waiting for the venous result. High-probability CH; lost time is lost IQ.
• fT4 normal but serum TSH PERSISTENTLY >20 mIU/L (around the 2nd week of life) -> treat.
• At ~4 weeks of age, TSH still >10 mIU/L -> treat.

THE AGE-GRADED VENOUS TSH CUTOFFS (fT4 normal — when an isolated high TSH earns LT4):

• 14 days and older with TSH 5-20 mIU/L: do NOT commit — recheck every 2-3 weeks and treat only if the control TSH is >10 AND on a RISING trend.

THE GREY ZONE (fT4 normal, TSH 6-20, clinically well baby): two defensible moves — (1) start LT4 now, or (2) work the cause up (family history, thyroid imaging, genetics) and if TSH is still up at 3-4 weeks, discuss with the family and start LT4 then; a 1-2 week re-test before committing is acceptable.

THE TRAP — the TRANSIENT mimics you must not over-treat (they cluster in exactly the fT4-normal, TSH 6-20 band):

• Transient neonatal hyperthyrotropinemia, prematurity, maternal IODINE deficiency or excess (Wolff-Chaikoff), heterozygous gene variants.
• The neurodevelopmental benefit of LT4 in this band is NOT established, so unnecessary treatment is the error to avoid — confirm persistence, and plan a trial off therapy around age 3 to prove permanence (the transient-cause list lives in §31).

THE TARGET once you treat:

• Start LT4 at 10-15 μg/kg/day (the higher end for severe CH / very low fT4); the goal is FAST normalization.
• Keep serum fT4 in the UPPER HALF of the age-specific reference range and TSH within age-specific normal limits.
• fT4 normalizes BEFORE TSH — the pituitary lags. Aiming for upper-half fT4 with rapid TSH normalization is what improves neurodevelopmental outcome; do not starve the brain chasing a perfect TSH first.

MONITORING SCHEDULE:

• 1-2 weeks after the first dose, then every 2 weeks until thyroid hormone levels normalize.
• First 12 months of life: every 1-3 months.
• 12 months to 3 years: every 2-4 months.
• Thereafter until growth is complete: every 3-6 months.
• After ANY dose change: re-measure fT4 and TSH at 4-6 weeks (steady state).

THE DOSING TRAP: do NOT cut the LT4 dose on a SINGLE higher-than-normal fT4 unless the TSH is SUPPRESSED or there are clinical signs of over-treatment (irritability, tachycardia). fT4 runs high transiently — especially when blood is drawn soon after the morning dose — and TSH is the steadier guide. One isolated high fT4 is not over-replacement.

Hashimoto (chronic autoimmune / lymphocytic) thyroiditis is the commonest cause of acquired hypothyroidism and acquired goiter in children and adolescents in iodine-replete regions — peak in adolescence, girls > boys (Salerno 2020).

MECHANISM: T-cell-mediated lymphocytic infiltration + anti-TPO (most sensitive) and anti-thyroglobulin antibodies. Strong autoimmune clustering — screen for and expect it alongside T1DM, celiac disease, vitiligo, and the autoimmune polyglandular syndromes; markedly over-represented in Turner, Down, and Klinefelter (screen TSH on schedule in those).

THE PRESENTATION SPECTRUM is the key idea — one disease, a sliding biochemical scale:

• EUTHYROID: positive antibodies +/- a firm bosselated goiter, normal TSH / fT4.
• SUBCLINICAL HYPOTHYROIDISM: HIGH TSH, NORMAL fT4.
• OVERT HYPOTHYROIDISM: HIGH TSH, LOW fT4 — fatigue, growth deceleration WITH weight gain, delayed bone age, cold intolerance, constipation, goiter; occasionally the Van Wyk-Grumbach picture (very high TSH cross-activating FSHR -> precocious puberty / macroorchidism, §63).
• A transient toxic phase = Hashitoxicosis (destructive leak, LOW RAIU; §30) can open the course.

Goiter is firm / rubbery and bosselated, or the gland ends up ATROPHIC. Ultrasound: heterogeneous, hypoechoic gland with pseudonodules.

WHEN TO TREAT (Salerno 2020):

• OVERT hypothyroidism (high TSH + low fT4) -> levothyroxine, always (dosing + monitoring in §32).
• SUBCLINICAL (high TSH, normal fT4): TSH persistently >10 mIU/L -> treat. TSH 5-10 with normal fT4 -> mostly OBSERVE — much of pediatric subclinical hypothyroidism (especially mild and antibody-negative) is benign and REMITS spontaneously, so resist reflexive LT4. Tip toward treating only with symptoms, a growing goiter, a rising TSH trend, positive anti-TPO with progression, or an at-risk syndrome.

The complete pediatric Graves treatment workflow per ETA 2022 (Mooij CF et al, Eur Thyroid J 2022). Diagnostic anchors are in §30 (big ideas / pediatric thyrotoxicosis differential).

ATD CHOICE (ETA 2022): MMI (methimazole / thiamazole) or carbimazole (CBZ — prodrug, metabolized to MMI). MMI 0.6 mg ≈ CBZ 1.0 mg — remember the 0.6:1 conversion. Propylthiouracil (PTU) CONTRAINDICATED in peds (FDA black-box hepatocellular failure, highest in children) except thyroid storm and 1st-trimester pregnancy. Both MMI/CBZ given ONCE DAILY.

DOSE TITRATION (DT) is the PREFERRED approach (recent RCT: equal biochemical control to BR with fewer adverse events):

Titration rule once euthyroid (around 4-6 wk): reduce dose by 25-50% if euthyroid, by 50% if hypothyroid. Long-term maintenance often drops to 2.5-5 mg MMI/CBZ daily. If raised TSH despite minimum daily dose (2.5 mg) AND normal TSHRAb -> consider remission, may stop.

BLOCK AND REPLACE (BR), alternative when DT compliance is unstable: MMI 0.3-0.5 mg/kg/day (CBZ 0.5-0.75 mg/kg/day) fixed — blocks endogenous synthesis. Add weight-appropriate LT4 when fT3 falls into the reference range. If fT3 doesn't fall as expected, escalate to MMI 1.0 mg/kg/day or CBZ 1.3 mg/kg/day. More side effects than DT in the recent RCT, so reserve for the right family.

MONITORING SCHEDULE: TFTs every 4 wk for the first 3 months, then every 2-3 months. The teaching point: USE fT4 AND fT3 to titrate in the first 4-6 months, NOT TSH. TSH stays SUPPRESSED for weeks-months after euthyroid because the long-suppressed thyrotropes recover slowly — a low TSH at 8 weeks doesn't mean undertreatment. After 4-6 months, TSH becomes interpretable and joins fT4/fT3 in titration.

PRE-TREATMENT BASELINE: FBC with neutrophil count, LFTs (because GD itself can lower WBC and raise transaminases — you need a baseline to interpret later changes).

SIDE EFFECTS (~15% of pediatric GD on ATD have at least one):

• Minor cutaneous (rash, urticaria) ~10% — often transient, symptomatic management.
• AGRANULOCYTOSIS (~0.2-0.5% on antithyroid drugs; median onset ~day 30): neutrophil count <0.5 x10⁹/L -> STOP ATD, alternative treatment. Neutrophil 0.5-1.5 -> close monitoring (1-2x/wk CBC).
• HEPATOTOXICITY with MMI/CBZ is CHOLESTATIC (resolves on stop), unlike PTU's hepatocellular damage. Transaminases >3x ULN -> STOP.
• Stevens-Johnson signal: mucosal blistering with rash -> STOP immediately.
• Most AEs in first 3 months; younger children at higher rate; dose-dependent. After serious AE -> never re-challenge.
• Family teaching: fever + sore throat + mouth ulcers = stop drug immediately, check CBC.

BETA-BLOCKER (propranolol or atenolol, weight-appropriate) for adrenergic symptoms during ramp-up. Contraindicated in asthma. Stop when biochemically euthyroid.

ATD DURATION + REMISSION: minimum 3 years, extend to 5+ years if remission predictors are weak. Remission rate ~20-30% at 2 years, rising to ~43% at 5-6 years and ~75% at 9 years.

GOOD-PROGNOSIS (remission-predicting) FACTORS — the more a child stacks, the more a long ATD course is worth it (ETA 2022, Table 5):

• OLDER age at diagnosis (pubertal > prepubertal).
• LOW TSHRAb titer at diagnosis — and crucially a titer that has NORMALISED by the planned stop point.
• SMALL goiter.
• MILD biochemical disturbance at diagnosis.
• LONGER ATD duration before the stop attempt.
• Female (adult data), Caucasian ethnicity.

TSHRAb falls ~90% over 3 years of ATD — ELEVATED TSHRAb at the planned stop point predicts relapse, so DON'T stop. Most relapses occur within 12 months of stopping — so don't stop during an exam-prep year.

THYROID STORM is decompensated thyrotoxicosis — a CLINICAL diagnosis, not a lab one: hyperpyrexia, tachyarrhythmia / high-output failure, CNS (agitation -> delirium -> seizure / coma), GI (vomiting, diarrhea, jaundice). It almost always has a PRECIPITANT — infection, surgery, DKA, trauma, iodine / contrast load, RAI, abrupt ATD withdrawal — so HUNT it. The bundle, in ORDER:

• ANTITHYROID DRUG FIRST: PTU is PREFERRED here (the one setting where PTU beats MMI in peds — it ALSO blocks peripheral T4 -> T3 conversion).
• IODINE (Lugol's / SSKI / KI) ≥1 HOUR AFTER the ATD, NEVER before — the ATD must block organification first, or the iodine just FUELS new hormone synthesis; once blocked, iodine shuts down hormone RELEASE (Wolff-Chaikoff).
• BETA-BLOCKER (propranolol) for the adrenergic storm — high doses also blunt T4 -> T3 conversion; esmolol IV if cardiac instability.
• GLUCOCORTICOID (hydrocortisone / dexamethasone): blocks T4 -> T3 conversion AND covers the relative adrenal insufficiency of severe thyrotoxicosis.
• SUPPORTIVE in the ICU: aggressive COOLING + fluids. Acetaminophen for fever, NOT aspirin (salicylates displace T4 from TBG -> RAISE free T4). Treat the precipitant.
• Refractory: bile-acid sequestrant (cholestyramine, interrupts enterohepatic T4 recirculation); plasmapheresis as the last resort.

DEFINITIVE TREATMENT (relapse, AE, non-compliance, large goiter, or GO):

• 131-I: aim for COMPLETE ABLATION (not euthyroidism) — prevents relapse AND removes long-term thyroid cancer risk in radiation- damaged residual tissue. Activity ~15 MBq/g thyroid mass (US- estimated) OR dosimetry ≥300 Gy. Stop ATD 3-7 days before, resume 1-2 days after if needed. Contraindications: pregnancy / breast- feeding, age <5 (absolute), age 5-10 (relative), active GO. ATD- refractory thyrotoxicosis on MMI ≥1.0 mg/kg/day or CBZ ≥1.3 mg/kg/day -> discuss definitive.
• TOTAL THYROIDECTOMY at high-volume pediatric thyroid surgeon (>50 cases/yr per unit). Indicated when fast euthyroidism needed, large or obstructive goiter, age <5 (RAI contraindicated). Must be biochemically euthyroid pre-op (Lugol's 5-10 drops or SSKI 1-4 drops TID for 1-2 wk — not beyond 2 wk, Wolff-Chaikoff escape — if ATD alone can't get there). Pre-op vitamin D repletion (or a short pre-op calcitriol course) reduces transient hypocalcemia. Mortality <0.1%; transient hypocalcemia 22%, permanent hypopara 2.5%, recurrent laryngeal nerve injury 5.4% transient / 0.4% permanent. Start LT4 weight-appropriate immediately post-op.
• GO: mild expectant, selenium supplementation 1-2 mcg/kg/day x 6 mo. Moderate-severe active GO -> IV corticosteroids per EUGOGO. Active GO is a relative contraindication for RAI (steroid cover required if RAI proceeds); SURGERY preferred over RAI in active GO.

GRAVES ORBITOPATHY (GO) in peds is usually mild (70-100% of cases). Eyelid retraction (72%) and proptosis (53%) lead; soft-tissue inflammation less common in kids than adults. Optic neuropathy and motility impairment are rare. Smoking, high TSHRAb, high fT4 at diagnosis are independent risk factors.

GO TREATMENT — restore euthyroidism, STOP SMOKING (the biggest modifiable driver), and stage by ACTIVITY (CAS) and severity:

• MILD (most peds GO): local measures — lubricants, sunglasses, head-of-bed up, prisms for diplopia — plus SELENIUM for 6 months in mild ACTIVE disease. Usually nothing more.
• MODERATE-TO-SEVERE, ACTIVE: first-line is IV methylprednisolone pulses (~4.5 g over 12 wk) PLUS oral mycophenolate — the COMBINATION is EUGOGO 2021 first-line (the MINGO RCT improved the 24-wk response vs IVMP alone, ~71% vs 53%); IVMP monotherapy / a higher 7.5 g course is the fallback. Second-line: teprotumumab, orbital radiotherapy, rituximab, tocilizumab.
• SIGHT-THREATENING (dysthyroid OPTIC NEUROPATHY): urgent high-dose IV steroid, then orbital DECOMPRESSION if it does not turn fast.
• INACTIVE / burnt-out with disfiguring proptosis or stable diplopia: rehabilitative surgery IN ORDER — decompression -> strabismus -> eyelid.

TEPROTUMUMAB is the mechanism-targeted drug: a human anti-IGF-1R monoclonal that breaks the IGF-1R / TSHR signaling complex on ORBITAL FIBROBLASTS, cutting hyaluronan, inflammation, and adipogenesis — so it actually REDUCES PROPTOSIS (~3 mm, decompression-like) and diplopia, in ACTIVE and chronic disease. IV 10 mg/kg then 20 mg/kg q3 weeks x8 (24 wk). The AEs to counsel: HYPERGLYCEMIA (worsens diabetes — optimize and monitor glucose) and SENSORINEURAL HEARING LOSS / tinnitus that can be PERMANENT (baseline + on-treatment audiometry); also muscle spasms, nausea, alopecia, IBD flare, and TERATOGENICITY (IGF-1R drives fetal growth — contraception required). PEDS: it is an ADULT therapy — pediatric GO is mostly mild, so teprotumumab is rarely needed and stays off-label for the rare moderate-severe / sight-threatening child.

NEONATAL HYPERTHYROIDISM — almost always TRANSPLACENTAL maternal Graves. Maternal STIMULATING TSHR antibodies (TRAb / TSI, an IgG) cross the placenta and flog the fetal/neonatal TSHR. It hits ~1-5% of babies born to GD mothers (Graves is ~96% of all pediatric hyperthyroidism, yet neonatal GD is rare overall, ~1 in 25,000-50,000 births) — and UNLIKE childhood Graves (girls >> boys) it strikes boys and girls EQUALLY, because it is the MOTHER'S antibody, not the child's own autoimmunity. Any mother with current OR PAST GD qualifies — even one rendered hypothyroid by prior RAI or thyroidectomy, because TRAb persist for years (RAI actually RAISES them).

THE ANTIBODY BALANCE sets the phenotype: stimulating TRAb (TSI) -> hyper, blocking TRAb (TBII) -> hypo. A baby can carry BOTH and FLIP — transient hypo first, then hyper as the blocking antibody (or maternal drug) clears faster than the stimulating one. (Classic vignette: one mother's three successive babies ran euthyroid, then hyper, then hypo.)

THE TIMING TRAP — the part that burns people: if the mother took an ATD (MMI / PTU cross the placenta too), the baby is born suppressed, looking EUTHYROID or even transiently HYPOthyroid, then REBOUNDS into thyrotoxicosis around DAY 3-5 as the drug washes out but the antibody lingers. Onset can lag a week, with case reports out to ~day 45. So: CORD-blood TRAb predicts risk (cord TSH/fT4 do NOT — skip them), and you recheck TSH + fT4/fT3 at DAYS 3-5 AND 10-14, sooner if symptomatic.

FETAL clues (poorly controlled maternal GD): tachycardia >160/min, goiter, IUGR, ADVANCED bone age (distal femoral epiphysis seen before ~31-32 wk), craniosynostosis / microcephaly, hydrops, prematurity. The fetal-goiter fork — maternal-ATD HYPOthyroid goiter vs TRAb-driven HYPERthyroid goiter (the HYPER end of the neonatal-goiter differential, §31) — is read on Doppler (flow at the RIM = hypo; flow THROUGHOUT = hyper) plus bone age and heart rate, and is managed by TITRATING THE MOTHER'S ATD: up for fetal hyper, down for fetal hypo, aiming to normalise the fetal heart rate.

NEONATAL signs: tachycardia, irritability / tremor, poor feeding with poor weight gain, sweating, stare / proptosis, goiter. SEVERE (and easily mistaken for SEPSIS): thrombocytopenia, hepatosplenomegaly, jaundice, pulmonary hypertension, high-output cardiac failure — mortality up to ~20% untreated, cardiac failure leading.

TREAT: MMI 0.2-0.5 mg/kg/day (some start up to ~1) divided q8-12h; add PROPRANOLOL 2 mg/kg/day for the adrenergic storm (tachycardia, hypertension). NO PTU in neonates (hepatotoxicity). SEVERE / hemodynamic compromise -> add inorganic IODIDE (Lugol's / SSKI) to choke hormone RELEASE, plus a glucocorticoid (blocks T4->T3 and release). Recheck TFTs q1-2 wk and titrate down; it is SELF-LIMITED — stop once TRAb are undetectable, usually 1-3 months. Afterward watch for craniosynostosis and — if the mother ran thyrotoxic and untreated for a long stretch — transient CENTRAL hypothyroidism (high transplacental T4 having suppressed the fetal HPT axis, so the baby can need brief LT4 even after the hyper resolves).

NON-AUTOIMMUNE = PERMANENT — suspect it when there is NO maternal GD history, or the hyperthyroidism does NOT clear on the antibody schedule: an ACTIVATING TSHR germline mutation (AD or de novo) or McCune-Albright (GNAS Gsα). MMI controls it medically but it does not remit — it ultimately needs DEFINITIVE therapy (total thyroidectomy +/- RAI), which can be delayed for months-years while it responds to drug.

Follow the JOURNEY of a thyroid-hormone molecule and the lesion sits at one of three stations: the gland makes T4, a transporter carries it INTO the cell (TRANSPORT), a deiodinase ACTIVATES it to T3 (METABOLISM), and T3 binds its nuclear receptor (ACTION). Break transport, metabolism, or action and you get the SAME biochemical signature — DISCORDANT TFTs (T4 and/or T3 raised with a NON-suppressed TSH) — because the pituitary feedback loop is reading a different signal than the tissue you actually care about. Name the station, then the gene. ETA 2024 (Persani / Chatterjee, Eur Thyroid J) is the framework. The trigger to start the workup is the same in all: discordant TFTs AFTER ruling out drugs (amiodarone, biotin, heparin, furosemide), illness (NTI), and assay interference (macro-TSH, heterophile / anti-iodothyronine antibodies, biotin, binding-protein variants).

ASSAY TRAPS — SPURIOUS TSH / TFTs (exclude these BEFORE chasing a genetic cause; the thyroid parallel to the prolactin hook / macroprolactin traps in §71) (Favresse 2018):

• MACRO-TSH: TSH complexed to anti-TSH IgG (the macroprolactin analog) -> spuriously HIGH TSH in a CLINICALLY EUTHYROID child with normal fT4. Confirm with PEG precipitation (low recovery) or gel filtration. The commonest cause of an isolated, unexplained high TSH.
• BIOTIN (high-dose supplements / "hair-skin-nails", MS megadoses): jams streptavidin-biotin immunoassays, and the DIRECTION DEPENDS ON FORMAT — sandwich assays (TSH) read FALSELY LOW, competitive assays (fT4, fT3) read FALSELY HIGH. The net picture MIMICS GRAVES (low TSH + high fT4/fT3) and can falsely raise TRAb — a factitious-thyrotoxicosis trap. Hold biotin 2-3 days and repeat.
• HETEROPHILE / HUMAN ANTI-ANIMAL ANTIBODIES (HAMA): bridge the capture and detection antibodies -> usually FALSELY HIGH TSH. Catch it with heterophile-blocking tubes, NON-LINEAR serial dilution, or a different analyzer platform.
• ANTI-IODOTHYRONINE (anti-T4 / anti-T3) AUTOANTIBODIES: spurious free-hormone values, often in autoimmune thyroid disease.
• BINDING-PROTEIN variants: FDH (familial dysalbuminemic hyperthyroxinemia, ALB variants) and thyroxine-binding globulin (TBG) excess / deficiency shift TOTAL T4/T3 while the FREE hormones are normal — the discordance shows up only on total-hormone assays.

The operative move: when the TFTs do not fit the child in front of you, suspect the ASSAY before the gland — repeat on a DIFFERENT platform, run PEG (macro-TSH), serial-dilute (heterophile), and ASK ABOUT BIOTIN.

TRANSPORT — MCT8 / SLC16A2 = ALLAN-HERNDON-DUDLEY syndrome (X-linked). MECHANISM, plainly: MCT8 is the main carrier that moves T3 INTO neurons; lose it and T3 cannot ENTER the brain, so the CNS is functionally HYPOTHYROID even though circulating T3 is high. The intraneuronal receptors are normal — it is a DELIVERY problem, not a reception problem. Peripheral tissues reached by other transporters (MCT10, OATPs) still take up the high circulating T3, so the PERIPHERY runs THYROTOXIC. Biochem: LOW T4, HIGH T3, low rT3, elevated T3/rT3 ratio, normal-to-high TSH, HIGH sex hormone-binding globulin (SHBG).

• Phenotype: severe global developmental delay, hypomyelination on MRI, persistent primitive reflexes, dystonia / bradykinesia, feeding difficulty -> failure to thrive.
• LT4 monotherapy WORSENS the periphery — DO NOT give.
• TRIAC (tiratricol, EU-approved) is the recommended therapy: enters cells independently of MCT8, lowers peripheral T3, rescues some neurodevelopment if started early. Target T3 1.4-2.5 nmol/L. DITPA is the backup analog.
• Prenatal SLC16A2 testing via CVS or amnio in MALE pregnancies of known-carrier families — early diagnosis enables in-utero TRIAC trials.
• LT4 + propylthiouracil (PTU) combo as a backup if analogs unavailable.
• Multidisciplinary: peds endo + peds neuro + PT/OT + dietitian (often need NG / G-tube), surveillance for scoliosis, hip subluxation, contractures.

TRANSPORT — OATP1C1 / SLCO1C1 = the BLOOD-BRAIN-BARRIER thyroxine transporter (the MCT8 counterpart at the BBB; it also carries rT3 into astrocytes). Loss gives a BRAIN-SPECIFIC hypothyroidism with essentially NORMAL serum TFTs — peripheral transport is intact — so, unlike MCT8, the clue is clinical / radiologic, not biochemical: childhood-onset PROGRESSIVE NEURODEGENERATION with cognitive decline and brain glucose HYPOMETABOLISM on FDG-PET. Rare (a handful of families). Brain-penetrant analogs (TRIAC, DITPA) are the experimental route. The contrast to hold: MCT8 screams on the labs (low T4 / high T3); OATP1C1 hides behind a NORMAL thyroid panel — suspect it in unexplained childhood neurodegeneration with normal TFTs. (The same shape recurs for the cell-membrane neutral-amino-acid transporter LAT1.)

THE DEIODINASE STORY (you need it for the two metabolism defects below, and for NTI in §30). Deiodinases are SELENOENZYMES that add or strip iodine to switch thyroid hormone ON or OFF — T4 is only a PROHORMONE, and deiodination decides its fate. The rule: OUTER-ring removal ACTIVATES (T4 -> T3); INNER-ring removal INACTIVATES (T4 -> reverse-T3, T3 -> T2). Three enzymes, three jobs:

• D1 (DIO1, liver + kidney) — the systemic T3 FACTORY that feeds CIRCULATING T3, and the rT3 GARBAGE DISPOSAL. Blocked by PTU and amiodarone.
• D2 (DIO2, brain / pituitary / brown fat / muscle, intracellular) — the LOCAL activator that sets INTRACELLULAR T3 where it acts; this is HOW THE PITUITARY SENSES T4 and how the brain makes its own T3. Rises in hypothyroidism to compensate.
• D3 (DIO3, placenta + fetal tissue + brain) — the OFF switch: inner-ring deiodination inactivates T4 -> rT3 and T3 -> T2, shielding the fetus and brain from excess. The dominant deiodinase of fetal life.

So D1/D2 turn hormone ON, D3 turns it OFF. NTI (sick-euthyroid, §30) is just this system dialed for illness — D1 DOWN + D3 UP -> low T3, high rT3. The two genetic metabolism defects now read straight off this map:

METABOLISM — SECISBP2 / SBP2 deficiency — the cell cannot BUILD selenoproteins, and ALL THREE deiodinases ARE selenoproteins, so the entire activation/inactivation machinery is crippled at once. (The SECIS-binding protein loads selenocysteine into 25+ selenoproteins; the other 20-plus explain the multisystem phenotype.)

• Biochem: HIGH T4, LOW-NORMAL T3, HIGH rT3, LOW plasma selenium.
• Phenotype: growth retardation + muscular dystrophy (vital capacity monitoring, polysomnography for nocturnal hypoventilation) + sensorineural hearing loss + thoracic aortic aneurysm + PHOTOSENSITIVITY (UV-sensitive skin) + male infertility.
• LIOTHYRONINE (LT3) for symptomatic low FT3.
• DO NOT supplement selenium (counterintuitive trap) — doesn't help and may worsen oxidative stress.
• Alpha-tocopherol antioxidant may be considered.

METABOLISM — DIO1 / deiodinase-1 defect (NEW in ETA 2024, AD) — lose the rT3 GARBAGE DISPOSAL and reverse-T3 piles up (D1 normally also makes circulating T3).

• Biochem: HIGH rT3, ELEVATED rT3:T3 ratio in the ABSENCE of NTI (the discriminator), otherwise minimal phenotype. Confuses TFTs in combination with other thyroid defects.
• Genetic confirmation by DIO1 sequencing in a family with discordant TFTs and dominant inheritance.
• No specific treatment.

CONSUMPTIVE HYPOTHYROIDISM — the ACQUIRED D3 disease: an infantile HEPATIC HEMANGIOMA (or a large cutaneous hemangioma) overexpresses D3, inactivating T4/T3 faster than the gland can replace -> severe hypothyroidism with HIGH rT3, needing very large LT4 +/- LT3. Treat the hemangioma (propranolol) and the thyroid demand falls. The only "T3-consumption" hypothyroidism — it reads straight off the D3 box above.

RECEPTOR — RTH-α (THRA LoF, the FORGOTTEN-COUSIN resistance — THRA is enriched in heart, bone, brain, GI).

• MECHANISM: the nuclear T3 receptor comes in two genes — THRA (TRα) dominates HEART, BONE, GUT and parts of the CNS, while THRB (TRβ) runs the pituitary feedback loop and liver. In RTH-α the TRα-driven tissues are DEAF to T3 (tissue hypothyroidism -> bradycardia, constipation, delayed bone age, short stature, developmental delay), but the pituitary still senses T3 through intact TRβ2 and keeps TSH normal — hence the NEAR-NORMAL labs that hide the disease.
• Biochem (NORMAL TFTs is the trap): TSH within normal range / mildly raised, low or low-normal T4, RAISED or high-normal T3, LOW rT3. T3:rT3 ratio elevated. Always check rT3 — it's the bedside flag.
• Phenotype "major" criteria (ETA 2024, sequence THRA if 3+ present): macrocephaly, short stature, constipation, the typical biochem profile, developmental delay.
• Minor: anemia (unexplained), hypothyroid-style dysmorphic facies, dyspraxia, skeletal dysplasia.
• Empirical LT4 trial in all patients (helps growth, constipation, GH axis); titrate above usual replacement dose. Reassess GH after LT4 — some kids look GH-deficient until the receptor signal is rescued.

RECEPTOR — RTH-β (THRB, the COMMON form).

• MECHANISM: THRB mutations are DOMINANT-NEGATIVE — the mutant receptor still binds DNA but cannot transactivate when T3 binds, and it POISONS the normal allele. The pituitary thyrotroph feeds back through TRβ2, so it goes DEAF to T3 -> it never shuts off TSH -> TSH stays up and drives T4/T3 higher, which the pituitary still cannot sense (the open loop = HIGH fT4/fT3 with NON-suppressed TSH). Meanwhile the TRα-rich HEART and BONE sense the excess normally -> tachycardia / AF and advanced bone age, while the TRβ1-rich LIVER stays resistant — so SHBG is NORMAL, not raised (the discriminator from TSH-oma / true thyrotoxicosis, where SHBG climbs). One patient, MIXED phenotype, because different tissues run on different receptor isoforms.
• Biochem: HIGH fT4 and fT3 with NON-SUPPRESSED TSH (genuine discordance, not assay artifact).
• First step: rule out drugs / illness / assay interference. Test first-degree relatives — abnormal TFTs in 50% confirms suspicion.
• Sequence THRB. If mutation absent in suspected case (vertical transmission to progeny), do NGS on tissues OTHER than blood — somatic mosaicism is increasingly recognized.

RTH-β vs THYROTROPINOMA / TSH-oma — the differential reviewers love:

RTH-β treatment:

• DO NOT routinely treat with ATD or thyroid ablation — TSH compensates, the patient is functionally euthyroid for most tissues. Treatment is reserved for severe / uncontrolled / life-threatening hyperthyroidism (homozygous RTH-β with thyrotoxic cardiomyopathy; large goiter with compression; coexisting Graves' / toxic nodule overlay).
• Beta-blocker for tremor, palpitations, AF.
• TRIAC may control symptoms and lower FT4 — only at expert centres; twice or thrice daily; cross-reacts with T3 immunoassays (use LC-MS/MS).
• Surveillance: thyroid US for nodules (TIRADS), atrial fibrillation (chemical / electrical cardioversion often fail), MASLD (metabolic dysfunction-associated steatotic liver disease, formerly NAFLD) with FibroScan, bone density (low BMD common), lipid + glucose + HbA1c at diagnosis.
• ADHD evaluation in children with formal neuropsych testing; consider educational support.
• Co-existent hypothyroidism (autoimmune or congenital, with raised TSH but normal-range TH) means UNDER-replacement — titrate LT4 to family- comparable FT4 or to high-normal TSH + high TH, monitor cardiac.
• Pregnancy rule: if maternal FT4 >150% of ULN, consider ATD during pregnancy to protect an UNAFFECTED fetus from SGA / LBW. Genetic diagnosis pre-conception ideally. Multidisciplinary obstetric + endocrine care.

Pediatric nodule prevalence is low (1-5%) but malignancy rate is 22-26% in children vs 5-10% in adults. Workup is more aggressive.

WHAT THE CANCERS ARE — "DIFFERENTIATED" is the organizing word:

• Differentiated thyroid carcinoma (DTC) = FOLLICULAR-cell-derived and still BEHAVES like a thyrocyte — TSH-responsive, makes thyroglobulin (Tg, the follow-up marker), takes up IODINE (so I-131 / RAI both images AND treats it). DTC = PTC + FTC + oncocytic/Hürthle-cell.
• PTC (papillary), ~90% of pediatric thyroid cancer — spreads by LYMPHATICS (neck nodes, lung); fusion-driven in kids (RET/PTC, NTRK, ALK); >98% survival despite advanced presentation.
• FTC (follicular), rare in kids — spreads HEMATOGENOUSLY (bone/lung); RAS-driven; needs capsular/vascular invasion on histology (FNA can't call it — a "follicular neoplasm" is a surgical question).
• Oncocytic / Hürthle-cell — oxyphilic follicular variant, less iodine-avid, behaves like FTC.
• Medullary thyroid carcinoma (MTC) is NOT differentiated — it is a PARAFOLLICULAR C-cell tumor, marked by CALCITONIN + CEA (NOT Tg) and NOT iodine-avid (RAI useless). Think MEN2 / germline RET. Deep-dive below.
• Poorly-differentiated / anaplastic = DEDIFFERENTIATED (lost Tg + iodine handling); essentially not seen in children.

So "differentiated" literally means the tumor still acts like a thyrocyte — which is exactly WHY Tg is a marker and RAI is therapy in DTC but in neither MTC nor anaplastic disease. Size alone is NOT the fine-needle aspiration cytology (FNAC) threshold — any US-suspicious nodule, any solid >1 cm, and small lesions in high-risk context (prior XRT, syndromic [DICER1, PTEN, APC, MEN2/RET, PHTS, Carney], strong family history) get FNAC.

The neck-US read answers two questions — does the NODULE itself look malignant, and is any cervical NODE involved — and each has a reassuring-vs-suspicious split (ATA 2015 ped + ETA 2022 ped). The NODULE:

And the cervical NODE — the metastatic look upgrades staging and earns its own biopsy:

SIZE alone does NOT discriminate — a reactive node can be big and a metastatic one small; read the architecture, not the centimeters. The thyroid drainage beds to scrutinize are central (level VI) and lateral (levels III-IV).

Quantified, the LYMPH-NODE features are the high-SPECIFICITY rule-in signs (ETA 2022, sens / spec) — a suspicious node earns its own FNB + thyroglobulin-washout:

ETA 2022 PEDIATRIC RISK STRATIFICATION uses ultrasound pattern + size to triage FNAC:

• Highly suspicious US pattern (microcalcifications + hypoechoic + irregular margins + taller-than-wide): FNAC any size if technically feasible.
• Intermediate suspicion (1-2 suspicious features): FNAC at ≥1 cm.
• Low suspicion (isoechoic / hyperechoic, smooth margins, no suspicious features): FNAC at ≥1.5 cm or surveillance.
• Pure cyst or spongiform: no FNAC unless symptomatic.

Every nodule still starts at TSH: a SUPPRESSED TSH routes to scintigraphy — a hot, autonomous nodule is almost never malignant, so treat the autonomy and skip the FNA — while a NORMAL / HIGH TSH means stratify the ultrasound and let the pattern + size set the FNAC threshold.

PATIENT-LEVEL HIGH-RISK CONTEXTS that lower the FNAC threshold:

• Prior neck irradiation (RT to head/neck/chest before age 16, including craniospinal RT in childhood ALL/medulloblastoma).
• Syndromic risk: DICER1 (multinodular goiter + Sertoli-Leydig + PPB), PTEN hamartoma (Cowden), APC (FAP-associated cribriform-morular PTC), MEN2/RET (medullary), Carney complex (PRKAR1A), Werner.
• Family history of pediatric DTC or medullary thyroid carcinoma (MTC).
• Iodine deficiency (less common in iodized-salt countries but still relevant in some populations).

WORRISOME CLINICAL FEATURES on history/exam: rapid growth, hard fixed nodule, hoarseness (recurrent laryngeal nerve involvement), neck mass with palpable lymphadenopathy, history of head/neck radiation, family history of MTC.

FNA-CALCITONIN WASHOUT for the MTC question: C-cells exfoliate poorly, so plain cytology UNDER-CALLS MTC — rinse the FNA needle and assay calcitonin in the washout. A high washout confirms C-cell / MTC tissue in an indeterminate nodule or a suspicious node, and is more sensitive than cytology for nodal staging in known MTC.

BETHESDA (TBSRTC) applied to pediatric FNAC. Malignancy risk is shifted UP in each category vs adults:

• I (non-diagnostic): repeat FNAC; lobectomy if persistent.
• II (benign): ~6% ped (vs ~3% adult). US surveillance.
• III (AUS / FLUS - atypia of undetermined significance): ~28% ped. Repeat FNAC, molecular panel, often lobectomy.
• IV (follicular neoplasm / suspicious for FN): ~50% ped. LOBECTOMY for histology (capsular and vascular invasion).
• V (suspicious for malignancy): ~80% ped. Total thyroidectomy.
• VI (malignant): ~98% ped. Total thyroidectomy +/- central neck dissection.

THE AUS (BETHESDA III) PATHWAY — the indeterminate result you actually sweat (ATA 2015 / ETA 2022). Pediatric AUS runs ~28% malignant (vs 5-15% in adults), so the adult instinct — "molecular-rule-out and watch" — does NOT transfer.

• FIRST AUS: repeat the FNB, but NOT right away — wait 3-6 months (ATA ≥3 mo, ETA ~6 mo), because reparative atypia after a fresh aspiration manufactures false-positive atypia if you re-stick too soon.
• DOUBLE AUS (the repeat is STILL indeterminate): this is the decision point — take it to the MDT and move to DIAGNOSTIC SURGERY. Diagnostic hemithyroidectomy (lobectomy + isthmusectomy) is the default; ATA explicitly FAVORS surgery over yet another aspiration for persistent indeterminate cytology, because the pre-test malignancy rate is too high to keep watching.
• BRAF V600E / molecular — a RULE-IN, never a rule-out, tool in children. A POSITIVE BRAF V600E on the FNB is essentially never seen in a benign nodule -> high probability of PTC (still confirm cytologically or by intra-operative frozen before committing to a TOTAL thyroidectomy). But a NEGATIVE panel does NOT clear the nodule, because pediatric PTC is FUSION-driven — RET/PTC, NTRK, and ALK fusions dominate (all three targetable — the ALK one matters for the TKI list below) and BRAF V600E is far less common than in adults (rare in the youngest). A BRAF-only test therefore misses most pediatric cancers; broader fusion panels (RET/PTC, NTRK, ALK) — plus, for the FOLLICULAR-pattern indeterminate categories (Bethesda IV), RAS and PAX8-PPARγ — may add yield in Bethesda 3-5 but are not yet standard of care (research setting per ETA 2022). The adult Afirma-style "negative molecular -> avoid surgery" logic is unsafe in a child.

Pediatric DTC is >90% papillary, frequently bilateral/multifocal, so total thyroidectomy is the default; lobectomy only in low-risk unilateral small lesions.

THE 2025 ATA SHIFT IS ADULT-ONLY — the adult guideline de-escalated hard (active surveillance for cT1a PTC, lobectomy first-line up to 2-4 cm, no prophylactic central-neck dissection in cN0, surveillance stopped after a sustained excellent response). In CHILDREN none of this applies: pediatric PTC is multifocal and node-positive with lung-metastasis potential and decades of recurrence risk, so TOTAL thyroidectomy stays the default, molecular stays rule-IN only, central-neck dissection is used more liberally, and surveillance is LIFELONG. Peds DTC runs on the ATA 2015 pediatric + ETA 2022 guidelines; the 2025 ATA is explicitly adult-only (it even dropped nodules).

ATA 2015 pediatric DTC follow-up uses DYNAMIC RISK. The response category achieved in the first 1-2 years post-treatment sets the TSH suppression target and surveillance intensity:

• Excellent (Tg <0.2 suppressed or <1 stimulated, AntiTg negative, imaging clean) -> TSH 0.5-1.0, annual US.
• Indeterminate -> TSH 0.1-0.5, q6-12 mo.
• Incomplete biochemical (Tg rising, AntiTg rising, imaging negative) -> TSH <0.1, q6 mo, possibly empiric RAI.
• Incomplete structural -> TSH <0.1, surgery / RAI / EBRT / TKI (lenvatinib for iodine-refractory; selpercatinib OR pralsetinib for RET-positive).

Pediatric considerations: protect growth and bone — TSH suppression at the minimum needed. AntiTg+ makes Tg unreliable; use the AntiTg trend.

NOW THE NON-DIFFERENTIATED ONE — MTC is a different disease (C-cell, not follicular), handled separately from all the PTC/FTC above:

MEDULLARY THYROID CARCINOMA — RET TIERS + POST-OP MARKERS (ATA 2015). ~75% sporadic (somatic RET/RAS), ~25% germline RET (MEN2A, MEN2B, FMTC). The codon sets the prophylactic-thyroidectomy clock:

POST-OP, the marker game (NOT Tg — MTC makes calcitonin + CEA): draw both ~3 months out, then follow by level —

• undetectable / normal -> exam + neck US + markers q6mo x1yr, then yearly.
• detectable but <150 pg/mL, no structural disease -> calcitonin + CEA q3-6mo to compute the DOUBLING TIME; neck US.
• calcitonin >150 pg/mL -> IMAGE for mets (neck US, chest CT, liver MRI, bone). Positive -> resect / EBRT / systemic; negative -> re-image q6-12mo.

CALCITONIN / CEA DOUBLING TIME is the prognostic engine: <6 months = POOR (~25% 5-yr survival), 6 months-2 years = intermediate, >2 years = good (near-normal survival). The SLOPE beats the absolute number.

SYSTEMIC therapy for progressive metastatic MTC: multikinase inhibitors vandetanib + cabozantinib; the RET-SELECTIVE selpercatinib / pralsetinib are now preferred when a RET mutation drives it (better tolerated). RAI does nothing in MTC.

Amiodarone is ~37% IODINE by weight and intensely lipophilic, with a half-life near 100 days — so a single drug delivers a massive, LONG-LASTING iodine load, and its thyroid effects can appear late and persist for months after it is stopped. It also blocks T4->T3 conversion (D1 inhibition) and T3-receptor binding, and is directly toxic to thyrocytes. One drug, OPPOSITE diseases — and it slots into the thyrotoxicosis differential of §30.

AMIODARONE-INDUCED THYROTOXICOSIS (AIT) — two types, and color-flow Doppler is the bedside discriminator:

• MIXED / indeterminate forms are common — when you cannot cleanly split them, treat with BOTH (glucocorticoid + thionamide).

AMIODARONE-INDUCED HYPOTHYROIDISM (AIH):

• Mechanism: the iodine load triggers the Wolff-Chaikoff effect, and a gland that CANNOT ESCAPE it — typically an autoimmune / Hashimoto background (TPO-antibody positive) — stays blocked and goes hypothyroid. More common in iodine-replete regions.
• Treatment: levothyroxine. You do NOT have to stop the amiodarone — replace the thyroid and keep the antiarrhythmic.

PEDIATRIC NOTE: amiodarone is used for refractory arrhythmias, including after congenital-heart surgery — check TFTs at baseline and periodically, because both AIT and AIH are reported in children on it.

WHO TO REFER a short child to peds-endo (when to start testing):

• Under 3 yr (birth weight ≥2500 g): height SDS <-3, or <-2.5 on repeated measures.
• 3 yr or older: height SDS <-2.5; OR height SDS <-2 PLUS any of — SGA without catch-up (§41), psychosocial deprivation, disproportion / dysmorphism, >1.6 SDS below the target-height range, or a fall of ≥1 SDS in height.

WHEN TO SUSPECT GH DEFICIENCY (auxology, after excluding systemic / skeletal / syndromic causes and hypothyroidism) — GHD is an AUXOLOGICAL-plus-biochemical call: the GROWTH PATTERN, not a single GH-stimulation number, selects who gets provocative testing + insulin-like growth factor (IGF-1) + pituitary MRI.

• Severe short stature (height <-3 SDS); OR
• Moderate short stature (-2 to -3 SDS) WITH growth-velocity SDS <-1 over a year, or (after age 2) a fall in height SDS >0.5 over a year; OR
• Height >1.5 SDS (~8-10 cm) below target height; OR
• WITHOUT short stature: growth velocity SDS <-2 over one year, or <-1.5 averaged over two years.

(rhGH dosing, treatment-response evaluation, and IGF-1 monitoring: §39.)

GHRH -> GHRHR (somatotrope) -> GH pulses -> dimeric GHR on hepatocyte/ chondrocyte -> JAK2/STAT5B -> IGF1, IGFBP3, IGFALS. Most circulating IGF-1 is in a ternary complex with IGFBP-3 and ALS. PAPP-A2 cleaves IGFBP-3 to release free IGF-1 at the growth plate.

GH IS NOT JUST FOR HEIGHT — it runs a fed-vs-fasting metabolic program, half of it DIRECT (GHR-mediated) and half INDIRECT (IGF-1-mediated). The one-line summary: GH is LIPOLYTIC, ANABOLIC (for protein), and ANTI-INSULIN (diabetogenic) all at once (Çocuk Endokrinolojisi ve Diyabet, Cilt 1):

• CARBOHYDRATE: anti-insulin / DIABETOGENIC — lowers insulin sensitivity -> compensatory hyperinsulinemia. Mild + reversible in GHD/ISS kids, but watch Turner, SGA, PWS (baseline glucose-intolerance risk). Don't withhold rhGH from a diabetic — start low, expect higher insulin needs.
• LIPID: LIPOLYTIC (direct, via hormone-sensitive lipase) -> ↑free fatty acids, ketogenic, ↓visceral fat; rhGH lowers LDL, raises HDL.
• PROTEIN: ANABOLIC (via IGF-1 + mTOR) -> ↑protein synthesis, ↑lean body mass. The "lipolytic + anabolic" combo = less fat, more muscle.
• BONE: osteoanabolic -> ↑bone turnover + bone mineral density; also widens the growth plate (part of why rapid growth predisposes to SCFE).
• MUSCLE: ↑mass + strength (clearest in PWS — even strengthens respiratory muscles + central ventilatory drive).
• WATER / SALT: SODIUM + WATER RETENTION -> edema, carpal-tunnel, arthralgia (these are adult overdose signs; rare in children).
• ENDOCRINE CROSS-TALK (clinically load-bearing): GH ↑ peripheral T4->T3 conversion (can UNMASK central hypothyroidism) and ↑ cortisol->cortisone inactivation via 11β-HSD1 (can UNMASK central adrenal insufficiency) — recheck the thyroid + adrenal axes after starting/uptitrating rhGH.

The DIRECT/INDIRECT split maps to fed-vs-fasting: FED state (insulin present) -> JAK2-STAT5 + PI3K-AKT -> IGF-1 + protein synthesis, lipolysis OFF (anabolic). FASTING -> direct GHR-PLCγ-PKC -> lipolysis ON, anabolic pathways quiet (the starvation-adaptation face: GH up, IGF-1 down).

Diseases at each step:

• GH1 deletion / nonsense = IGHD IA — AR, GH UNDETECTABLE, severe from infancy. KEY POINT: the fetal thymus never sees GH peptide, so there is no central tolerance -> the first rhGH injection raises ANTI-GH antibodies -> neutralization -> initial response then plateau. Switch to mecasermin (rhIGF-1).
• GH1 splice-site OR GHRHR LoF = IGHD IB — AR, GH LOW but DETECTABLE, milder than IA, and NO antibody problem on rhGH (residual endogenous GH gave tolerance). GHRHR loss is the classic IB cause: the somatotrope cannot hear GHRH.
• GH1 dominant negative = IGHD II — AD. Exon-3 splice 17.5-kDa isoform dimerizes with wild-type GH intracellularly and gums up secretion; somatotrophs progressively die -> evolving multi-axis hypopituitarism.
• IGHD III — X-linked: BTK (with agammaglobulinemia) or SOX3.
• Bioinactive GH (Kowarski) — the trap that's not IGHD-II. GH1 missense (C53S, R77C); GH peptide is secreted in NORMAL or HIGH amount on immunoassay, but cannot dimerize GHR. IGF-1 is low. Treat with rhGH.
• GHR LoF = Laron. High GH, low IGF-1, growth failure -6 SDS untreated, hypoglycemia, microphallus. Mecasermin works, rhGH is futile.
• STAT5B LoF = Laron-like + immunodeficiency (eczema, lymphocytic interstitial pneumonia from broken TREG/IL-2R signaling).
• IGFALS LoF = mild SS, low total IGF-1, NORMAL free IGF-1.
• PAPP-A2 LoF = HIGH total IGF-1 (bound), LOW free IGF-1. The counterintuitive lab pattern. Treat with rhIGF-1, not rhGH (the somatotropes work fine; the BLOCKING STEP is at IGFBP-3 cleavage). NOT to confuse with PAPSS2 (steroid-sulfation enzyme, adrenal androgens + brachyolmia, see §13). Different gene, different pathway, different organ system, opposite-direction labs.
• IGF1 LoF = IUGR, microcephaly, deafness.
• IGF1R haploinsufficiency / mutation = SGA without catch-up PLUS the discriminator: MICROCEPHALY +/- developmental delay and a HIGH IGF-1 (the resistance pattern — normal/high GH, IGF-1 elevated because the receptor can't respond).

TESTING THE AXIS — auxology selects WHO gets tested; the tests themselves are imperfect:

• GH PROVOCATION (insulin-tolerance, glucagon, clonidine, arginine, L-dopa) is the traditional "GHD" confirmation but a POOR gold standard. The traps: PRIME peripubertal children with sex steroids first (the low-sex-steroid juvenile state gives a FALSE-POSITIVE "GHD"); the >7-10 ng/mL pass cutoff is ASSAY-dependent (not transferable between labs); OBESITY blunts the peak; and two separate tests are required because either alone is poorly reproducible. Auxology + IGF-1 + a pituitary MRI outweigh any single stimulated number.
• IGF-1 GENERATION TEST — the GH-RESISTANCE discriminator: give rhGH for ~4-7 days, then measure the IGF-1 rise. A blunted / absent rise = GH-resistance (Laron, post-receptor / STAT5B); a brisk rise = the axis can respond (points back to GH-deficiency). Use it when GH is HIGH and IGF-1 is LOW; reproducibility is limited, so read it against the phenotype.

IDIOPATHIC SHORT STATURE (ISS) — the diagnosis of EXCLUSION: height <-2 SDS with a NORMAL GH/IGF-1 axis, normal birth size (NOT SGA), and no dysmorphism, skeletal dysplasia, systemic illness, malnutrition, or endocrinopathy. It is a DESCRIPTION, not a mechanism — familial short stature and self-limited delayed puberty (SLDP) sit inside it, and a real fraction hides mild SHOX, ACAN, or NPR2 variants (§43), so the phenotype + a targeted gene panel keep peeling cases out of the "idiopathic" bucket. Workup = the standard short-stature screen (CBC, metabolic / renal / coeliac / thyroid, IGF-1 +/- IGFBP3, bone age, karyotype in girls) coming back normal. GH is approved for ISS (height roughly <-2.25 SDS) but the average gain is MODEST (~3-7 cm over years), dosed to auxology with an IGF-1 target ~+1 SDS (§39) — honest counseling about the small benefit is part of consent.

MONITORING CADENCE: follow in a peds-endo centre. Growth velocity (cm/yr) + height SDS every 3-6 months in year 1, then every 6-12 months. The best read on response is the change in height SDS over the FIRST YEAR; the 4-MONTH growth velocity is the strongest EARLY predictor of the first-2-year height-SDS gain, and the first-year response also predicts adult height. Annual left hand / wrist film for bone age; track puberty onset and tempo (puberty shortens the runway).

Insulin-like growth factor (IGF-1) — the adherence / efficacy / safety dial (check yearly, sooner after a dose change; it moves BEFORE growth velocity):

• LOW IGF-1 -> non-adherence, or another factor blunting the response.
• HIGH IGF-1 with POOR growth -> IGF-1 insensitivity.

LONG-ACTING (WEEKLY) GH (LAGH) is the main change in the field (Maniatis 2025 consensus): weekly SC dosing that, in 52-week phase-3 trials, gives NON-INFERIOR height velocity and the SAME safety as daily rhGH — the whole draw is adherence (52 shots a year, not 365). Three agents, three DIFFERENT half-life tricks:

• LONAPEGSOMATROPIN — a TransCon PRODRUG: UNMODIFIED somatropin transiently linked to an inert mPEG carrier; at body pH / temperature the linker hydrolyses and releases NATIVE GH across the week, so what reaches the receptor is ordinary somatropin (half-life ~25 h vs ~3 h).
• SOMATROGON — a FUSION PROTEIN: rhGH carrying three copies of the hCG beta-subunit C-TERMINAL PEPTIDE (CTP); the added size + glycosylation (~47 kDa) slows clearance. A bigger, modified molecule.
• SOMAPACITAN — an ALBUMIN-BINDING derivative: a side chain reversibly grabs endogenous albumin to dodge clearance (the same trick as long-acting insulin / GLP-1).

THE MONITORING CATCH: on weekly dosing IGF-1 SWINGS across the interval (peak ~day 2, trough ~day 7), so a RANDOM IGF-1 is uninterpretable — you must TIME the draw to the last injection. A sample ~DAY 4 post-dose (somatrogon, somapacitan) or ~day 4.5 (lonapegsomatropin) reads the AVERAGE IGF-1 (no adjustment needed); ~day 2 reads the PEAK. Always record the day-and-time of the draw relative to the shot, or the IGF-1 SDS means nothing.

The safety profile is otherwise the SAME as daily rhGH — the shared signals (SCFE, intracranial hypertension, scoliosis, insulin sensitivity, the mortality / cancer question) are general to all rhGH and live in the SAFETY block below, NOT specific to the long-acting formulation.

TRANSITION — retesting off rhGH: at the end of childhood treatment, retest off rhGH for ≥1 month — but SKIP the retest (the GHD is permanent) if any of: ≥3 pituitary hormone deficiencies, IGF-1 <-2 SDS, a causal genetic defect, or a structural hypothalamic-pituitary lesion.

• A 20% GH dose change ~= 1 SDS of IGF-1. Dosing to IGF-1 lets you use lower doses.
• Target by indication: GHD ~= 0 SDS; ISS ~= +1 SDS; REDUCE the dose if IGF-1 stays >+2.0 SDS without insensitivity. The PARTIAL-IGF-1-insensitivity states — SGA (§41), Silver-Russell, PWS, IGF1R defect, Turner — may NEED IGF-1 above +2 SDS to grow, so individualise to the auxology, not the number. No IGF-1 ceiling has been tied to harm in children, but keep it at or below the mean in cancer-predisposition / survivors (benefit unproven). IGFBP3, free IGF-1, ALS, and the IGF-1/IGFBP3 ratio add nothing to monitoring.

UNMASKING ON GH — the high-yield trap in multiple pituitary hormone deficiency. GH accelerates two peripheral conversions:

• cortisol -> cortisone (11β-HSD) — can expose a compensated ACTH deficiency (adrenal-crisis risk).
• T4 -> T3 — can expose central hypothyroidism.

So re-screen the adrenal and thyroid axes after starting GH and replace as needed (L-thyroxine to a normal free T4; hydrocortisone ~7-10 mg/m²/day with stress dosing; induce puberty with sex steroids at the appropriate age).

WHEN TO STOP: continue toward final height — growth velocity <2 cm/yr, or bone age 14 (girls) / 16 (boys).

INADEQUATE RESPONSE (GHD) — any of: first 6-12 months Δgrowth velocity <2 cm/yr, growth-velocity SDS <-1, or Δheight SDS <0.3/yr (severe GHD <0.4/yr). Work the ladder: adherence, injection technique, dose, occult hypothyroidism, nutrition, puberty status, chronic disease — then re-question the diagnosis. In GH1 deletion, check neutralising anti-GH antibodies. If the diagnosis holds, adherence is good, and IGF-1 sits below target -> raise the dose 25-50%; if a suboptimal response persists, STOP. Use etiology- and sex-specific response charts; use disease-specific growth curves for syndromes; spinal-irradiation survivors grow disproportionately and respond less.

SAFETY (>30 years of use; complications rare, <3%, but monitor long-term — even after stopping):

The cancer / mortality signal, kept honest: in children GH does NOT raise new-primary-tumour risk or recurrence; cancer survivors carry a small secondary-tumour excess that is highest early and falls with time, with a bone-tumour and (in irradiated patients) meningioma signal. Noonan / PTPN11 carries an intrinsic ~3.5x cancer risk — assess tumour predisposition before starting. A mortality signal (all-cause, bone-tumour, cardiovascular / hemorrhagic) was raised at very high doses (>50 μg/kg/day) but has NOT been confirmed at standard pediatric doses, and there is NO excess in low-risk groups (idiopathic GHD, ISS) — where a signal does appear it tracks the underlying high-risk condition, not the GH. So it is a dose-ceiling caution, not a reason to withhold replacement.

Bone age (BA, skeletal maturity) is the body's biological clock: how far ossification has progressed, read off a single LEFT hand-wrist radiograph. It is a better predictor of REMAINING growth, pubertal tempo, and adult height than the calendar. The whole game is BA versus chronological age (CA) — the gap, and its direction, is the diagnostic signal that recurs across this book.

THREE METHODS:

THE OSSIFICATION TIMELINE — what appears, and when (females run ~1-2 yr AHEAD of males throughout; ages approximate):

TWO QUICK BEDSIDE RULES off that table:

• Carpal count — from ~1 to 7 years the NUMBER of ossified carpal centers roughly equals the age in years. A fast first-pass sanity check on a young child's film.
• Sesamoid sign — the thumb adductor sesamoid lighting up means the pubertal spurt is UNDERWAY (Tanner ~3-4 / near PHV). It marks tempo, not the switch-on of puberty — present in essentially all girls by ~13-14 and boys by ~15.

INTERPRETATION — the BA-vs-CA gap and its differential:

A delayed BA is a hopeful number in a short child — growth potential is BANKED, and height prediction off the delayed BA usually beats the gloomy mid-parental estimate. An advanced BA in a short child is the opposite warning: the window is closing early (the ACAN trap, and the reason untreated CPP costs adult height).

HEIGHT PREDICTION rides on BA: Bayley-Pinneau (off the GP bone age) is the classic "% of adult height already achieved" method; Tanner-Whitehouse and Khamis-Roche are alternatives. All are estimates with wide error bands — track the TREND, do not quote a single centimeter as destiny.

DEFINITION: birth weight and/or length <-2 SDS for gestational age (needs accurate dating + actual birth measurements). SGA is a SIZE-AT-BIRTH label, not a mechanism — distinct from IUGR (which is a process).

CATCH-UP is the fork. ~85-90% of SGA infants catch up to height >-2 SDS by age 2 (preterm SGA by ~4). The ~10% who DO NOT are the persistent-short-stature group — and the GH-treatment population.

GH INDICATION (the persistent-SGA short child): NO catch-up by age 2-4 with height roughly <-2.5 SDS (EMA: from age 4; US: from age 2 at a less-strict cutoff). Doses run HIGHER than classic GHD (~35-70 μg/kg/day). The first-year response predicts; treat to a normalised height, then toward target. SGA behaves as a PARTIAL insulin-like growth factor (IGF-1)-insensitivity state — IGF-1 often runs high and may need to exceed +2 SDS for effective growth — so dose to the auxology, not to the IGF-1 number (§39).

WORK IT UP before settling on idiopathic SGA: confirm TRUE SGA, then hunt a syndrome when there is dysmorphism, body disproportion, or failure to catch up — Silver-Russell (11p15 ICR1 hypomethylation or maternal UPD7; relative macrocephaly, body asymmetry, feeding difficulty, scored on the Netchine-Harbison criteria; §46), 3-M, IMAGe, Bloom, and others. Use disease-specific growth curves where they exist.

THE METABOLIC TAIL (why SGA matters beyond height): SGA — and especially RAPID postnatal weight catch-up — programs later INSULIN RESISTANCE, type 2 diabetes, dyslipidemia, central adiposity, and hypertension. Add premature adrenarche, earlier / faster puberty (which further shortens the time to grow), and PCOS in girls. GH itself transiently lowers insulin sensitivity, so screen glucose in the at-risk. The international consensus (Hokken-Koelega 2023) frames management as a continuum — infancy (feeding, neurodevelopment) through puberty to the adult metabolic and transition picture.

SHOX sits in pseudoautosomal region 1 (PAR1) of Xp and Yp. It escapes X-inactivation, so dosage matters: everyone normally has two functional copies.

LERI-WEILL DYSCHONDROSTEOSIS = heterozygous SHOX loss (deletions most common, point mutations less so). One functional copy. Inheritance looks autosomal dominant (technically pseudoautosomal). Mesomelic short stature — the middle segments of the limbs (forearm, lower leg) are disproportionately short. MADELUNG DEFORMITY is the wrist sign: bowing of the distal radius with dorsal subluxation of the distal ulna, V-shaped wrist articulation on X-ray, limited wrist extension and supination. More penetrant in females, often emerging at puberty under estrogen. Variable expression — some heterozygotes are asymptomatic with mild short stature, others have full Madelung. GH is FDA-approved for SHOX deficiency.

LANGER MESOMELIC DYSPLASIA = biallelic SHOX loss. Both copies gone. Profound mesomelic dwarfism, severe hypoplasia or aplasia of ulna and fibula, marked Madelung, severely short forearms and legs. Was once considered a separate disease until SHOX biology was unraveled — it's the double-dose effect of the same lesion.

The TURNER connection: Turner syndrome short stature is largely SHOX haploinsufficiency. Turner is missing one X, so missing one copy of PAR1 SHOX. That is why GH works in Turner.

The dose ladder: 2 copies = normal stature. 1 copy = Leri-Weill or SHOX-related isolated short stature. 0 copies = Langer.

These aren't generic "short stature" disorders, they're growth-plate- specific diseases. The phenotype is DISPROPORTIONATE — limbs out of proportion to trunk, or middle segments mismatched with proximal — and the radiographs give it away. Three genes own the axis: FGFR3 brakes chondrocyte proliferation, ACAN provides the structural matrix, NPR2 receives the C-type natriuretic peptide (CNP) growth-promoting signal.

THE LIMB-SEGMENT VOCABULARY — say WHERE the shortening is and you've half-localized the gene. A limb has three segments: PROXIMAL (rhizo-, "root" = femur / humerus), MIDDLE (meso- = forearm radius+ulna / lower leg tibia+fibula), and DISTAL (acro- = hands / feet). Which segment is short names the pattern:

Two axes layer on top of the segment term. SHORT-LIMB vs SHORT-TRUNK: limbs short against a near-normal trunk = the FGFR3 group; trunk short against relatively normal limbs = the spondylo- / COL2A1 group (e.g. spondyloepiphyseal dysplasia). The same radiograph also gets read for bowing, angulation, and Madelung deformity (the dorsal-radial wrist deformity of SHOX loss).

DYSPLASIA vs DYSOSTOSIS — the distinction the limb-segment words quietly assume. A dysplasia (osteochondrodysplasia) is a fault in the bone / cartilage TISSUE itself: GENERALIZED across the skeleton wherever that tissue is laid down, and because the abnormal tissue keeps growing, the phenotype EVOLVES and worsens with age. Achondroplasia — FGFR3 in every growth plate — is the prototype. A dysostosis is a fault in the PATTERNING of a SPECIFIC bone or group of bones, a blueprint error set during embryonic blastogenesis: LOCALIZED to the bones it mis-specifies and STATIC — the shape is fixed at birth and does not progress. So rhizo- / meso- / acromelia are DYSPLASIA vocabulary: they describe how a generalized tissue defect shortens whole long-bone segments. BRACHYDACTYLY is the other category — short DIGITS from a DYSOSTOSIS of named hand / foot bones, classified on its own (the Bell types), NOT a limb-segment term. (Brachydactyly can also ride along as a SIGN inside a dysplasia or syndrome — COMP, TRPV4, AHO, Turner — but there it is a feature of the larger disorder, not isolated Bell-type brachydactyly.)

THE BELL BRACHYDACTYLY TYPES — which bone is short names the type:

BRACHYDACTYLY TYPE E is the one that earns its keep in an endocrine text, because it is the skeletal readout of the PTHrP -> PTH1R -> Gsα -> cAMP -> PKA axis at the growth plate of the short tubular bones — the SAME cascade the iPPSD umbrella is named for (§24). PTHrP normally HOLDS chondrocytes in the proliferative pool and keeps them from differentiating too early; lose the signal anywhere along that cascade and the metacarpal / metatarsal growth plates differentiate and fuse prematurely, so those bones finish SHORT. One mechanism, one hand, several genes:

• ISOLATED BDE: PTHLH (the PTHrP ligand itself; BDE2) or HOXD13. These account for only a minority — most isolated BDE is still molecularly unsolved.
• AHO (PHP1A / PPHP): GNAS / Gsα — BDE as one piece of the full AHO phenotype, with multihormone resistance in PHP1A.
• 2q37 deletion / HDAC4 — the "AHO-like" phenocopy (brachydactyly-mental-retardation syndrome): same hand, often intellectual disability, but NORMAL Gsα activity, which is the lab line that separates it from true AHO.
• HYPERTENSION-WITH-BRACHYDACTYLY (HTNB / Bilginturan, PDE3A = iPPSD6) and the short 4th metacarpal of Turner (a BDE forme fruste).

The AHO bedside shortcut for this hand is the Archibald sign (§24): on making a fist, a DIMPLE replaces the knuckle over the short 4th metacarpal.

ACHONDROPLASIA (FGFR3 G380R, AD, fully penetrant) is the classic short-limbed dwarfism. ~1 in 25,000 births. ~80% are de novo with a paternal-origin G380R and the advanced-paternal-age effect.

WHY IT'S ALWAYS THE SAME VARIANT, AND WHY IT'S USUALLY NEW. Achondroplasia is the textbook example of a disease driven by ONE recurrent nucleotide change: ~99% of patients carry c.1138G>A (a minority c.1138G>C), both encoding the identical G380R. Position 1138 sits in a CpG dinucleotide, and methylated cytosine spontaneously deaminates to thymine — so this base is a mutational hotspot that gets hit over and over, which is why the mutation is recurrent rather than scattered across the gene. On top of that CpG baseline, the G380R change is ALSO a gain-of-function in the spermatogonial stem cells that carry it: the mutant clone proliferates faster and is positively selected in the aging testis, so mutant sperm are progressively ENRICHED with paternal age (selfish spermatogonial selection, Goriely & Wilkie). The two effects compound — a hypermutable CpG site PLUS a clonal-expansion advantage — which is why ~80% of cases are de novo and the new mutation is almost always paternal in origin. The practical upshot, state it plainly: ADVANCED PATERNAL AGE RAISES THE RISK OF ACHONDROPLASIA. An older father has a higher chance of an affected child (the risk climbs progressively, steepest past ~40), because his testis has accumulated more of these selfishly-expanded G380R spermatogonial clones — it is paternal, not maternal, age that matters, and the mechanism is the new mutation arising in sperm, not anything inherited. The same selfish-selection logic explains the paternal-age effect in Apert (FGFR2), MEN2 (RET), and Costello (HRAS).

The mechanistic twist: G380R is CONSTITUTIVELY ACTIVE FGFR3, but activation of FGFR3 at the growth plate SUPPRESSES chondrocyte proliferation. Gain-of-function means LESS growth here — opposite of the usual GoF intuition.

Phenotype: rhizomelic (proximal) limb shortening, macrocephaly, frontal bossing, midface hypoplasia, trident hand, lumbar lordosis. Adult height ~130 cm (♂) / ~124 cm (♀).

The complications you cannot miss: foramen magnum stenosis in infancy (brainstem compression, sudden death risk — screen with PSG + MRI + neurosurg referral), thoracolumbar kyphosis, spinal stenosis in adulthood, OSA, recurrent otitis media.

Vosoritide (FDA-approved to increase linear growth in achondroplasia with open epiphyses) is the established CNP-analog drug. It's a CNP analog that binds NPR-B (the NPR2 receptor), activates cGMP, and ANTAGONIZES FGFR3 downstream signaling. So a drug targeting NPR2 fixes an FGFR3 disease — the axis is interconnected. ~1.5 cm/yr extra growth velocity. Requires open growth plates. Navepegritide, a once-weekly CNP-analog prodrug, is also FDA-approved for achondroplasia with open epiphyses — same NPR-B target, weekly dosing instead of daily. Infigratinib (a selective FGFR1-3 TKI) is in late-phase trials. GH gives marginal ~3 cm gain and isn't standard of care.

Hypochondroplasia is achondroplasia's milder cousin (FGFR3 N540K most common). Less obvious short limbs, no macrocephaly, mild or absent midface hypoplasia. Adult height 130-150 cm. The kid with subtle disproportion and "idiopathic short stature" — sequence FGFR3.

Thanatophoric dysplasia (FGFR3 K650E, severe GoF) is the lethal neonatal end of the spectrum. Cloverleaf skull, very short limbs.

Pseudoachondroplasia (COMP, AD) — named for the resemblance but a COMPLETELY different disease from achondroplasia: different gene, different mechanism, different face. COMP (cartilage oligomeric matrix protein) mutations make a misfolded protein that jams the chondrocyte ER -> ER stress -> chondrocyte death. The discriminators that keep the two apart: the head and face are NORMAL (no macrocephaly, no frontal bossing, no midface hypoplasia), the child is normal-sized AT BIRTH and only falls off the curve around age 2 with a waddling gait + marked ligamentous laxity, and FGFR3 is normal. Otherwise disproportionate short-limbed stature, brachydactyly, early severe osteoarthritis, normal intelligence. The same gene also causes the milder multiple epiphyseal dysplasia. The one-line trap: similar NAME, but normal CRANIOFACIES + later onset + COMP-not-FGFR3.

ACAN-related short stature (ACAN haploinsufficiency, AD). Aggrecan is the major proteoglycan of growth-plate cartilage. Lose one copy and you get short stature -2 to -4 SDS with ADVANCED bone age (the paradox — GHD and CDGP both have DELAYED bone age), midface hypoplasia, brachydactyly, early-onset osteoarthritis, AD family pedigree where the whole family is short and ages early.

The recognition trick: short child with ADVANCED bone age = ACAN. GH partially helps. Pairing GH with a GnRHa to slow bone-age progression is being studied. Aromatase inhibitor in pubertal boys helps final height.

NPR2 (CNP receptor) sits on the same axis:

The FGFR3-ACAN-NPR2 triad: FGFR3 brakes, CNP via NPR2 promotes, aggrecan is the matrix. Different diseases hit the same axis from different angles. Vosoritide tipping NPR2 to overcome FGFR3 is the proof of concept.

Other skeletal dysplasia genes worth knowing as one-liners: COL2A1 (spondyloepiphyseal dysplasia, Stickler-related), COL10A1 (Schmid metaphyseal chondrodysplasia), NPPC / CNP itself (rare overgrowth), TRPV4 (SMD spectrum). (IHH / brachydactyly type A1 is a dysostosis, not a dysplasia — it lives with the Bell types above.)

3M syndrome (CUL7 / OBSL1 / CCDC8, AR) — a PRIMORDIAL dwarfism, NOT a GH-axis defect. Severe PRE- AND POSTNATAL growth restriction (adult height ~-5 to -6 SDS), NORMAL intelligence, relatively spared head size (looks macrocephalic), and a recognizable face (triangular, frontal bossing, midface hypoplasia, fleshy nose tip, full lips). Radiographs: TALL slender vertebral bodies + slender long bones + foreshortened long bones. The CUL7-OBSL1-CCDC8 complex regulates microtubule dynamics + insulin-like growth factor (IGF) signaling. The discriminator vs SRS: AR inheritance, NORMAL cognition, NORMAL 11p15 methylation, and the vertebral/long-bone signature. GH response is POOR — they are not GH-deficient. Send the 3-gene panel in unexplained severe SGA + postnatal short stature with normal 11p15.

The RASopathies are the family of disorders caused by gain-of-function mutations along the RAS-MAPK pathway. They share a core phenotype — short stature + characteristic face + congenital heart disease + variable ID + cancer risk — and the specific gene determines the flavor.

NOONAN SYNDROME (1 in 1000-2500 live births, AD) is the most common RASopathy:

• Short stature ~70%.
• Pulmonary VALVE stenosis — the most characteristic cardiac lesion. Plus hypertrophic cardiomyopathy (HCM) and septal defects.
• Face: low-set posteriorly rotated ears, ptosis, hypertelorism, downslanting palpebral fissures, deeply grooved philtrum, webbed / short neck.
• Cryptorchidism in boys.
• Bleeding diathesis (variable platelet dysfunction, factor XI deficiency).
• Lymphatic dysplasia (lymphedema, chylothorax).
• Cognition usually mild ID or normal IQ.
• Cancer: juvenile myelomonocytic leukemia (JMML), especially with PTPN11. Also ALL.

SCORING Noonan — the van der Burgt criteria (2007), still the clinical anchor when the gene panel is negative (a real ~20%). The TYPICAL FACE is the entry key; the rest is graded Major vs minor:

DIAGNOSE Noonan with: TYPICAL face + 1 Major (from categories 2-6) OR + 2 minor; or SUGGESTIVE face + 2 Major (categories 2-6) OR + 3 minor. Molecular testing (the RAS-MAPK panel) confirms ~80% and increasingly leads — the criteria matter most for the test-NEGATIVE child and for deciding WHO to send.

Gene by gene, the phenotype shifts subtly — this is the part to internalize because the gene affects treatment response and surveillance:

GH treatment: FDA-approved for Noonan. Response is suboptimal compared with GHD or Turner — ~1 SDS over years on average. PTPN11 patients respond less well than non-PTPN11. No clear evidence of increased malignancy on GH but hematology surveillance is sensible given the JMML/ALL background risk.

The other RASopathies share biology but differ in severity:

• Costello syndrome (HRAS GoF): more severe than Noonan. Prominent papillomata around mouth and nose, cardiac arrhythmias, characteristic coarse face, ID, malignancy risk (rhabdomyosarcoma, neuroblastoma), severe short stature.
• Cardio-facio-cutaneous (CFC) syndrome (BRAF, MAP2K1/2, KRAS): more severe ID, prominent ectodermal abnormalities (sparse hair, keratosis pilaris), cardiac disease, FTT.
• LEOPARD syndrome (now NOONAN SYNDROME WITH MULTIPLE LENTIGINES, specific PTPN11 mutations): Lentigines + EKG abnormalities + Ocular hypertelorism + Pulmonary stenosis + Abnormal genitalia + Retarded growth + Deafness.

NF1 also activates RAS-MAPK — there's clinical overlap with Noonan (Noonan-NF1 spectrum).

The unifying biology: GoF anywhere in RAS-MAPK gives a RASopathy. The phenotype reads out when and where the pathway is dysregulated developmentally. Treatment principles share across the family — monitor for HCM, gene-specific cancer surveillance, GH if short stature warrants.

CONSTITUTIONAL ADVANCE OF GROWTH (CAG) — the MIRROR IMAGE of SLDP (formerly CDGP), and the place to start when a tall child shows up without any other pathology marker. From Corredor / Bozzola 2019 (Curr Pediatr Rev) — the operational definition:

• ACCELERATED growth velocity after birth.
• Peak centile reached by 2-4 years of age.
• Child grows ALONG the 97th centile until ~9 years.
• Growth rate then drops to the 50th centile.
• Pubertal entry tends to be on the EARLIER side of normal.
• Final adult height NORMAL for the parents (bone-age advance allows earlier epiphyseal closure — the kid "uses up" their growth window earlier rather than growing taller than expected from mid-parental height, MPH).

How CAG differs from FAMILIAL TALL STATURE (FTS) — this is the clinical discrimination question on the tall-child consult:

So the question to ask the parents: "Was this baby longer than average at birth, and was the family height itself just normal?" Yes to both = CAG. Tall family + normal birth length + tracking with MPH = FTS.

PROPOSED MECHANISM for CAG: increased IGF-II secretion and elevated IGF / IGFBP molar ratio prepuberty. The increase in bioavailable insulin-like growth factor (IGF) activity drives the early growth acceleration AND the bone-age advance, which is why the kid ends up at a normal MPH-aligned adult height — the growth potential is just compressed into an earlier window. The obesity link may also operate via this IGF / IGFBP shift (paracrine nutrient signaling).

CLINICAL DECISION POINTS for CAG:

• Reassurance is the main move. Parents need to understand the kid will grow more SLOWLY in mid-childhood — this is normal.
• No GH workup unless growth velocity actually accelerates BEYOND the 97th-centile track or there are dysmorphic / syndromic features.
• Track for OBESITY signals in mid-childhood, since CAG may flag a predisposition.
• Pubertal timing should be monitored but rarely needs intervention — earlier-end-of-normal puberty in CAG is not precocious puberty.
• If puberty is genuinely precocious (CPP criteria), THAT is a different diagnosis and needs workup (see §47).

CAG sits in the differential before you reach for the aromatase / overgrowth / endocrine causes below. It is the "mirror of SLDP" — self-limited, family-history-friendly, normal final outcome.

Connective: Marfan (FBN1, upward lens dislocation), Loeys-Dietz (TGFBR, bifid uvula, hypertelorism, aggressive aortic disease, no ectopia lentis), homocystinuria (CBS, DOWNWARD lens dislocation, thromboembolism, some pyridoxine-responsive).

DIAGNOSING Marfan — the revised Ghent nosology (2010). Two CARDINAL features carry it: aortic root dilatation / aneurysm (Z-score ≥2) and ectopia lentis (UPWARD lens dislocation). The pathway forks on family history:

• NO family history — any ONE of: aortic root Z≥2 + ectopia lentis; OR aortic root Z≥2 + a causative FBN1 variant; OR aortic root Z≥2 + systemic score ≥7/20; OR ectopia lentis + an FBN1 variant already known to cause aortic disease.
• FAMILY history of Marfan — any ONE of: ectopia lentis; OR systemic score ≥7; OR aortic root Z≥2 (age ≥20) / Z≥3 (age <20).

The systemic score (≥7 of 20) tallies the soft signs — wrist + thumb signs, pectus, hindfoot deformity / pes planus, reduced upper:lower-segment ratio + increased arm span, scoliosis, reduced elbow extension, the face (3 of 5), skin striae, myopia >3D, mitral-valve prolapse, dural ectasia, protrusio acetabuli, pneumothorax. The 2010 revision deliberately WEIGHTS the aorta + lens, DEMOTES the soft signs, and pushes FBN1 testing to the centre. KEY PEDIATRIC CATCH: a child can be too young to have declared aortic dilation or ectopia lentis yet, so a kid who falls short is "potential / emerging MFS" — re-image the root and re-examine on a schedule, don't clear them. Discriminators sit in the line above: Loeys-Dietz has the aggressive aorta but NO ectopia lentis (plus bifid uvula, arterial tortuosity); homocystinuria dislocates the lens DOWNWARD (plus thromboembolism, ID).

Overgrowth: Sotos (NSD1), Weaver (EZH2), PRC2 family, Simpson-Golabi-Behmel (GPC3 LoF, X-linked males — pre- and post-natal overgrowth + macroglossia / organomegaly + Wilms risk; glypican-3 normally restrains IGF2, so it's the mechanistic mirror of BWS).

Imprinting: Beckwith-Wiedemann (11p15, IGF2 overexpression, CDKN1C loss).

Endocrine: GH excess (X-LAG/GPR101 microduplication, AIP, McCune- Albright); aromatase deficiency.

Aromatase deficiency (CYP19A1 LoF) is the trap. In a 46,XY boy the testis makes testosterone normally, virilizes the fetus normally, drives normal puberty. ESTROGEN closes the growth plate. Without aromatase, T accumulates, estrogen cannot be made, growth plate stays open into the 20s/30s. Adult height drifts up 1-2 cm/year. Osteopenia and insulin resistance follow. Treatment is transdermal estradiol. So 46,XY aromatase deficiency is TALL with delayed fusion — NOT undervirilized.

Sex chromosome: Klinefelter 47,XXY (an extra SHOX copy [3, not a triplication], but the dominant driver is hypogonadism removes brake on epiphyseal closure -> tall, eunuchoid, small firm testes < 4 mL in Tanner 3-4 with delayed puberty, gynecomastia 30-50%, azoospermia in classical KS); 47,XYY modest tall.

Monogenic obesity with tall stature: MC4R LoF (tall plus hyperphagia).

Beckwith-Wiedemann (BWS) is the COMMONEST overgrowth syndrome. Same locus, opposite phenotypes: BWS and Silver-Russell are the cleanest example in clinical genetics of BIDIRECTIONAL IMPRINTING — the SAME chromosomal region (11p15.5) can be tipped one way to drive overgrowth or the OTHER way to drive growth failure.

THE LOCUS. 11p15.5 holds two adjacent imprinted clusters with opposite imprinting directions:

• IC1 (telomeric, H19/IGF2 DMR). Paternal allele methylated -> H19 silenced + IGF2 expressed. Maternal allele unmethylated -> H19 expressed + IGF2 silenced. Net normal = ONE active IGF2 (paternal).
• IC2 (centromeric, KCNQ1OT1/CDKN1C DMR). Maternal allele methylated -> KCNQ1OT1 silenced + CDKN1C expressed (growth brake). Paternal allele unmethylated -> KCNQ1OT1 expressed + CDKN1C silenced. Net normal = ONE active CDKN1C (maternal).

So the locus runs ONE pro-growth signal (IGF2) and ONE anti-growth brake (CDKN1C), each from a single parental allele. Disturb the balance in EITHER direction and you get a phenotype.

SCORING SRS — the Netchine-Harbison Clinical Scoring System (NH-CSS, 2016/2017 international consensus). SRS is a CLINICAL diagnosis, made at ≥4 of these 6:

≥4/6 = clinical SRS; send (epi)genetics at ≥4/6 (or ≥3/6 with strong suspicion). THE HEAD RULE: if a child scores 4/6 but LACKS both relative macrocephaly AND the prominent forehead, doubt SRS and hunt another cause — those two head features carry the diagnostic weight. Molecular confirmation lands in only ~60% (11p15 LOM ~50%, maternal UPD7 ~10%), so ~40% of genuine SRS is test-NEGATIVE and stays a clinical call on the NH-CSS.

The deeper lesson: imprinting disorders break the "one gene = one disease" rule because the SAME variant has opposite phenotypes depending on which parent transmits, and the SAME phenotype can come from MULTIPLE mechanisms (CNV, UPD, methylation defect, coding mutation) acting on the SAME pathway. DICER1, MEN1, PHP1A/PPHP and Carney complex all hint at this — the imprinted-locus stories in particular (PWS/AS at 15q11, BWS/SRS at 11p15, PHP1A/PPHP at GNAS) are the cleanest mirrors.

Turner (45,X 40-50% pure, mosaics common):

• Short stature from SHOX haploinsufficiency, ~143 cm untreated.
• Lymphedema embryology explains webbed neck (fetal cystic hygroma compresses aortic arch -> coarctation).
• Cardiac: BAV (15-30%), coarctation (10-12%), aortic dilation/dissection LIFELONG. Echo at dx, CMR at 9-11 then q5-10y.
• Streak gonads. Spontaneous puberty in ~30% of mosaics, sustained menses ~5-10%.
• Renal anomalies ~30%.
• Hashimoto, celiac, IBD.
• Hormone treatment (GH, estrogen induction, fertility preservation) — full timeline at the end of this section.
• 45,X/46,XY mosaic with female phenotype: gonadectomy for gonadoblastoma risk.

The "45,X" label hides a lot. Karyotype distribution drives phenotype and surveillance intensity — count at least 30 metaphases or you'll miss mosaicism.

The take-home: karyotype tells you who keeps an ovary, who needs gonadectomy for gonadoblastoma risk, and how aggressively to screen for the rest.

Surveillance baseline regardless of variant: cardiac echo at diagnosis in all, CMR at 9-11 yr then q5-10 yr, renal US at diagnosis (horseshoe, duplex, malrotation in 30%), audiology q2-3 yr, TSH yearly from age 4, celiac q2-5 yr from age 2, DXA + HbA1c + lipid panel periodic from adolescence.

Pregnancy in Turner is the one to flag. Aortic dissection during pregnancy is a recognized cause of maternal mortality — pre-pregnancy CMR within 2 years, intensive cardiology comanagement throughout pregnancy. Spontaneous pregnancy is rare but possible in mosaics; donor-oocyte ART covers permanent gonadal failure.

Y-MOSAICISM IN TURNER — the gonadoblastoma trap. ~5% of clinical Turner patients carry hidden Y-chromosome material (45,X/46,XY mosaicism, marker chromosomes, structurally abnormal Y). The risk locus is the GBY (gonadoblastoma on Y) region around TSPY1 (Yp11.2) — testis-specific protein on Y, when expressed in a streak/dysgenetic gonad it drives gonadoblastoma. POU5F1 (OCT4) expression in dysgenetic gonad cells is the histologic marker for in-situ gonadoblastoma / germ cell neoplasia. Operational: do FISH for SRY/centromeric Y probes on any Turner karyotype with virilizing signs, ambiguity, OR persistent mosaicism. If Y material is present -> PROPHYLACTIC GONADECTOMY by adolescence (gonadoblastoma risk 15-30% over lifetime).

TURNER — THE HORMONE TREATMENT TIMELINE (Gravholt 2024):

GROWTH (rhGH):

• START as soon as growth falters / height drops across centiles — early is better (often by age 4-6, can be earlier). Dose 45-50 μg/kg/day (1.3-1.5 mg/m²/day) — HIGHER than classic GHD, because Turner is a partial GH-resistance state.
• OXANDROLONE add-on for the poor-prognosis / late-diagnosed very short girl: from ~age 10, 0.03 mg/kg/day (max 0.05) — watch HDL drop and virilization.
• STOP when growth potential is spent: bone age >14 with growth velocity <2 cm/yr, or once a satisfactory height is reached.

ESTROGEN (puberty induction — do NOT sacrifice pubertal timing for height, the old delay-for-height practice is abandoned):

• Track LH / FSH / AMH from 8-9 yr, yearly. FSH elevated on ≥2 occasions = the ovary has failed and she needs induction.
• START low-dose transdermal 17β-estradiol at 11-12 yr (physiologic timing), CONCURRENT with GH. Transdermal preferred (oral second; ethinyl estradiol last resort).
• TITRATE slowly to adult dose over 2-4 years (adult target E2 ~100-150 pg/mL).
• ADD cyclic progesterone once breakthrough bleeding appears (usually after ~18-24 months of unopposed estrogen): micronized progesterone 200 mg x 10-12 days/month.
• CONTINUE cyclic estrogen + progesterone to the usual menopausal age (~50), then re-evaluate.
• LATE diagnosis (>12 yr) with remaining growth potential: start low-dose E2 SIMULTANEOUSLY with GH — don't wait.

FERTILITY PRESERVATION (Nawroth 2020 — genotype + ovarian reserve (AMH) + cardiac fitness decide; AMH ~0.5 ng/mL is the hinge):

Klinefelter (47,XXY, 1 in 450-600 boys, 50-75% undiagnosed) — course and care per the EAA guideline (Zitzmann 2021):

• Tall, eunuchoid (arm span > height, long lower segment).
• Testes <4 mL in Tanner 3-4 with delayed puberty (this is the developmental cutoff, not a flat <6 mL rule).
• Primary testicular failure: high FSH/LH, low T, azoospermia.
• Karyotype is a DOSE gradient: each extra X worsens it (48,XXXY, 49,XXXXY, 48,XXYY -> ID, malformations, deeper hypogonadism); 47,XXY/46,XY mosaics are milder and occasionally fertile. Confirm prenatal/screening calls on a postnatal karyotype before counselling.
• Gynecomastia 30-50%, low T-to-E2 ratio.
• Cognitive: ~10 IQ points below sibs, verbal > performance affected, ADHD and autism-spectrum. Speech therapy + learning support early.
• Metabolic: visceral adiposity, IR, T2DM, MetS — INTRINSIC (hypogonadal body composition + X-dosage), so plain central adiposity is just KS and the lever is testosterone (when deficient) + lifestyle, not melanocortin drugs. But KS is NOT a hyperphagic syndrome — severe early-onset hyperphagic obesity is out of proportion -> think DUAL etiology (a second, melanocortin / hypothalamic cause), not the karyotype.
• Cardiovascular SMR 1.3; the thromboembolic risk is real (DVT/PE) — give thromboprophylaxis before long-haul flights / immobilization. Baseline 12-lead ECG (KS runs a reduced QTc).
• Mediastinal extragonadal GCT: karyotype before you biopsy.
• Male breast cancer SIR 4-5 — examine the breasts.
• TESTOSTERONE: none in infancy/childhood, and NOT routinely in adolescence just for the karyotype — treat only DELAYED PUBERTY or symptomatic + biochemically confirmed hypogonadism. Then yearly endocrine review, with DXA + vitamin D for bone.
• FERTILITY: NO prepubertal testicular-tissue cryopreservation — there is no method of in-vitro spermatogenesis, so it is not justified. For any post-pubertal patient wanting paternity, semen analysis + sperm cryopreservation first; if azoospermic, (micro)-TESE retrieves sperm in ~30-60% of KS men (no proven gain of micro- over conventional TESE). Do it BEFORE TRT (TRT suppresses spermatogenesis).
• TESE YIELD BY AGE — earlier is NOT better: only ~10% at age 13-14, jumping to ~45% at 15-19, and 15-19 is no better than young adulthood (20-30). So there is no fertility reason to harvest in early adolescence; higher serum LH predicts lower yield.

The HPG axis fires three times in life: fetal, mini-puberty (0-6 mo), and true puberty.

Mini-puberty is THE WINDOW for diagnosing CHH/Kallmann, Klinefelter, and certain DSDs by hormone phenotype alone. A boy with bilateral cryptorchidism + micropenis: draw LH, FSH, T, inhibin B, AMH between 4 weeks and 4 months. Flat gonadotropins + low T + low inhibin B = CHH.

MINIPUBERTY — the choreography, and why it is more than a diagnostic convenience (Rohayem 2024 Endocr Rev; Serbis 2025 Endocrines). At birth the placental-estrogen brake lifts and the axis fires, differently by sex:

• BOYS (LH/T-driven, BRIEF): LH rises in week 2; testosterone peaks at 1-3 months (near-pubertal), gone by ~6 months; INSL3 peaks ~1 month. FSH meanwhile drives the SERTOLI readout — AMH and inhibin B peak LATER (~4 months) and stay up through childhood; testis volume peaks 4-5 months. (That later, persistent Sertoli signal is what makes the FSH x AMH / FSH x inhibin B product readable in infancy, §52.)
• GIRLS (FSH/E2-driven, LONGER): FSH-predominant, with fluctuating estradiol and follicular waves; breast / uterine changes can run to age 2-3 years (the tail behind benign premature thelarche and neonatal ovarian cysts).
• PRETERM / SGA: PRETERM amplifies and prolongs it (higher, longer LH/FSH/T, faster genital growth); SGA shifts its duration and testosterone by birth weight — both want their own infant reference ranges, not the term ones.

The point beyond diagnosis: the FSH surge of minipuberty EXPANDS THE SERTOLI-CELL POOL, and that pool sets the LIFETIME spermatogenic ceiling. A boy with severe CHH gets NO minipuberty -> a permanently small Sertoli pool -> future infertility that later treatment cannot fully rescue. That is the rationale for neonatal gonadotropin replacement (mimic the signal with rFSH + LH/hCG, or a GnRH pump; §54) — and why the 4-12 week window is the one cheap, stimulation-free read of the entire axis before it goes quiet until puberty a decade later. And it is not only reproductive: the same surge is proposed to shape sex-specific brain circuits — a candidate window for later cognitive / behavioral programming, biology plausible but outcome data still thin (Serbis 2025).

Quiescent phase (~6mo - 6-8y): GABAergic + NPY + MKRN3 + DLK1 brakes on kisspeptin neurons.

Pubertal onset is reawakening of GnRH pulsatility, driven by kisspeptin (KISS1/KISS1R). Two kisspeptin populations: arcuate KNDy (pulse generator) and AVPV (surge generator, female only).

MKRN3 and DLK1 are PATERNALLY EXPRESSED brakes. LoF causes CPP, and the inheritance pattern follows imprinting: FATHER TRANSMITS, mother is silent carrier. A girl with CPP whose father had early puberty (or who appears unaffected because he was at the early end of normal) — that's your pedigree.

Tanner anchors:

• Girls: thelarche typical first sign, normal 8-13. PHV at B2-B3 (~12y). Menarche at B3-B4 (mean B4), ~2.5y after thelarche.
• Boys: testis ≥4 mL = puberty, normal 9-14. PHV at G3-G4 (~14y). Spermarche at G3-G4 (13-14y).

FSH vs LH FEEDBACK — the asymmetry that makes ratios diagnostic:

• LH brake = sex steroids (T in ♂, E2 in ♀).
• FSH brake = INHIBIN B (Sertoli ♂ / granulosa ♀). Sex steroids weaker.
• LH wants HIGH GnRH pulse frequency (~1/hr); FSH wants LOW (~q2-3h).
• Only LH gets E2 positive feedback (midcycle surge).

So "why is FSH high?" almost always = LOSS OF INHIBIN B.

Why Klinefelter and Turner are FSH-dominant:

• Klinefelter: tubule failure -> Sertoli dysfunction -> inhibin B falls first -> FSH ↑↑↑ leads, LH ↑ follows, T low-normal. Adolescent pattern is "FSH leads the way".
• Turner: streak gonads lose granulosa (inhibin B) AND theca E2. Both rise, but FSH > LH from infancy (the "infantile FSH peak").

Clinical reads in one line each:

• High FSH > high LH + low sex steroids = primary gonadal failure (Klinefelter, Turner, POI, chemo/RT).
• High LH > high FSH + hyperandrogenic girl = PCOS/PMOS (lean phenotype; obese blunts the ratio).
• Low LH + low FSH + low sex steroids = central (CHH, acquired).
• High LH + high T + 46,XY undervirilized = AIS.
• Inappropriately NORMAL LH/FSH at low T or E2 = functionally hypogonadotropic.

Estrogen leaves a visible SIGNATURE on the target tissues — breast and the vulvovaginal mucosa. Reading that signature does two jobs: it CONFIRMS true central activation (not isolated thelarche), and it gauges estrogen exposure when the serum estradiol sits BELOW the sensitivity of routine assays — which, at pubertal-onset levels, it usually does. The tissues are the bioassay the lab can't run.

BREAST — estrogen drives DUCTAL elongation + stromal / fat deposition; the palpable breast bud (Tanner B2, glandular tissue under the areola) is the FIRST sign of central puberty in girls. Stage by PALPATION, not inspection — adipose alone (pseudo-thelarche in the obese girl) is soft and has no firm subareolar disc. Areolar enlargement + the secondary mound (B4) come later. The bud is the switch; the vulvovaginal changes below corroborate that the switch is estrogen.

VULVOVAGINAL ESTROGEN EFFECT — the under-used half of the exam:

So a girl with a firm B2 bud PLUS pink dull mucosa, redundant hymen, and leukorrhea is genuinely estrogenized and centrally pubertal. A girl with breast tissue but a RED thin mucosa and no leukorrhea is far more likely premature thelarche — breast without systemic estrogenization (§50). The vaginal smear / maturation index is the old-school estrogen bioassay still useful exactly when the E2 assay reads "<20".

WHY THE "TUBULAR" BREAST HAPPENS — shape is set by the ORDER and TEMPO of two hormones. ESTROGEN builds the DUCTAL tree (terminal end buds, ductal elongation). PROGESTERONE — which only turns cyclic AFTER ovulatory cycles begin, i.e. months to years post-menarche — then fills it out with LOBULOALVEOLAR side-branching and the stroma / fat that ROUND the breast. Normal puberty gives a long, slow, low-dose unopposed-estrogen runway first, so ducts elongate gradually while the areola and stroma keep pace. Force estrogen in TOO FAST or TOO HIGH — over-rapid pubertal induction, or jumping straight to an adult estrogen dose — and ductal overgrowth OUTRUNS the stromal / areolar development: tissue herniates forward into the areola and the breast base stays constricted -> the conical tubular ("tuberous") breast with a puffy enlarged areola. It is largely IRREVERSIBLE, which is the whole reason induction is START-LOW-GO-SLOW: titrate transdermal estradiol up over ~2-3 years to mimic the physiologic ramp and protect breast contour (§54, and the same logic in Turner / POI induction, §47, §55).

CPP = GnRH-DEPENDENT puberty signs <8 girls, <9 boys — premature reactivation of pulsatile GnRH secretion (the central axis switching on early). The 2026 Endocrine Society guideline rebuilt the workup around one finding: most slowly-progressive early puberty reaches a NORMAL adult height untreated — so the default posture is OBSERVE, not reflex-test-and-treat. It reads as ten decision points, five diagnostic and five therapeutic, woven through below.

CONFIRM CENTRAL ACTIVATION — basal LH FIRST. The opening test is a basal (unstimulated) LH on an ULTRASENSITIVE assay (one that reliably reads down to <0.05-0.1 IU/L); basal LH ≥0.3 alone proves central activation. Reserve the GnRH/GnRHa STIMULATION test for the equivocal case — a low basal LH with persistent clinical suspicion, or when a fast answer is needed (rapid tempo). The stimulated pattern is unchanged: peak LH ≥5 with LH/FSH >0.66 = central; E2 >10-20 or T >30 supports. First-morning (8-10 AM) basal LH is best in early puberty, but any time of day is acceptable.

WHEN TO EVEN WORK IT UP — watchful waiting is now the default for early breast development: most regresses or stalls (progression to true CPP runs only ~0-31% across series), so OBSERVE before you test.

• Girls 7-8 yr with thelarche (Tanner B2): WATCHFUL WAITING — serial exams (growth velocity + Tanner) q4-6mo, NOT immediate labs/imaging.
• Girls <7 yr with thelarche: a 4-6 month OBSERVATION period first, to sort unsustained / slowly-progressive from rapidly-progressive puberty before any workup.
• Skip the wait and evaluate NOW for: Tanner B3+ at presentation, progression to B3 within 6 months of thelarche onset, growth ACCELERATION (climbing height percentiles, growth velocity >6 cm/yr), or ANY CNS sign (headache, seizure, visual-field deficit).
• Trap: lipomastia (fatty tissue in an overweight girl) mimicking breast — PALPATE for true glandular tissue before you call it thelarche.

PREMATURE THELARCHE — the benign isolated-breast mimic, and the first thing to exclude before working a girl up for CPP. Isolated breast development with NOTHING else: no growth acceleration, bone age = chronological age, no pubic hair, no vaginal bleeding, prepubertal uterus on US. Peaks in the first 1-3 years (a tail of mini-puberty — transient FSH-driven ovarian follicular activity throwing off small estradiol pulses) and is typically NON-progressive or waxing/waning. The lab giveaway on GnRH stim is an FSH-PREDOMINANT response with LH staying prepubertal — the mirror of CPP's LH-predominant peak. Management is REASSURE + SERIAL OBSERVATION (growth velocity, Tanner, bone age q6mo); no GnRHa, no estrogen workup beyond the basics. The trap is the thelarche variant / slowly-progressive thelarche — mild growth and bone-age acceleration that can evolve into true CPP — so any breast that PROGRESSES, accelerates growth, advances bone age, adds pubic hair, or bleeds gets re-worked as CPP. Premature thelarche is to estrogen what benign premature adrenarche (§60) is to androgen: an isolated, non-progressive partial-puberty variant you follow, not treat.

PREPUBERTAL VAGINAL BLEEDING — before calling ANY of it "puberty," clear the NON-endocrine causes in a girl with no other pubertal signs: a vaginal FOREIGN BODY (foul bloody discharge — the classic), trauma / straddle injury or SEXUAL ABUSE (always consider), urethral prolapse, lichen sclerosus, vulvovaginitis, and exogenous estrogen. Then clear estrogen- driven bleeding: McCune-Albright / peripheral PP, severe primary hypothyroidism (Van Wyk-Grumbach), an estrogen-secreting tumor. BENIGN (ISOLATED) PREMATURE MENARCHE is the diagnosis of EXCLUSION left over — one or a few episodes of cyclical bleeding WITHOUT thelarche, growth acceleration, or bone-age advance, with prepubertal LH/FSH and estradiol; self-limited, non-progressive. It is the bleeding-side member of the isolated-variant family (premature thelarche / adrenarche) — you follow it, but ONLY after the foreign-body / abuse / tumor workup is clean.

BRAIN MRI — no longer reflexive. The 2026 guideline restricts routine imaging because the yield is AGE-driven, not just sex-driven. Pathologic intracranial lesions (hamartoma or tumor) by group:

So: MRI the high-yield young — girls <6 and boys <8 — plus ANY child with CNS findings at any age. Do NOT routinely image girls 6-8 or boys 8-9 who lack CNS signs — the line that retires the old "every boy gets an MRI" reflex. Overall any-abnormality rate ~18%, pathologic ~5.5%; the remainder are incidentalomas that buy only cost and anxiety.

Etiologies:

• Idiopathic dominates girls (>95%).
• CNS organic: hamartoma (gelastic seizures), glioma (NF1), hydrocephalus, post-XRT, post-trauma, post-infection.
• Genetic: MKRN3 LoF (most common monogenic, paternally expressed), DLK1 LoF (paternally expressed, with metabolic phenotype), KISS1/ KISS1R GoF (rare).
• Peripheral (GnRH-INDEPENDENT) PP gets its OWN section (§51) — McCune-Albright, testotoxicosis, hCG tumors, gonadal / adrenal tumors, CAH, and severe primary hypothyroidism (Van Wyk-Grumbach).

GENETIC TESTING is NOT routine. Reserve TARGETED testing (MKRN3 / DLK1 / MECP2 — not broad genomic sequencing) for FAMILIAL CPP, offered through shared decision-making: a hit reshapes recurrence counseling but rarely changes this child's own management.

TREATMENT — the five therapeutic decision points:

• WHO benefits: GnRHa helps MANY, not all. Net benefit is driven by ADULT HEIGHT and is largest started YOUNG. Subgroups that do NOT net-benefit on height — weigh treatment harder, lean toward observation: girls 7-8 with SLOWLY-progressive CPP, and any child already AT or BEYOND peak height velocity (the bone age tells you). Decide WITH the family.
• WHICH drug: GnRHa is standard — leuprolide depot, triptorelin, histrelin SC implant — continuous receptor occupation desensitizes the gonadotrope. If you intend long-term suppression, START with a LONG-ACTING preparation (≥3-month depot, 6-month depot, or 12-month implant) rather than starting monthly then switching; start monthly only if the family plans to stay on monthly.
• MONITOR clinically, not biochemically: growth velocity, Tanner stage, and an ANNUAL bone age suffice. Do NOT routinely draw LH / sex-steroid "suppression" labs. Reserve biochemical testing for SUSPECTED failure — breast or testis progression, re-accelerating growth velocity — and first check ADHERENCE and injection technique / implant integrity.
• GROWTH HORMONE: do NOT routinely add GH to GnRHa, unless the child has a separate, established GH indication.
• WHEN to STOP: don't routinely continue past CHRONOLOGIC age 10-11 or bone age 11-12 (girls), or CA 11-12 / BA 12-13 (boys) — EARLIER than the old "stop at bone age ~12.5" reflex, because continuing longer adds little height. Go beyond these ages only for individualized reasons (growth trajectory, psychosocial, neurocognitive delay).

GnRHa SIDE EFFECTS & SAFETY — mostly mild, but two FDA-labeled pediatric warnings to know. The MECHANISM sets up the first issue: a GnRH AGONIST STIMULATES before it desensitizes, so the opening 1-2 weeks carry a FLARE — a transient gonadotropin / estrogen surge that can cause withdrawal VAGINAL BLEEDING and brief pubertal progression in girls. Counsel for it; it settles as the gonadotrope downregulates.

• INJECTION-SITE reactions and STERILE ABSCESS (depot leuprolide / triptorelin) — and an abscess matters beyond cosmetics: poor drug absorption can mean FAILED SUPPRESSION / breakthrough puberty, so an abscess + rising LH / estradiol means check the device, don't assume non-adherence. Histrelin implant: site reaction, breakage, difficult removal.
• HYPOESTROGENIC / menopause-like: hot flushes, headache, mood / emotional lability, nausea, increased appetite — usually mild and self-limited.
• PSEUDOTUMOR CEREBRI (idiopathic intracranial hypertension) — the 2022 FDA class warning. Headache, visual change (blurring, diplopia), papilledema, tinnitus -> fundoscopy, and stop if confirmed. (Same signal you already counsel for on GH, §39.)
• CONVULSIONS — a labeled warning, reported WITH and WITHOUT a prior seizure history or CNS risk factors.
• BONE MINERAL DENSITY — a transient DIP in BMD accrual during treatment that RECOVERS after stopping; peak bone mass is NOT compromised — reassure, no routine DXA needed for CPP dosing.
• WEIGHT — appetite can rise, but GnRHa is essentially WEIGHT-NEUTRAL over time; the BMI elevation often seen in CPP is mostly baseline, not drug-caused. Counsel diet / activity, don't blame the drug.
• SCFE — rare but reported during GnRHa for CPP (hypoestrogenic growth-plate change); knee / hip pain or a limp on treatment earns a frog-leg lateral film (§39).

The long-game reassurance families actually want: the axis RECOVERS — menses return ~12-18 months after stopping, puberty completes normally, and adult reproductive function and fertility are preserved.

PERIPHERAL PP (gonadotropin-INDEPENDENT, GIPP) = sex steroids arriving from a source OTHER than an awakened HPG axis, so on GnRH stim the LH STAYS SUPPRESSED — the mirror of CPP's LH rise (§50). The steroid comes from an autonomous gonad, the adrenal, a tumor, hCG, or outside the child. CRUCIAL caveat: ANY sustained peripheral sex-steroid exposure can mature the hypothalamus and trigger SECONDARY CENTRAL PP — the GnRH-stim then flips to LH-dominant, and a GnRHa becomes useful ON TOP of treating the source.

THE FULL DIFFERENTIAL, BY SOURCE of the sex steroid (each entity tagged ISOsexual = matches the child's sex, or CONTRAsexual):

1. GONADAL — autonomous activation or a steroid-secreting tumor:

• OVARY (girls): autonomous follicular CYST (McCune-Albright, the single commonest peripheral cause in girls — ISO, estrogen); JUVENILE GRANULOSA-CELL tumor (the commonest estrogen-secreting ovarian tumor; ISO precocity in ~70-80% — ISO, estrogen); SEX-CORD TUMOR WITH ANNULAR TUBULES (SCTAT, the Peutz-Jeghers ovarian tumor -> estrogen — ISO); thecoma; a Leydig / steroid-cell tumor or gonadoblastoma instead makes ANDROGEN (CONTRA, virilizing).
• TESTIS (boys): LEYDIG-CELL tumor (testosterone, UNILATERAL — ISO); familial male-limited PP (testotoxicosis, LHCGR GoF, BILATERAL — ISO); LARGE-CELL CALCIFYING SERTOLI-CELL TUMOR (LCCSCT) — overexpresses aromatase -> ESTRADIOL -> GYNECOMASTIA + advanced bone age (CONTRA), and flags Carney complex (PRKAR1A) or Peutz-Jeghers (STK11); McCune-Albright (autonomous Leydig). Boy-gynecomastia differential: §62.

2. ADRENAL: CAH (21-OH > 11β-OH) — androgen, ISO in boys but VIRILIZING (CONTRA) in girls (§10); adrenocortical ADENOMA / CARCINOMA — usually virilizing (± Cushing), rarely feminizing. 3. hCG-SECRETING TUMORS (hCG -> LHCGR -> Leydig testosterone, so BOYS ONLY — a girl's ovary needs FSH too): hepatoblastoma (liver), mediastinal / pineal / intracranial GERMINOMA, choriocarcinoma, teratoma. Always check beta-hCG in a boy. 4. McCUNE-ALBRIGHT (mosaic GNAS R201) — autonomous gonadal activation (box below). 5. SEVERE PRIMARY HYPOTHYROIDISM (Van Wyk-Grumbach) — very high TSH cross-activates the FSHR (box below); the ONLY peripheral cause with DELAYED rather than advanced bone age. 6. EXOGENOUS / ENVIRONMENTAL: estrogen or testosterone cream / gel (caregiver transfer), oral or anabolic steroids, endocrine-disrupting chemicals (§81). 7. PERIPHERAL AROMATASE EXCESS (aromatase excess syndrome, CYP19A1 gain) — runaway T -> E2 -> familial gynecomastia / feminization + early fusion (§9).

THE KILLER DISCRIMINATOR IN A BOY IS TESTICULAR SIZE — the drive is NOT gonadotropin, so the testes usually stay PREPUBERTAL / small, and the exceptions LOCALIZE the source:

• SMALL testes + virilization => ADRENAL source (CAH or adrenal tumor) — adrenal androgens virilize but do NOT grow the testis. Check 17-OHP, DHEAS, androstenedione.
• BILATERAL enlargement => the drive is hitting the LHCGR: testotoxicosis, an hCG-secreting tumor (check beta-hCG), or McCune-Albright.
• UNILATERAL enlargement => Leydig-cell tumor (testicular US).

(The big-testes differential is §63.)

WORKUP: confirm PERIPHERAL with a suppressed basal + GnRH-stimulated LH, then LOCALIZE by the steroid pattern — estradiol / testosterone, adrenal androgens (17-OHP, DHEAS), beta-hCG, and a TSH (don't miss Van Wyk-Grumbach) — with imaging directed by the pattern (pelvic / adrenal US, testicular US). Bone age is ADVANCED in all of these EXCEPT Van Wyk-Grumbach, where it is paradoxically DELAYED.

TREATMENT = treat / remove the source, then BLOCK whichever steroid arm is still firing: an AROMATASE INHIBITOR (stops the estrogen that races the bone age) +/- an ANDROGEN-RECEPTOR BLOCKER; add a GnRHa ONLY once secondary central PP supervenes. The source-specific regimens:

McCUNE-ALBRIGHT SYNDROME (MAS) — one POST-ZYGOTIC mosaic GNAS R201 activating mutation (constitutive Gsα -> cAMP), so the phenotype is patchy and each manifestation is wherever the mutant clone landed. The classic TRIAD: (1) café-au-lait macules with JAGGED "coast of Maine" borders that respect the midline; (2) POLYOSTOTIC FIBROUS DYSPLASIA (fracture, pain, deformity); (3) AUTONOMOUS hyperfunction of cAMP-driven endocrine glands — all gonadotropin/TSH/ACTH-INDEPENDENT, so SCREEN each:

• Peripheral PP — recurrent OVARIAN CYSTS (the commonest MAS endocrinopathy in girls; vaginal bleeding is often the first sign) -> aromatase inhibitor (letrozole) / tamoxifen / fulvestrant.
• Autonomous HYPERTHYROIDISM (nodular, §34); GH excess / gigantism (§79); neonatal CUSHING from autonomous adrenal nodules (§16).
• FGF23-driven HYPOPHOSPHATEMIA from the fibrous-dysplasia lesions (§20).

Because GNAS is mosaic, a BLOOD test can be NEGATIVE — test affected tissue. (Contrast: GERMLINE GNAS = AHO / PHP, §24.)

VAN WYK-GRUMBACH — the striking paradox: a child with profound long-standing primary hypothyroidism (untreated congenital, autoimmune Hashimoto, post-ablative without replacement) develops PERIPHERAL PP with breast development in girls, multicystic ovaries on US, sometimes vaginal bleeding, AND macroorchidism in boys WITHOUT virilization. Bone age is DELAYED (thyroid hormone is needed for bone maturation) while sexual maturation runs ahead: bone age << chronological age << pubertal stage.

The mechanism is glycoprotein-hormone cross-talk. TSH, FSH, LH, and hCG share a common alpha-subunit; they differ only in their beta-subunits. At physiologic concentrations the FSHR discriminates ligand cleanly. But when TSH climbs into the hundreds (often >500 mIU/L in long-standing untreated primary hypothyroidism), the receptor cross-reactivity becomes biologically meaningful. At these supraphysiologic levels TSH cross-activates the FSHR more readily than the LHCGR (the FSHR is simply the more PROMISCUOUS receptor here), so FSH-like ovarian / testicular signaling dominates while LH-driven steroidogenesis is largely spared. This is why girls show breast development and multicystic ovaries (FSH effect at granulosa) WITHOUT precocious adrenarche or acne, and boys show macroorchidism (FSH effect on Sertoli cells driving seminiferous tubule expansion) without significant testosterone production or virilization. PRL is also often elevated because chronic TRH cross-stimulates lactotrophs (separate axis, same TRH).

The Rx is purely thyroid hormone replacement — L-T4 corrects the TSH, the cross-stim collapses, and the pubertal signs regress. Do NOT start GnRHa. Tip-offs: dry skin, slow growth velocity, goiter, delayed bone age in a child with "peripheral PP". TSH first.

TESTOTOXICOSIS (familial male-limited PP, LHCGR GoF): the testis is AUTONOMOUS, so a GnRHa ALONE FAILS (nothing for it to suppress). Block the two downstream arms — an ANDROGEN-RECEPTOR BLOCKER + an AROMATASE INHIBITOR: the modern combo is bicalutamide + anastrozole / letrozole (the AI also stops the T->E2 that races the bone age); older regimens are spironolactone + testolactone, or ketoconazole / cyproterone (steroidogenesis block). ADD a GnRHa ONLY once SECONDARY CENTRAL puberty supervenes — chronic sex-steroid exposure matures the HPG axis, and the now-active gonadotrope becomes suppressible.

The mirror of CPP. Same axis, opposite end. Abaci & Besci 2024 (J Clin Res Pediatr Endocrinol) is the framework to quote.

DEFINITION:

• Boys: NO testicular enlargement (TV <4 mL) by age 14.
• Girls: NO glandular breast tissue (B1) by age 13.
• Statistical anchor: 2 to 2.5 SD above the population mean for onset.
• Delayed PROGRESSION / pubertal arrest: time from onset to completion >5 years in boys, OR no menarche by 15-16 (15 in the Abaci framework; 16 in older primary-amenorrhea definitions), OR no menarche within 3 years of thelarche in girls.

NOMENCLATURE — the term you use matters: SELF-LIMITED DELAYED PUBERTY (SLDP) is the CURRENT PREFERRED LABEL. The older name was CONSTITUTIONAL DELAY OF GROWTH AND PUBERTY (CDGP, also "constitutional delay of puberty"). The field shifted because "constitutional" implied a fixed trait, whereas the entity is by definition transient — pubertal maturation completes spontaneously by age 18, full sexual development within 2-4 years of pubertal onset. SLDP captures the biology: late, but self-resolving. CDGP is now a legacy synonym you'll still see in older texts. Going forward in this compendium I'll use SLDP as the primary term, with CDGP in parentheses where the older literature demands recognition.

THREE BUCKETS (Abaci classification):

• Hypergonadotropic hypogonadism (primary gonadal failure).
• Permanent hypogonadotropic hypogonadism (CHH and acquired central failure).
• Transient hypogonadotropic hypogonadism = SLDP (CDGP) + Functional HH.

So in boys, SLDP dominates (~6 of 10 delayed-pubertal boys are SLDP). In girls the differential is more evenly spread; SLDP is less of a default.

SLDP (formerly CDGP): diagnosis of EXCLUSION. Self-limited by definition. 50-80% have AD family history of delayed puberty, bone age lags chronological age but matches pubertal stage when it arrives. Gold standard for distinguishing SLDP from CHH = CLINICAL OBSERVATION UNTIL AGE 18 (testis volume reaches ≥4 mL by 18, or reaches ≥4 mL between 14 and 18 and progresses to ≥15 mL within 4 years). If endogenous puberty kicks in by then, it was SLDP. Anything short of 18 is provisional.

ACQUIRED CENTRAL HYPOGONADISM: tumor (craniopharyngioma, germinoma, prolactinoma), LCH, sarcoidosis, post-XRT, post-trauma, post- meningitis/encephalitis.

FUNCTIONAL HH: underlying systemic disease drives REVERSIBLE suppression. Drivers:

• Chronic disease: celiac, IBD, CF, asthma, sickle cell, thalassemia, CKD.
• Endocrine: hypothyroidism, hyperprolactinemia, GHD, hypercortisolism, poorly-controlled T1DM.
• Energy deficit: anorexia nervosa, restrictive eating, excessive exercise (gymnastics, ballet, distance running, weight-class sports).
• Medications: antipsychotics (via PRL), antidepressants, opioids, chronic glucocorticoids.

Functional HH reverses once the underlying disease is treated. 10-20% of total delayed puberty.

HYPERGONADOTROPIC HYPOGONADISM:

• Congenital: Turner, Klinefelter, Noonan, Down, Fragile X premutation, 46,XX or 46,XY gonadal dysgenesis (Swyer), AIS, testicular regression / anorchia, galactosemia, steroidogenic defects (5aRD2, 17,20-lyase, lipoid CAH, 17b-HSD), Sertoli-cell-only syndrome.
• Acquired: chemotherapy (alkylators most gonadotoxic), radiation (gonadal or cranial), autoimmune oophoritis (APS-1, APS-2), gonadal infection (mumps orchitis, coxsackie), surgery, trauma, torsion.