Off-label use of recombinant IGF-I for growth promotion: Is it warranted?

Promotion of off-label use of recombinant human insulin-like growth factor I began as soon as FDA orphan drug approval was given late in 2005 for its use in growth hormone insensitivity due to growth hormone receptor deficiency or growth hormone inactivating antibodies following recombinant human growth hormone treatment of children with growth hormone gene deletion. This promotion was based on the hypothesis that milder forms of growth hormone insensitivity accounted for a substantial proportion of unexplained short stature.

Questions raised

Arlan L. Rosenbloom

The appellation “primary IGF-I deficiency” was given to this hypothetical condition. However, the novel designation “primary IGF deficiency” was vague, ambiguous and nonspecific and also not used as a diagnostic category in contemporary works, except in reference to patients with IGF-I gene deletion or mutation. Furthermore, the diagnosis of IGF deficiency based on a single IGF-I determination is highly unreliable. Pitfalls of assays and causes of variability among laboratories include variable efficacy of extraction, impurity of the international reference standards, and difference among manufacturers in procedures and matrices for dissolving standards. Further, post-sampling proteolysis of samples occurs. IGF-I concentrations are highly variable with age, pubertal status, illness and nutritional status. Differences among IGF-I values on different days in the same individual range from –36% to 38%. To further compound the problem, many — if not most — children designated as having IGF deficiency are children with constitutional delay in growth and maturation whose IGF-I levels have been inappropriately interpreted for chronologic rather than developmental (bone) age.

The promotion of rhIGF-I to “replace what’s missing” is misleading. The concept ignores the critical importance of local production of IGF-I, which would not be matched by even pharmacologic provision of endocrine IGF-I. This was demonstrated by the inferior growth response to rhIGF-I of patients with GH insensitivity compared with replacement therapy with rhGH in GH deficiency. Also of concern was the potential for IGF-I to suppress GH secretion in children with GH deficiency and compromise the local effects of GH on chondrocyte proliferation and stimulation of paracrine/autocrine IGF-I production.

The approximately 40% of children with idiopathic short stature considered to have IGF-I deficiency respond no differently to rhGH than those with more normal IGF-I levels and, in fact, show an inverse correlation between baseline IGF-I concentration and the first-year growth response — exactly the opposite of what would be expected with IGF-I deficiency due to GH insensitivity. There is also a lack of correlation of growth response to rhGH in ISS relative to peak stimulated GH levels — further evidence against GH insensitivity in these individuals.

The animal experiment widely used in promotional materials to emphasize the centrality of IGF-I in growth is manifestly inappropriate. It states that 35% of growth is regulated by IGF-I alone; 34% by both GH and IGF-I; 14% by GH alone; and 17% unrelated to either, based on targeted gene knockout studies in mice. The fact that 50% (not 17%) of normal human growth is neither IGF-I nor GH dependent makes these mice an inadequate model for understanding human growth. Among individuals with the most dramatic IGF-I deficiency seen thus far, those with IGF-I mutations affecting both endocrine and autocrine/paracrine synthesis had adult heights of 119 cm and 112 cm. If 69% of growth was dependent on IGF-I, their heights would have been less than half of that — an incredible 51.2 cm and 52.7 cm (based on their mean parental target heights).

IGF-I deficiency a viable diagnosis?

Presentations during the past year and the first peer-reviewed report of an open-label trial of rhIGF-I therapy in 120 “short children with low IGF-I levels,” defined as both height and IGF-I standard deviation scores <–2, confirmed these reservations and provided additional concerns about safety.

Patients in the recently published report had growth velocities that were considered adequate to maintain their height SD score — that is, they were growing normally, inconsistent with GH insensitivity or IGF-I deficiency. Levels of circulating IGF-I had to be maintained at +2 SD score for age to obtain growth acceleration, pharmacologic rather than physiologic dosing inconsistent with treatment of a hormonal deficiency state. IGF-I generation test data demonstrated normal GH sensitivity with brisk responses in IGF-I and IGF binding-protein-3 concentrations, which were substantially greater percentage-wise than reference standards. The untreated control patients had increased mean IGF-I concentrations by 54% during the year of observation, such that at the end of the year greater than 50% (12 of 23) had IGF-I SD scores greater than the diagnostic level for “IGF deficiency,” emphasizing the evanescence of this arbitrary diagnosis.

In a recent meeting, evidence was presented that once-daily dosing of rhIGF-I was more effective in some patients than twice-daily dosing, which can only be explained by lack of suppression of GH for half of the day, a further strong indication of normal GH sensitivity.

Another recent presentation described the early results of combination therapy, rhIGF-I plus rhGH, in ISS — an undertaking that acknowledged that these children have normal GH sensitivity.

Is IGF-I growth-promoting in ISS?

Despite much greater increases in circulating IGF-I concentrations in the off-label–treated patients than in those treated for GH insensitivity, the growth increments in the recent report were only one-third to one-half the growth velocity increments of those treated for the approved orphan indication.

Growth prognosis was not enhanced because bone ages advanced more rapidly than chronologic age in treated patients and more slowly than chronologic age in controls, resulting in annual growth velocity per year of bone age for controls of 6.5 cm; for the lower dose rhIGF-I treatment group, 6.4 cm; and for the higher dose group, 6.6 cm.

Another indication that growth prognosis may not have been enhanced is the observation that 14% of the treated patients began puberty during the year they were administered rhIGF-I vs. only 4% of the controls.

Acceptable safety profile?

Because there is no health or life-threatening issue involved nor even evidence that growth promotion enhances quality of life, safety is of paramount importance in considering any treatment of ISS. In addition to the recently published report, data are available from FDA post-marketing adverse event reports.

Adverse events from the published results of rhIGF-I for short stature with low IGF-I included: hypoglycemia in 9% of those administered the lower dose and 20% assigned to the higher dose of rhIGF-I; headache in 3% of low-dose and 41% of high-dose patients; and vomiting in 20% of low-dose and 27% of high-dose patients. Other adverse events included lipohypertrophy, arthralgia and myalgia. Tonsillar or adenoidal hypertrophy occurred in three patients and snoring in four of the 95 patients administered rhIGF-I and also appeared to be dose-related. Treatment of GH insensitivity was associated with a cumulative prevalence of snoring from 4% in the first year, as noted in this study, to 65% after several years. Along with hypoacusis, this led to the need for tonsillectomy–adenoidectomy in 11% of patients.

There were two instances of intracranial hypertension in 95 patient-years of rhIGF-I treatment of ISS, which contrasted markedly with complete absence of this complication in 25,000 patient-years of rhGH treatment of ISS.

Other adverse events reported included a single case of anaphylaxis and one of generalized urticaria. FDA reports noted 10 children hospitalized: one each for retinal hemorrhage, increased intracranial hypertension, depression and hypoglycemia; two for tonsillar hypertrophy; and four for allergic reactions, along with 67 children with adverse events not requiring hospitalization. These included hypersensitivity, abnormal behavior, tonsillar hypertrophy, gynecomastia, hypoglycemia, headache, facial palsy, abnormal appetite, increased hepatic enzymes, diplopia, and hyperglycemia. These reports represented a substantial number of adverse events for the estimated 1,000 children who have been treated off-label with rhIGF-I.

Potential long-term risk

Epidemiologic data in humans have suggested that higher levels of circulating IGF-I are associated with increased risk for certain types of cancer. Overexpression of IGF-I is associated with tumor formation in mice. IGF-I favors cellular proliferation instead of arrest and cellular survival instead of apoptosis. Thus, it is mitogenic and anti-apoptotic. This may result in stepwise accumulation of genetic damage in individuals with higher IGF-I levels because of a slightly higher rate of cell division (increasing the risk for errors) and the reduced probability of apoptosis of cells with a small number of “hits.” Whether or not the groundwork for increased cancer risk is being laid by pharmacologic growth therapy producing high levels of IGF-I in non-GH–deficient or resistant short children cannot be known for decades.

Arlan L. Rosenbloom, MD, is Adjunct Distinguished Service Professor Emeritus in the Department of Pediatrics at the University of Florida College of Medicine, Gainesville, Fla.

For more information:

• Buckway C. J Clin Endocrinol Metab. 2001;86:5176-5183.
• Chernausek SD. J Clin Endocrinol Metab. 2007;92:902-910.
• Guevara-Aguirre J. J Clin Endocrinol Metab.1997;82:629-633.
• Longo VD. Science. 2003;299:1342-1346.
• Lupu F. Dev Biol. 2001;229:141-162.
• Midyett LK. #O1-742. Presented at: Annual Meeting of The Endocrine Society; June 10-13, 2009; Washington, D.C.
• Midyett LK. J Clin Endocrinol Metab. 2010;95:611-619.
• Moore WV. Pediatr Endocrinol Rev. 2008;6:4-7.
• Pollak MN. Nat Rev Cancer. 2004;4:505-518.
• Rosenbloom AL. J Clin Endorinol Metab. 2010;95:505-508.
• Rosenbloom AL. Lancet. 2006;368:612-616.
• Walenkamp MJ. J Clin Endocrinol Metab. 2005;90:2855-2864.
• Woods KA. N Engl J Med. 1996;335:1363-1367.