Recombinant somatropin, the GH–IGF-1 axis, and the epiphyseal growth plate — from GHD diagnosis to predicted adult height
Growth hormone deficiency (GHD) is diagnosed by combining longitudinal auxology (height and height-velocity tracking against CDC/WHO growth charts) with skeletal maturity assessment (bone age) and biochemical confirmation via GH stimulation testing. A child whose bone age lags meaningfully behind chronological age, with a subnormal growth velocity, is the classic presentation that triggers pediatric endocrine referral.
Growth is plotted on CDC (US) or WHO (international, <2 yr) reference charts at every well-child visit. Two auxological triggers prompt an endocrine growth evaluation:
• Height below the 3rd percentile (or >2 SD below the population mean) for age and sex • Height crossing downward across two major percentile channels within 12–18 months, even if still within the normal range • Height velocity below the 25th percentile for bone age over a well-documented 6–12 month interval
Height velocity (HV) is calculated in cm/year from two accurately measured stadiometer heights at least 6 months apart: HV = (height2 − height1) / (interval in years). A healthy mid-childhood HV is roughly 5–6 cm/yr; the case profiled here — 3.8 cm/yr at chronological age 8 — falls well below the 25th percentile and, combined with a low IGF-1 SDS, is the auxological hallmark of GHD.
Skeletal (bone) age is assessed from a single postero-anterior radiograph of the left hand and wrist — chosen because it captures dozens of independently maturing ossification centers (epiphyses of the phalanges, metacarpals, radius, ulna, and carpal bones) in one low-dose image.
Greulich-Pyle (GP) method: the radiograph is visually matched to the closest-fitting reference plate in the 1959 Greulich and Pyle atlas, a set of standard films from the Brush Foundation growth study, arranged by age and sex. It is fast (minutes) and remains the most widely used clinical method, though matching a whole hand to a single reference plate has interobserver variability of roughly ±0.4–0.7 years.
Tanner-Whitehouse (TW2/TW3) method: a more granular, quantitative alternative. Each of 20 individual bones (13 in the RUS — radius, ulna, short bones — subset commonly used clinically) is scored on a maturity stage (A through H/I), and stage scores are summed into a total maturity score that converts to a bone age via published tables. TW methods are more reproducible but slower and less commonly used outside research or forensic settings.
In this case, a chronological age of 8.0 years with a bone age reading of 6.0 years documents a 2-year skeletal delay — consistent with, though not exclusively diagnostic of, growth hormone deficiency (constitutional delay of growth and puberty produces a similar pattern without a GH defect).
Because GH is secreted in pulses (mainly during slow-wave sleep) and a single random serum level is uninterpretable, GHD is confirmed provocatively: a pharmacologic stimulus is given to force a GH pulse, and serum GH is sampled serially to find the peak.
Commonly used provocative agents: • Insulin tolerance test (ITT) — the historical gold standard. IV insulin (0.05–0.1 U/kg) induces hypoglycemia (documented glucose <40 mg/dL); GH sampled every 15–30 min for 90–120 min. Requires close medical supervision for hypoglycemia symptoms. • Arginine infusion — IV arginine (0.5 g/kg over 30 min); commonly paired with a second agent (arginine + clonidine, or arginine + GHRH) to improve specificity. • Clonidine — oral α2-agonist (0.15 mg/m²); simple to administer, widely used as a first-line outpatient test. • Glucagon stimulation test — IM glucagon (1 mg); increasingly used in adults and where ITT is contraindicated.
A peak GH below 10 ng/mL on a traditional polyclonal RIA (or below roughly 5–7 ng/mL on newer, more specific monoclonal immunoassays) is considered subnormal. Because any single test has a high false-positive rate (up to 20%) from normal variability in pulsatile secretion, guidelines from the GH Research Society and the Pediatric Endocrine Society require two failed stimulation tests, in the context of consistent auxology and a low IGF-1/IGFBP-3, before a GHD diagnosis is made.
Bone age alone is never diagnostic. A 2-year skeletal delay can reflect constitutional delay, chronic illness, malnutrition, hypothyroidism, or GHD. The diagnosis of GHD requires the triad: subnormal growth velocity for bone age, low IGF-1/IGFBP-3, and two failed GH stimulation tests — per Pediatric Endocrine Society and GH Research Society consensus guidelines.
Recombinant human growth hormone (somatropin) is a 191-amino-acid, 22 kDa protein identical in sequence to pituitary-derived GH, manufactured by recombinant DNA technology in E. coli or mammalian (CHO) cell lines. Daily subcutaneous injection restores physiologic GH exposure, driving hepatic and local IGF-1 production through the JAK2/STAT5 signal transduction pathway — the pharmacologic engine behind catch-up growth.
Several biosimilar and originator somatropin products are in routine pediatric use, all delivering the identical 191-residue GH sequence: Genotropin (Pfizer), Norditropin (Novo Nordisk), Humatrope (Eli Lilly), Saizen (Merck Serono/Emd Serono), and Omnitrope (Sandoz, the first FDA-approved biosimilar GH). Formulations differ mainly in device design — prefilled pens, cartridge-based pen injectors, and needle-free jet injectors — and in minor excipients affecting stability and injection comfort.
Dosing is weight-based (mg/kg/day) and given once daily by subcutaneous injection, classically in the evening to approximate the physiologic nocturnal GH secretory peak that occurs during slow-wave sleep. Injection sites are rotated (abdomen, thigh, buttock, upper arm) to prevent lipohypertrophy. Adherence is increasingly tracked electronically via connected pen injectors (e.g., easypod, a device that logs each dose and transmits adherence data to the care team) — non-adherence is one of the most common, and most correctable, causes of a poor growth response.
GH exerts its effects by binding the GH receptor (GHR), a single-pass transmembrane receptor that is pre-associated at the cell surface as an inactive dimer. GH binding induces a conformational change in the GHR dimer that activates JAK2 (Janus kinase 2), a receptor-associated tyrosine kinase, docked on the receptor’s cytoplasmic tail.
Activated JAK2 trans-phosphorylates itself and the GHR cytoplasmic domain, creating docking sites for STAT5b (Signal Transducer and Activator of Transcription 5b). Phosphorylated STAT5b dimerizes, translocates to the nucleus, and binds GH-response elements in the promoter of the IGF1 gene, driving its transcription. Hepatocytes are the dominant source of circulating (endocrine) IGF-1, though growth-plate chondrocytes, muscle, and other tissues also produce IGF-1 locally (paracrine/autocrine) under direct GH stimulation — the basis of the "dual effector" model of GH action described in the next stage.
Once secreted, over 90% of circulating IGF-1 travels bound in a ternary complex with IGF-binding protein 3 (IGFBP-3) and acid-labile subunit (ALS), which extends its serum half-life from minutes (free IGF-1) to many hours, buffering tissue exposure and providing a stable circulating reservoir.
Because the growth response to GH varies with dose, adherence, pubertal status, and individual sensitivity, therapy is titrated using serial biomarkers and auxology rather than a fixed dose:
• IGF-1 and IGFBP-3 measured every 3–6 months; the dose is adjusted to keep IGF-1 within roughly 0 to +2 SDS for age and sex — high enough to drive an adequate growth response, but avoiding sustained supraphysiologic IGF-1 levels, a theoretical long-term safety concern examined extensively in large post-marketing surveillance cohorts (e.g., the Genentech National Cooperative Growth Study, and the international KIGS/GeNeSIS registries). • Height velocity is re-assessed at 6 and 12 months; a first-year gain less than roughly 50% of the model-predicted response (see Stage 4) prompts a structured work-up for non-adherence, incorrect dosing, undiagnosed hypothyroidism, celiac disease, or an alternative diagnosis. • Bone age is re-imaged approximately annually, both to track how quickly skeletal maturation is advancing relative to height gain, and — critically — to recalculate predicted adult height and to time eventual treatment discontinuation.
GH does not act only through circulating IGF-1. Landmark work by Olle Isaksson, Judith Green, and colleagues in the 1980s (the "dual effector theory") showed GH acts directly on growth-plate prechondrocytes to trigger their differentiation, while both endocrine (hepatic) and locally produced IGF-1 drive the subsequent clonal expansion of chondrocyte columns — explaining why isolated liver-specific IGF-1 knockout mice still grow nearly normally.
Longitudinal bone growth occurs entirely at the epiphyseal (growth) plate, a thin disc of cartilage sandwiched between the epiphysis and metaphysis of every long bone. GH and IGF-1 converge on this structure to accelerate a tightly organized cellular assembly line — chondrocyte proliferation, hypertrophy, and replacement by bone — that is the true cellular engine behind every centimeter gained on a growth chart.
A cross-section of an active growth plate reveals four histologically distinct zones, arranged in sequence from the epiphysis toward the metaphysis:
• Resting (reserve) zone: scattered, slowly-dividing stem-like chondrocytes anchored near the epiphyseal blood supply; this zone acts as a reservoir that periodically recruits new columns of proliferative cells. • Proliferative zone: chondrocytes flatten, align, and divide rapidly in the direction of bone growth, stacking into tall parallel columns — the "coin stacks" visible on histology — each column derived clonally from a single resting-zone progenitor. • Hypertrophic zone: proliferation ceases and cells enlarge dramatically (roughly 5–10-fold in volume), the single largest contributor to plate elongation; these cells also mineralize their surrounding matrix and secrete factors (including VEGF) that trigger vascular invasion from the metaphyseal side. • Zone of ossification (metaphysis): blood vessels invade the mineralized cartilage scaffold, hypertrophic chondrocytes undergo apoptosis, and osteoblasts deposit new bone (primary spongiosa) directly onto the residual cartilage matrix — permanently converting cartilage template into bone.
Growth-plate chondrocytes express abundant IGF-1 receptor (IGF1R), a receptor tyrosine kinase. IGF-1 binding — whether delivered endocrine from the liver or produced locally within the plate — activates the PI3K/AKT pathway, promoting chondrocyte survival, proliferation-zone cell-cycle progression, and hypertrophic differentiation, together with parallel MAPK/ERK signaling that reinforces proliferative drive.
GH itself also acts directly at the plate: GH receptor is expressed on resting-zone prechondrocytes, and local GH binding triggers their initial commitment into the proliferative program — the direct arm of the dual-effector model. IGF-1 then amplifies and sustains the clonal expansion of each committed column. This two-tier control (GH recruits new columns; IGF-1 expands them) allows growth rate to be tuned continuously rather than as an all-or-nothing switch.
The net rate of bone lengthening equals the rate at which hypertrophic chondrocytes are added to the column, multiplied by their peak cell height, minus the rate at which the ossification front advances to consume them. During a rapid childhood growth phase this nets out to roughly 3 mm of longitudinal growth per month at an actively growing plate — summed across the many growth plates of the long bones, this produces the centimeters-per-year height velocity tracked clinically.
Because the entire process depends on a finite, non-regenerating supply of resting-zone chondrocytes, growth plates have a built-in "expiration": as puberty progresses and sex-steroid exposure rises, the resting zone is progressively depleted and the proliferative program slows, setting up the eventual plate fusion described in Stage 5.
The dual-effector theory (Green, Isaksson & Sjögren, 1985) reframed decades of growth physiology: GH is not simply "upstream" of IGF-1 in a single linear hormonal cascade. GH directly commits growth-plate stem-like chondrocytes to the proliferative program, while endocrine and locally-produced IGF-1 drive the clonal expansion of the resulting columns — a two-signal system that recombinant GH therapy re-activates at both levels simultaneously.
The clinical signature of a good GH response is a distinctive, well-characterized velocity curve: a sharp first-year acceleration — "catch-up growth" — that substantially outpaces the pretreatment baseline, followed by a gradual deceleration over subsequent years toward a sustainable, near-normal prepubertal growth rate as the child settles into a higher growth-chart percentile channel.
When GH-deficient children begin somatropin therapy, height velocity typically more than doubles within the first 6–12 months — commonly exceeding 10–12 cm/yr in a child whose pretreatment velocity was 3–4 cm/yr. This is driven by the simultaneous recruitment of a large cohort of resting-zone chondrocytes across many growth plates at once, compressing what would have been several years of subnormal growth into a single accelerated year.
Catch-up growth is not indefinite: as the child's height moves back toward its genetically-determined percentile channel, the "growth deficit" that fueled the acceleration is progressively exhausted, and velocity naturally decelerates even with unchanged GH dosing.
By year 2, height velocity typically falls to roughly 7–8 cm/yr, and by year 3 onward settles into a 5–6 cm/yr range — close to the normal prepubertal rate for a treated, non-GHD child, sustained until puberty brings its own growth-velocity peak.
Several validated prediction models (e.g., the Ranke index, and country-specific models derived from the KIGS international GH registry) estimate expected first-year growth response from baseline variables — chronological age, bone-age delay, baseline height SDS, GH dose, and body weight — and are used clinically as a benchmark: an actual response substantially below the model prediction is one of the strongest signals to investigate adherence or reconsider the diagnosis.
Height is plotted at every visit against CDC (US, ages 2–20) or WHO (international, ages 0–2) percentile curves, and height velocity is recalculated every 6 months from stadiometer measurements. A "good responder" gains at least 50% of the model-predicted first-year increment and visibly crosses upward across percentile channels on the chart.
A "poor responder" — less than roughly 50% of predicted gain — triggers a structured re-evaluation: confirm injection technique and adherence (often via connected-pen adherence logs), re-check IGF-1 to confirm biochemical response to the given dose, screen for undiagnosed hypothyroidism or celiac disease (both blunt GH responsiveness), and consider dose adjustment before assuming treatment failure.
A treated GHD child commonly gains 10–12 cm in year 1 versus a pretreatment trajectory of 3–4 cm/yr — a height-velocity multiple of roughly 3× — before decelerating over 2–3 years to a sustainable prepubertal rate. This first-year velocity peak is the most consistent, best-documented signature of an effective GH replacement regimen.
As treatment progresses, the clinical question shifts from "is the child growing faster" to "how tall will this child ultimately become, and when should therapy stop." Predicted adult height is estimated from bone age using the Bayley-Pinneau method, while the biological clock that ultimately ends growth is estrogen-driven fusion of the epiphyseal growth plates — the same steroid pathway in both sexes.
Bayley and Pinneau (1952) published tables giving, for a given bone age and sex, the percentage of eventual adult height already attained. Predicted adult height (PAH) is then simply:
PAH = current height ÷ (% of adult height attained at current bone age)
Bone age is read from the hand/wrist film (Greulich-Pyle), current standing height is measured, and the appropriate percentage is looked up from separate average- and advanced/retarded-maturation tables (the method further distinguishes children whose bone age is notably ahead of or behind chronological age, since skeletal maturity tempo itself affects the remaining growth percentage). In contemporary practice, PAH is frequently cross-checked against the Tanner-Whitehouse Mark II RUS-based prediction tables, which use the more granular TW bone-age score rather than a single GP-atlas match.
PAH is recalculated roughly annually throughout treatment, both to counsel the family on expected outcome and to detect an unexpectedly rapid advance in bone age (which would erode, rather than improve, the predicted adult height gain from therapy).
Growth plates do not close because cartilage simply "runs out" — closure is an actively signaled, hormonally timed event. Estrogen, acting through estrogen receptor-α expressed on growth-plate chondrocytes, drives progressive depletion of the resting-zone stem-like cell pool, slows proliferative-zone column height, and ultimately triggers a senescence-like arrest in which the remaining cartilage is fully replaced by bone, permanently sealing the growth plate.
Remarkably, this estrogen-dependent mechanism operates in boys as well as girls: circulating testosterone is peripherally aromatized to estradiol (via the enzyme aromatase) in bone and other tissues, and it is this locally generated estrogen — not testosterone directly — that ultimately closes the male growth plate. This was demonstrated definitively in rare cases of aromatase deficiency and estrogen-receptor mutations, in which affected males failed to fuse their growth plates and continued growing into adulthood despite normal or high testosterone levels.
Because rising pubertal sex steroids both accelerate bone-age advancement and eventually fuse the growth plate, the pace of bone-age progression relative to height gain — not chronological age — is what ultimately determines the growth window. This is why bone-age re-imaging, not a birthday, is what drives the decision of when to stop GH therapy.
GH therapy is generally discontinued once the growth plates are near-fused and further height gain would be minimal, using converging criteria:
• Bone age beyond approximately 14 years in girls or 16 years in boys (near-complete epiphyseal fusion on Greulich-Pyle reading) • Height velocity falling below roughly 2 cm/yr despite continued therapy • Growth plates read as "fused" or "near-fused" on serial hand/wrist films
Long-term cohort studies and international registries (KIGS, and the US National Cooperative Growth Study) following GHD children treated to near-adult height report an average gain of roughly +1.0 to +1.5 height SDS compared with predicted untreated adult height — translating, in an average-stature population, to several additional centimeters of final adult height, with the largest benefit seen in children who start therapy earlier and maintain good adherence throughout the treatment course. Patients with persistent GHD are subsequently re-screened in early adulthood, since childhood-onset GHD frequently persists and may warrant continued, lower-dose adult GH replacement.