Long-Term Follow-Up Program, Department of Pediatrics, Memorial Sloan-Kettering Cancer Center, New York, New York 10021
Address correspondence and requests for reprints to: Charles Sklar, M.D., Director, Long-Term Follow-Up Program, Department of Pediatrics, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, New York 10021.
![]() |
Introduction |
---|
![]() ![]() ![]() |
---|
In contrast to incidence rates, which have either increased slightly or remained steady for the past 25 yr, mortality rates have decreased dramatically during the same time period. The overall decline in mortality is estimated to have been 40% from 19751995 (1). The current overall 5-yr survival rate for childhood cancers is in excess of 70%; survival rates are currently 80% for children with acute lymphoblastic leukemia and greater than 90% for children and adolescents diagnosed with Hodgkins disease.
What is responsible for these remarkable improvements in survival? Clearly, advances in supportive care (e.g. prevention and treatment of infections, blood, and blood product support) have contributed substantially to the reduced mortality rates. The most striking improvements, however, are due to the changes in therapy that have occurred over the past 2530 yr. This includes the use of combined modality therapies (e.g. the use of surgery combined with chemotherapy and radiation therapy) and the use of aggressive multiagent chemotherapeutic regimens. Data from the Childhood Cancer Survivor Study, a cohort of nearly 14,000 5-yr survivors of childhood cancer from North America, indicate that greater than 85% of current survivors were treated with at least two different modalities (e.g. chemotherapy and radiation therapy), whereas 44% of survivors received a combination of surgery, chemotherapy, and radiation therapy. Thus, the vast majority of the estimated 200,000 survivors of childhood cancer currently residing in the United States will have been exposed to multimodality therapies during the course of their cancer treatment.
What are the long-term consequences of such exposures when they occur during childhood and adolescence? From our experience at Memorial Sloan-Kettering Cancer Center, it seems that approximately two thirds of pediatric cancer survivors will develop some type of medical complication or disability that can be directly attributed to their previous cancer treatment (2). These late complications span a spectrum from minor and treatable (e.g. primary hypothyroidism) to serious and occasionally fatal [e.g. subsequent new malignancies (SNMs)]. The most prevalent late effects of cancer therapy are endocrine disturbances, which have been documented in some 40% of survivors.
GH deficiency is among the most common endocrine disorders observed after therapy for childhood cancer. It can be seen in individuals who suffer from tumors arising in or near the region of the hypothalamus and pituitary (e.g. germinomas, optic nerve gliomas), either as a direct result of the tumor or as a consequence of the surgery required to remove the tumor. More frequently, however, GH deficiency is diagnosed after exposure of the hypothalamus (less commonly after exposure of the pituitary) to high-dose, external beam radiotherapy. Radiation-induced GH deficiency can develop in a host of different settings but most often is seen following whole brain irradiation for acute leukemia or a variety of CNS tumors, most of which originate at a distance from the hypothalamus and pituitary; after localized radiotherapy for sarcomas and carcinomas of the orbit, face, and nasopharynx; or following total body irradiation, which is used as preparative therapy for bone marrow/stem cell transplantation (3).
GH deficiency following radiotherapy is both dose and time dependent. At higher doses (e.g. >30 Gy) of external beam radiation GH deficiency typically develops within 5 yr of treatment, whereas after lower doses (e.g. 1824 Gy) GH deficiency may not become evident for 10 or more years (4). Once established, however, radiation-induced GH deficiency is usually permanent and irreversible. The impact of developing GH deficiency after a cancer diagnosis will vary depending on such factors as age at diagnosis and the presence of other comorbid events such as premature sexual maturation. Untreated, GH deficiency that originates during childhood will cause suboptimal linear growth and a reduced final height, while young adults may experience the varied metabolic derangements (e.g. increased body fat, raised plasma lipids, decreased bone density) and psychosocial abnormalities that have come to be recognized as the Adult GH Deficiency syndrome.
According to data from the National Cooperative Growth Study, the postmarketing surveillance study established by Genentech, Inc., nearly 10% of the 20,000 children receiving their recombinant GH in North America between 1985 and 1994 carried a diagnosis of organic GH deficiency secondary to the diagnosis or treatment of a childhood cancer or neoplasm (5). It is likely that there are at least an equal number of pediatric cancer survivors with treatment-related GH deficiency who could benefit from GH therapy but remain either undiagnosed or diagnosed but untreated.
A nagging question that has haunted all of ussurvivors, their families, and health care providers alikehas been whether or not treating cancer survivors with GH replacement poses any serious risks to their health. Specifically, does treatment with a potent growth-promoting agent such as GH, with its mitogenic and proliferating properties, increase the risk of recurrence of the primary cancer or heighten the risk for the development of SNMs? These are extremely serious issues, given the fact that either disease recurrence or development of a SNM is often associated with a dismal prognosis and a high mortality rate.
The evidence implicating a role for GH in the development of malignant diseases comes from a variety of different quarters, including experimental animal data, limited human studies, particularly the association between acromegaly and the risk of colon cancer (6), and the flurry of recent epidemiological studies that have shown a correlation between high plasma concentrations of the GH-dependent growth factor insulin-like growth factor I (IGF-I) and an increased risk for the common cancers of adulthood (e.g. breast, prostate, colon, and lung) (7). Moreover, in vitro studies have demonstrated that IGF-I acts both as a strong mitogen and as an antiapoptotic agent in a wide variety of cancers.
Whereas previously published reports have failed to detect an increase in the recurrence rate of childhood cancers in survivors treated with GH (8, 9), most of these studies suffer from one, some, or all of the following limitations: 1) single or limited institution study with reduced statistical power due to small sample size; 2) study confined to subjects previously diagnosed with and treated for a CNS tumor, with little or no data on the outcome of survivors diagnosed with other solid tumors of the face/head or acute leukemias; 3) short follow-up time with termination of data capture at the cessation of GH therapy; 4) incomplete ascertainment and lack of validation of outcomes; and 5) lack of appropriate controls. Fortunately, the study by Swerdlow et al. (10), published in this issue of JCEM, overcomes several of these shortcomings and provides very important data that should help allay our fears and anxieties about the risk of disease recurrence in survivors of childhood brain tumors treated with GH replacement therapy.
This is a multi-institutional study from three large pediatric neuro-oncology centers in the United Kingdom that compares the risk of first recurrence between 180 brain tumor survivors who were treated with GH and 891 survivors who did not receive GH. Patients were diagnosed and treated between 1965 and 1996. The predominate tumor types were medulloblastoma and astrocytoma, in keeping with the experience of most pediatric endocrine units that see large numbers of brain tumor survivors. GH-treated survivors were followed for an average of 6.4 yr after initiation of therapy. Of note, the dose of GH used in these patients (0.5 IU/kg/week) is approximately half the dose currently being used in the United States to treat similar patients. This most likely reflects both the era during which many of these patients were treated and the fact that, historically, our European colleagues have tended to use lower doses of GH than are generally used in North America.
The relative risk (RR) of a first recurrence (RR, 0.6) was actually reduced in the GH-treated group compared with those who did not receive GH. Neither duration of GH treatment nor time since initiation of GH treatment was associated with an increased RR of first recurrence. Additionally, the RR of death (RR, 0.5) was also reduced for the GH-treated compared with the nontreated survivors. How do we explain these findings? Is GH replacement therapy, in some way, protective? Most likely, these low RR estimates are due, at least in part, to some inherent selection bias that favored GH treatment in survivors with a better prognosis. Although the investigators attempted to adjust for some of these factors, several important prognostic variables were not accounted for in the analysis. For example, for individuals with a medulloblastoma, both extent of disease (localized vs. metastatic) at diagnosis and volume of postoperative residual tumor are major determinants of outcome but were not included as confounding variables. Similarly, histologic grade is a powerful prognostic feature in patients with an astrocytoma but was not controlled for in the current study. Although it is probable that these data were just not available to the researchers, nonetheless, it is not difficult to imagine that such information could have influenced the treating physicians decision whether or not to use GH.
Despite these minor concerns, the study of Swerdlow et al. (10) has been carefully done and is methodologically sound. The large sample size makes this the most robust study, to date, and allows us to have a high level of confidence in the results. The data are very reassuring and effectively exclude the likelihood that GH treatment of pediatric brain tumor survivors has a major and clinically important adverse effect on the risk of tumor recurrence. This is, indeed, good news for our patients. Hopefully, these new data will help to remove some of the barriers that have existed in the past, namely reluctance on the part of some pediatric (neuro)oncologists to refer these patients to their endocrine colleagues and a hesitancy on the part of some pediatric endocrinologists to treat brain tumor survivors with GH.
We are still left, unfortunately, with several unanswered questions. Due to the paucity of data, there remains considerable uncertainty about the risk of disease recurrence when GH therapy is administered to survivors of pediatric cancers other than brain tumors. Given the differences in tumor biology, presence of receptors for GH and IGF-I, etc. between the various tumor types, it is probably inappropriate and imprudent to extrapolate the data from the experience with GH in survivors of pediatric brain tumors and apply it to survivors of other types of pediatric cancer. Thus, there is a pressing need for well designed studies to determine whether GH therapy raises the risk of disease recurrence for the sizable pool of survivors of other cancers, such as acute leukemia and soft tissue sarcomas of the face/head, who are rendered GH deficient by virtue of their cancer therapy and who will benefit from treatment with GH. Given the relatively small numbers available at any single institution, this will require the cooperation of researchers from a number of centers at a national or even international level. A study assessing the safety and efficacy of GH replacement therapy in pediatric cancer survivors, using the 14,000 participants of the Childhood Cancer Survivor Study, has been completed recently. That study should help resolve some of the concerns about the impact of GH therapy on the risk of recurrence for survivors of a variety of different pediatric cancers.
Similarly, we know virtually nothing about the effect of GH replacement therapy on the risk of SNMs in pediatric cancer survivors. Survivors of childhood cancer are known to be at an increased risk of developing a new cancer later in life; recent data suggest a cumulative estimated risk of SNMs of 3% 20 yr after diagnosis of the primary cancer, which is approximately six times the rate observed for new cancers in the general population after adjustment for age and sex (11). An individuals risk of developing a SNM is determined primarily by his or her cancer treatment (e.g. exposure to radiation and certain chemotherapeutic agents), as well as genetic factors for the small subset of patients who have a genetic predisposition to cancer. Ultimately, it will be important to determine what, if any, interaction exists between GH therapy and these other determinants of SNM. This will be a formidable task, however, given the relatively small number of events and the long latency between completion of cancer treatment and the development of most SNMs.
Finally, what do we know about cancer risk in adults who are receiving GH replacement for either childhood or adult onset GH deficiency? For those previously treated for a neoplasm, primarily pituitary tumors, preliminary studies have failed to show a significantly increased rate of recurrence of the primary tumor, but the follow-up time remains relatively brief (12). In the case of de novo cancers, overall the incidences have not differed markedly compared with population norms, but, again, follow-up time is limited. The much higher background incidence of cancers in adulthood combined with the epidemiological data correlating adult cancer risk with circulating levels of IGF-I mandate that we place a very high priority on obtaining more data on the risks of cancer in adults receiving long-term GH replacement therapy.
Given what we know and do not know about GH therapy and both cancer recurrence and the risk of new cancers, what do we say to patients and their families and how do we decide when and whom to treat? Obviously, there is no simple or single answer to such a broad and complex question. We clearly have a moral and ethical obligation to educate and inform our patients and their families about the issues, as fully and completely as is practical. Also, as physicians we need to counsel our patients and help them weigh both the benefits and the risks of any proposed treatment strategy. In my own work with a large number of pediatric cancer survivors, it has been my experience that for the vast majority of survivors who are candidates for GH therapy, the benefits of treatment are known and generally achievable, whereas the risks of serious problems are probably very low and largely theoretical.
Received October 16, 2000.
Accepted October 19, 2000.
![]() |
References |
---|
![]() ![]() ![]() |
---|
HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |