Growth hormone treatment and neoplasia-coincidence or consequence?

The discovery that GH action in primates was species specific and the ability to extract GH from human pituitaries ushered in the era of GH therapy for humans in the 1950s. At the outset, these were investigational programs facilitated by a centralized and usually government-sponsored distribution program that also enabled monitoring of data on beneficial or adverse effects. Recombinant DNA-derived GH became available in 1985, shortly after human-derived GH was discontinued because of the occurrence of CreutzfeldtJakob disease in several recipients from different countries. With the introduction of recombinant DNA-derived GH, the indications for GH replacement therapy were broadened to include Turner’s syndrome, chronic renal failure, PraderWilli syndrome, and small-for-gestational-age, in addition to established GH deficiency in children. However, monitoring of potential adverse effects was meticulously continued, facilitated by scrupulously maintained, and continually enlarging, databases involving both government and pharmaceutical companies in concert with pediatric investigators. The Lawson Wilkins Pediatric Endocrine Society (LWPES) has closely and carefully monitored reports of adverse events via its Drug and Therapeutics Committee, which regularly informed the LWPES membership (1, 2). Early reports of an increase in childhood leukemia following GH therapy proved to be unfounded when the growing international databases were carefully monitored. Existing evidence indicates that GH treatment does NOT increase tumor recurrence in those whose primary lesion has been successfully treated. The LWPES has also endorsed the prudent policy of waiting for 1 yr after the completion of tumor treatment with no further evidence of tumor recurrence or growth before starting GH treatment in these children. Because a person treated for a malignancy is at risk for a second malignancy, ongoing surveillance of such persons is recommended, especially in those with genetic syndromes that are known to confer an increased risk of malignancy (Down’s syndrome, Bloom’s syndrome, Fanconi’s anemia, neurofibromatosis-1). These recommendations also are part of the recent consensus conference by the Growth Hormone Research Society, which the LWPES has endorsed (3). This careful monitoring by LWPES contributes to the remarkable record of safety and efficacy of GH treatment for recommended indications in children and adolescents. Although adverse events are known to occur (raised intracranial pressure, edema, slipped capital femoral epiphyses, hyperglycemia, gynecomastia), an increased risk of cancer has not been an issue in pediatric patients treated with GH (2). The article by Swerdlow et al. (4), in Lancet, reports on the risk of cancer in patients treated with human pituitary GH in the United Kingdom from 1959 to 1985. The cohort consisted of 1848 patients in the United Kingdom treated in childhood and early adulthood with human pituitary GH and followed for cancer incidence through December 1995 and for mortality through December 2000. Risks of cancer were compared with that in the general population after controlling for age, sex, and calendar period. Among this cohort, risks of mortality from cancer overall was increased approximately 3-fold and from colorectal cancer and Hodgkin’s disease approximately 11-fold. The incidence of colorectal cancer was increased approximately 8-fold, and both incidence and mortality of colorectal cancer as well as that of Hodgkin’s disease were increased even after excluding patients whose original diagnosis gave them a higher risk of cancer. These results are potentially important but clearly preliminary, as the authors themselves point out. Only two deaths each of colorectal cancer and Hodgkin’s disease were reported. Although highly more prevalent than expected in the control population, a chance occurrence of one case or of one death each from colon cancer or Hodgkin’s disease may have skewed the results. It would also be essential to know whether the incidence of such malignancies is increased in patients with GH deficiency who did not receive GH therapy. Are the incidence rates of patients with GH deficiency who are not treated with GH comparable with the matched general population? The relatively small cohort and small number of cases argue strongly for caution in the rush to judgment or conclusion. Similar concerns with leukemia proved unfounded when larger cohorts were carefully examined; some cases of leukemia actually preceded the institution of GH therapy. Failure to treat GH deficiency is itself associated with premature mortality, especially from cardiovascular causes (4). Despite these caveats, the authors are to be applauded for bringing this issue to scrutiny. They have carefully analyzed their cohort and cautiously presented their inferences and conclusions. They provide a potential rational basis through the known effects of IGF-I on promoting cell growth and inhibiting apoptosis. They point out reported associations between status of GH excess such as acromegaly and gastrointestinal cancer, although they do not mention several reports to the contrary (5). They infer that former dosing schedules with human-derived GH may have resulted in higher IGF-I levels than are likely with modern regimens, although they cannot substantiate this assertion, nor is it likely based on GH purity, potency, and dosages used in the 1959–1985 era (6–8). The available data in pediatrics do not support any concern for excess malignancy at this time. In the two largest international databases and surveillance studies, with a total of some 86,000 patients on GH, representing almost 250,000 GH treatment years, there is only one report of a gastrointestinal carcinoma—an adenocarcinoma in a 15-yr-old girl initially treated by radiotherapy and chemotherapy for a 0013-7227/02/$15.00/0 The Journal of Clinical Endocrinology & Metabolism 87(12):5351–5352 Printed in U.S.A. Copyright © 2002 by The Endocrine Society doi: 10.1210/jc.2002-021467