Longitudinal surveillance of children on recombinant growth hormone — dose titration, IGF-1 targeting, adherence tracking, and adverse effect monitoring after therapy starts
Growth hormone response is not uniform — two children with identical diagnoses can show dramatically different first-year height velocities on the same weight-based dose. Prediction models built on large observational registries let clinicians set realistic expectations up front, and — critically — give a benchmark against which a disappointing real-world response can be flagged early rather than discovered a year later.
KIGS (Kabi International Growth Study, later Pfizer International Growth Database) pooled outcomes from tens of thousands of GH-treated children across dozens of countries, enabling regression models that predict first-year height velocity (HV1) from variables known before the first injection:
• Chronological age at start — younger children generally show a larger relative HV gain • Auxological deficit — height SDS below mid-parental target height SDS at baseline • Peak GH on stimulation testing — lower peak (more severe deficiency) predicts a larger response • Body weight and starting dose (mg/kg/week) — dose-response relationship is well established • Maximum growth velocity in the year before treatment — a proxy for residual growth potential • IGF-1 SDS at baseline — very low pretreatment IGF-1 associates with brisker early response
Separate models exist for GH deficiency, Turner syndrome, small-for-gestational-age short stature, and chronic kidney disease, because the response magnitude and its determinants differ meaningfully by diagnosis.
A predicted HV1 becomes the individualized target for the first follow-up visit. A child tracking well below their model-predicted curve at month 3 is a poor-responder flag — prompting an adherence check, dose reassessment, or a search for a coexisting condition (undiagnosed hypothyroidism, celiac disease) rather than an assumption that GH "isn't working."
Before dosing starts, the baseline visit captures the reference values every later visit is compared against: standing height and sitting height (converted to SDS using population references), weight, pubertal (Tanner) stage, bone age via left-hand X-ray, and a baseline IGF-1 and IGFBP-3. This baseline bone age also anchors the eventual discontinuation decision — treatment is generally stopped once bone age reaches roughly 14 years in girls or 16 years in boys, when growth plates are nearly fused.
The first three months on therapy are the single most information-dense window in the entire treatment course: the gap between predicted and observed velocity at the 3-month visit is a far stronger early signal of true response than waiting a full year.
The first year on GH therapy is an active titration period, not a "start and wait" prescription. Daily subcutaneous injections begin at a weight-based starting dose, and every three-month visit is an opportunity to adjust that dose upward as the child gains weight and as early IGF-1 and growth data accumulate.
Recombinant human GH is dosed daily rather than weekly because the growth-promoting effect depends on the intermittent pulsatile pattern that daily evening injections best approximate; once-weekly long-acting GH formulations are increasingly used but require their own dose-equivalence tables.
Each quarterly visit in year one repeats a fixed panel: standing height (mid-parental-corrected SDS), weight, injection log review, IGF-1, and a brief systems check for headache or hip/knee pain. Dose is recalculated in mg/kg/week using current weight, since a growing, gaining child on a fixed mg dose is gradually being under-dosed relative to body mass.
A typical trajectory: month 3 velocity extrapolated to an annualized rate near 8–10 cm/yr, tapering somewhat by month 12 as the initial catch-up component wanes — the classic GH response curve is front-loaded, not linear.
Three broad patterns emerge by the month-6 or month-9 visit:
• Robust responder — velocity meeting or exceeding the model prediction, IGF-1 rising appropriately: continue current mg/kg dose, reassess at 12 months • Partial responder with rising IGF-1 already near the upper target band: hold dose rather than escalate, to avoid pushing IGF-1 supraphysiologic • Blunted responder with IGF-1 still low despite adequate mg/kg dosing: the differential includes missed doses (ask about the injection log before assuming true GH resistance), primary IGF-1 deficiency, undertreated hypothyroidism, unrecognized celiac disease, or poor injection technique delivering a subtherapeutic effective dose
By month 12, the observed height velocity is compared formally against the KIGS-style predicted value; a response under roughly 50% of predicted triggers a structured non-responder work-up rather than an automatic dose increase.
Insulin-like growth factor 1 (IGF-1) is the principal downstream mediator of GH's growth-promoting action, and its level in blood is far more stable day-to-day than pulsatile GH itself — making it the practical biomarker for dose titration. The modern approach keeps IGF-1 SDS within a defined target window rather than simply escalating a fixed mg/kg dose as the child grows.
Weight-based dosing alone assumes a fixed dose-response relationship across children, but GH sensitivity varies substantially by individual, pubertal stage, and diagnosis. Two children on an identical mg/kg dose can show IGF-1 SDS values that differ by more than two standard deviations.
IGF-1-based titration instead sets the biomarker as the target: dose is adjusted, up or down, to keep IGF-1 SDS within roughly 0 to +2 relative to age- and sex-matched normal ranges. This approach, validated in randomized comparisons against fixed-dose regimens, achieves comparable or superior height outcomes while reducing the fraction of children exposed to sustained supraphysiologic IGF-1 levels.
Practically: if IGF-1 SDS is below 0 despite adequate adherence, the dose is increased; if IGF-1 SDS drifts persistently above +2, the dose is trimmed even if height velocity looks favorable — because the ceiling on IGF-1 is a safety constraint, not merely an efficacy target.
Epidemiological and mechanistic data link chronically elevated IGF-1 to increased risk of several cancers in adults (notably colorectal and breast), and IGF-1 is a recognized mitogen in numerous cell lines. Large long-term cohorts of GH-treated children (SAGhE, and others) have not shown a clear excess of de novo malignancy from GH therapy itself at conventional doses, but they have prompted caution around avoidable, sustained supraphysiologic exposure.
This is the rationale for treating +2 SDS as a soft ceiling: it is a precautionary target grounded in biological plausibility and imperfect long-term safety data, not a value tied to any single confirmed clinical outcome. Because IGF-1 immunoassays vary meaningfully between platforms, clinics track the trend on a consistent assay over time rather than reacting to any single borderline value.
Because a single IGF-1 draw can be affected by acute illness, recent injection timing, and assay variability, dose changes are generally made on two consecutive elevated or two consecutive low readings — not a single outlier — to avoid over-correcting the dose in response to measurement noise.
A therapy that must be self- or parent-administered by injection nearly every day for years runs into a problem no dose-titration algorithm can solve on its own: adherence. Real-world studies consistently show that a substantial minority of patients miss doses, and missed doses — not biological non-response — are one of the most common, and most correctable, causes of a disappointing growth outcome.
Self-reported adherence is notoriously optimistic; caregivers under-report missed doses, and pharmacy refill timing only loosely approximates actual injections taken. Electronic injection devices — auto-injectors and needle-free devices with onboard memory — record the date, time, and dose of every actuation, giving clinicians an objective adherence log to review at each visit rather than relying on recall.
These logs commonly reveal patterns invisible to self-report: doses clustered around school mornings instead of the intended evening administration, multi-day gaps around travel or device malfunction, or a slow adherence decline over the second and third year of therapy as novelty wears off and the daily burden accumulates — a pattern sometimes called "adherence fatigue."
A sustained adherence rate below roughly 85% measurably blunts height velocity, and very poor adherence (missing more than 2–3 doses per week) can produce a growth trajectory indistinguishable from a true biological non-responder, so distinguishing the two is the first branch point in any disappointing-response work-up.
Daily injections in children carry a real psychological burden. Needle anxiety is common at treatment initiation and can independently drive missed doses regardless of family motivation; needle-free jet injectors, smaller-gauge auto-injector needles, topical anesthetic, and structured injection-training sessions with a nurse educator meaningfully reduce this barrier.
Technique matters beyond simply "getting the dose in": repeated injection into the same small skin area causes lipohypertrophy — localized fatty tissue thickening that impairs and delays local GH absorption, producing an apparent partial non-response even when every dose was technically administered. Structured site rotation across the abdomen, upper thighs, buttocks, and upper arms, following a written rotation schedule reviewed at each visit, is a standard and inexpensive countermeasure.
A lipohypertrophic nodule discovered on exam during a "poor responder" visit is one of the more satisfying catches in GH monitoring — the fix is behavioral, not a dose increase.
When a child's growth trajectory unexpectedly flattens, the three questions asked before ordering additional labs are: is the injection log showing gaps, is there a palpable lipohypertrophic nodule at the usual injection site, and has body weight increased faster than the last dose recalculation accounted for.
GH therapy is generally well tolerated over years of use, but longitudinal monitoring exists precisely to catch the uncommon but clinically significant adverse effects early, while also tracking the metric that ultimately justifies the entire treatment course: progress toward an adult height meaningfully closer to genetic potential, followed by a clear, criteria-based decision about when to stop.
Each routine visit includes a brief, structured screen for the adverse effects specifically associated with GH-driven rapid growth and its metabolic effects, alongside the growth measurements:
• Slipped capital femoral epiphysis (SCFE) — new hip, groin, thigh, or referred knee pain, or a limp, prompts hip exam and imaging; rapid growth during GH therapy is a recognized risk factor, particularly in children who are overweight or hypothyroid • Benign intracranial hypertension (pseudotumor cerebri) — new persistent headache, visual disturbance, or papilledema on exam warrants urgent ophthalmologic assessment; onset is typically within the first weeks to months of therapy and resolves with dose reduction or discontinuation • Scoliosis progression — spine exam at each visit, since accelerated growth velocity can unmask or worsen curvature, especially in syndromic diagnoses like Turner or Prader-Willi syndrome where baseline scoliosis risk is already elevated • Glucose intolerance — GH has an anti-insulin, mildly diabetogenic effect; fasting glucose or HbA1c is checked periodically, particularly in children with obesity, a family history of diabetes, or Prader-Willi syndrome • Tonsillar/adenoidal hypertrophy and sleep-disordered breathing — relevant chiefly in Prader-Willi syndrome, where baseline sleep study is often obtained before and after starting therapy
Bone age, reassessed roughly annually, is the clock that ultimately ends therapy: growth plates approaching fusion mean the height ceiling has nearly been reached regardless of continued dosing. Conventional discontinuation criteria are a bone age of approximately 14 years in girls or 16 years in boys, or a height velocity that has fallen below about 2 cm per year despite continued therapy and good adherence — whichever comes first.
Near-adult height outcomes in appropriately monitored GH-deficient children average several centimeters of net gain over predicted untreated adult height, with the largest gains seen in children started earlier, with more severe deficiency, and with sustained good adherence throughout — which is precisely why the preceding four stages of monitoring (accurate baseline prediction, careful first-year titration, IGF-1-guided dosing, and adherence support) compound into the final outcome rather than being separate concerns.
After discontinuation, a final near-adult height measurement is typically confirmed once growth velocity has been under 2 cm/yr for several months, closing out a monitoring relationship that, for many patients, spans the better part of a decade.
Long-term post-marketing cohorts of GH-treated children who completed therapy show a generally reassuring cardiovascular and malignancy safety profile at conventional dosing — the ongoing surveillance described here is precisely what keeps that record reassuring, by catching the individual outlier early rather than relying on population-level safety data alone.