Bone Mineral Status During and After Therapy of Childhood Cancer: An Increasing Population with Multiple Risk Factors for Impaired Bone Health

With survival rates for childhood cancers now approaching 70%,1, 2 there is an urgent need to assess the nature, prevalence, and risk factors for long-term treatment sequelae. Survival rates for pediatric malignancies increased from 20% in 1965 to 66% in 1995.3-5 As a result, it is anticipated that by the year 2000, 1 in every 900 individuals between the ages of 15 and 45 years will be a survivor of childhood cancer.5, 6 This growing population of long-term childhood cancer survivors represents a new group at risk for early onset osteoporosis. Adverse late effects resulting from impaired bone mineral density (BMD) (i.e., increased risk of early onset osteoporosis, fracture, and other bone defects) have been observed in cohorts of childhood cancer survivors of several primary tumor types.7-14 The frequency of diminished lumbar spine BMD has ranged from 8–23% of survivors who received chemotherapy prior to physeal closure.7, 8, 15 The wide range of reported frequencies probably reflects the study of relatively small numbers of patients with diverse primary tumor types and treatments. The imprecise information about the prevalence, risk factors for, and treatment approaches for diminished BMD in pediatric patients further emphasizes the need for prospective study in these patients. Cancer therapy in children and adolescents, particularly when combined with suboptimal nutritional status and inactivity, can impair bone mass accretion. Metabolic changes induced by cancer therapy may compound the physiologic changes associated with growth, pubertal development, and aging and result in premature onset and/or increased prevalence or severity of osteoporotic complications in childhood cancer survivors.7, 16-19 Bone mineral accumulation normally occurs during the period corresponding to the onset of many childhood cancers. The bulk of peak bone mass is usually achieved during puberty, with little if any subsequent accretion.20-24 BMD is known to be influenced by genetic factors25-29 and by factors that can be affected by cancer therapy, including physical activity levels,19, 24, 30 nutritional status,19, 31-34 and age at onset of puberty.19, 24, 35-37 The relative contribution of these factors to diminished BMD in childhood cancer survivors has not been elucidated. However, it is known that untreated diminished BMD predisposes to the early development of osteoporosis and related bone defects. Thus, treatment of childhood cancer survivors who have decrements in BMD might offset the osteoporotic effects of therapy and reduce health risks associated with osteoporosis observed during aging. The study of bone mineral status in children with cancer reported in this month's JBMR by Arikoski and colleagues emphasizes the importance of prospective evaluations of childhood cancer survivor cohorts in elucidating the true incidence, timing of onset, and clinical risk factors for the development of deficiencies in bone mineralization. The authors confirm by careful diagnostic imaging and biochemical evaluations that childhood cancer patients experience a negative balance in bone turnover which may predispose to impaired development of peak bone mass. The reported study focused on the early phases of treatment and have demonstrated that at baseline prior to cancer therapy, biochemical markers of bone formation are diminished and remain low during therapy. The findings are similar to those of Atkinson et al. who recently evaluated bone metabolism in 40 children diagnosed and treated for acute lymphocytic leukemia (ALL). At diagnosis, more than 70% had abnormalities of bone metabolism markers at diagnosis. During treatment, bone mineral content decreased significantly, implicating the multifactorial influence provided by the leukemic process and treatment.38 The relatively brief longitudinal follow-up in the study by Arikoski et al. (74% of cohort evaluated were still receiving cancer therapy at 6 months) limits conclusions about the frequency and severity of bone mineral deficits persisting after completion and recovery from the acute effects of cancer therapy. Other researchers have shown that alteration in bone metabolism markers and growth deficits during therapy may demonstrate catch-up recovery.39-41 However, Henderson et al. found the contrary to be true, i.e., at 1 year from completion of chemotherapy, recovery of BMD was not observed.42 Thus, further investigations are clearly warranted to clarify these aspects. The significance of marginal deficits in bone mineralization that may be exacerbated by late-onset treatment sequelae, e.g., pituitary hormone dysfunction, or the physiologic changes anticipated with pubertal development and aging, requires long-term (i.e., beyond 5 years) follow-up in order to be properly evaluated. The finding of asymptomatic vertebral fractures on radiographic evaluation of long-term ALL survivors studied at our institution confirms the potential for significant sequelae related to osteopenia. Studies examining larger numbers of patients, at diagnosis, during therapy, and after therapy, are badly needed to more clearly define the incidence and impact of diminished BMD in a long-term cancer survivor. The actual prevalence of risk factors for osteopenia in patients treated with contemporary therapy remains unclear. In two previous studies, severe abnormalities in BMD (more than 2 SD below the age- and gender-appropriate mean) were seen exclusively in patients treated with cranial irradiation.7, 9 However, others have observed that BMD deficits are not confined to patients treated with cranial irradiation. Investigations by Boot et al. implicated corticosteroids and cytostatic therapy as contributing to diminished BMD in ALL survivors.12 Identification of risk factors for diminished BMD would facilitate effective post-treatment follow-up and planning for risk-adapted interventions. Further, the ability to prospectively identify childhood cancer patients at risk for decrements in BMD might permit the early institution of preventive measures that may even be initiated during chemotherapy. In the general population, BMD is largely genetically determined but can be influenced by other factors, including diet and physical activity.22, 43, 44 Birth weight,45 menstrual history,46 oral contraceptive use,47 smoking,48 and alcohol use46 may all modulate BMD; collecting information on these covariates is important. A prior history of fracture has been shown to be predictive of future fractures49; therefore, inquiry concerning fracture history is mandatory. However, to date, the contributions of these many factors to diminished BMD in populations other than postmenopausal women are not known. Genetic predisposition influences BMD in “normal” populations. It has been estimated that as much as 80% of peak bone mass is genetically determined,50 with the remainder being environmentally controlled. In pre- and postmenopausal women, a vitamin D receptor (VDR) allele is associated with lower BMD and more rapid loss of bone mineral.27, 28, 50-54 In this population, the VDR genotype is the strongest predictor of lumbar spine and femoral neck bone density, after controlling for height and weight.27, 28, 50-52 The b allele is associated with a higher bone density and has a codominant effect.50 An excess representation of the BB genotype (p < 0.001) was seen in postmenopausal patients whose bone density was more than 2 SD below the young normal mean. Further 74% of women studied with the BB genotype had bone density below the fracture threshold, whereas 61% with the bb genotype were above the fracture threshold.50 Calcium supplementation (500 mg/day) for a 2-year period preserved bone mineralization in all genotypic groups and counteracted the differences between genotype.28 Systematic data on the influence of genotypic determination on BMD and on response to therapeutic interventions designed to increase BMD are as yet unavailable. Information on family history that may indirectly reflect genetic predisposition to osteoporosis is also important. Recently, the National Osteoporosis Foundation recognized maternal history of hip fracture as a strong risk factor for diminished BMD.5 Given the age of our population, we are collecting information on grandparents' history of hip fractures and evaluating VDR genotype to determine the genetic contribution to bone mineral deficits in childhood cancer survivors. Antimetabolites and glucocorticoids—key components in the treatment of childhood ALL—have been implicated in detrimental effects on bone metabolism including methotrexate,13, 18, 55-63 glucocorticoids,59, 60, 64-67 cyclosporine,55, 59, 60, 64, 68-70 and ifosfamide.61, 62, 71-74 Methotrexate is the backbone of antimetabolite-based therapy during post-remission (continuation) therapy for ALL. The mechanism by which methotrexate induces bone mineral loss is unknown. An increased excretion of urinary and fecal calcium in children treated with methotrexate suggests that increased bone resorption and excretion, rather than diminished bone formation, may be the causal mechanism.58, 61, 63 Glucocorticoids are also routinely included in clinical trials for ALL. Children, like postmenopausal women, are particularly at risk for steroid-induced osteoporosis because of their more rapid bone turnover.36, 66 The mechanisms of glucocorticoid-induced osteoporosis are multiple, with the greatest impact observed on the vertebral trabecular bone.75 Low bone mass (or BMD) can be determined many years in advance of fracture, at a time when preventive measures may be effective.76 BMD accounts for 75–85% of the variance in bone strength and can be determined only by direct measurement.77 Prospective studies have shown that a decrease in bone mass of 1 SD is associated with a 1.5-to 3-fold increase in the relative risk of fracture.78 dual-energy X-ray absorptiometry is widely available for determination of BMD. The cost and sparse availability of quantitative computerized tomography has limited its acceptance as a screening tool for BMD. Bone density determination by broadband ultrasound20, 31, 32, 78, 79 and quantitative magnetic resonance imaging and magnetic resonance microscopy techniques79 show promise but are not yet fully developed for clinical use, especially for pediatric imaging. In the paper by Arikoski et al.,80 the authors define an impairment of the development of BMD during chemotherapy. In their analysis of 46 newly diagnosed children (aged 2.9–16.0 years), areal bone mineral density (BMD areal, gm/cm2) of both lumbar spine and left femoral neck were measured by dual-energy X-ray absorptiometry at baseline and post 6 months of therapy. Dual-energy X-ray absorptiometry provides the advantage of ready availability, affordability, and a low level of ionizing radiation, useful for pediatric imaging. Studies of osteoporotic late effects and evaluation of promising treatment regimens in childhood cancer survivors could reduce cancer-related morbidity. To date, information on overt complications resulting from diminished BMD in long-term survivors of ALL and other childhood cancers is largely anecdotal. Because significant progress in developing curative therapy was not seen until the late 1970s and early 1980s, follow-up of childhood cancer survivors into adulthood is available for only a relatively small population. Further, late effects of therapy may not yet be apparent in survivors studied in the prepubertal period.80 Hence, the prevalence, severity, and implications of this adverse treatment effect remain largely unknown. Also unclear are the relative contributions of genetic predisposition, treatment effects, altered nutritional status, and decreased physical activity to diminished BMD in this population. Elucidation of this information is critical to the development of effective, comprehensive follow-up monitoring and care after curative treatment for pediatric malignancies. Longitudinal evaluations of larger, well-characterized patient cohorts are required to improve understanding of the relationship of treatment factors, physiologic conditions, genetic constitution, and health habits to adverse effects on bone metabolism. This information is essential to determine the appropriate timing for the institution of prophylactic or corrective interventions in relationship to cancer treatment and to determine the at-risk patients to target for these interventions. Further refinement of techniques to optimize quantitation of BMD in children, expansion of normative databases, and increased availability (e.g., quantitative computerized tomography, broadband ultrasound, quantitative magnetic resonance imaging, and magnetic resonance microscopy techniques79) should provide reasonably priced sensitive and specific phenotyping on which to base clinical treatment. Future studies may identify interventions that hold promise for decreasing the adverse late effects faced by this population. Findings in this patient group can then be used to develop cost-effective and quality follow-up screening and intervention trials for long-term survivors of other childhood cancers. The overall goal is to reduce cancer-related morbidity from treatment-induced bone loss in the increasing population of adolescents and young adults who have been successfully treated for pediatric malignancies. This work was supported by Pediatric Pharmacology Research Units, Leukemia Program Project Grant CA20180, Cancer Center Support Grant CA21765, by the American Lebanese Syrian Associated Charities (ALSAC), and by the Crippled Childrens Foundation Research Center.