Discussion
Bisphosphonates are manifestly effective in the treatment of osteoporosis in adults by inhibiting osteoclast-mediated bone resorption. Experience in children with low bone mineral is less extensive but generally supports the same conclusion.19 This is exemplified by a report on a small number (n=22) of children with a variety of chronic illnesses associated with loss of bone mineral, none with malignant disease.31 In a randomized clinical trial (RCT) the participants received either oral alendronate weekly or placebo for one year. Volumetric bone density of the lumbar spine increased significantly in the experimental group but not in those receiving placebo. Likewise, the cross-sectional moment of inertia per unit length of the femoral shaft - an estimate of mechanical strength - increased significantly only in the experimental group.31
A recent report from the Cincinatti Children’s Hospital Medical Center is of a 7 year retrospective chart review describing the outcomes of intravenous bisphosphonate infusions in patients less than 21 years of age.32 Among the patients who were excluded were all of those (N=12) who had received cancer chemotherapy, but 29 of the study sample (N=123) were categorized as having glucocorticoid-induced osteoporosis. In the 42 patients who had at least two DXA examinations available for comparison there was a significant gain in LS aBMD) at one year after bisphosphonate infusion.
In the non-randomized study reported here, involving only children with ALL (n=68) on active therapy who received oral alendronate weekly for a shorter time than in the RCT reported by Rudge et al.,31 LS aBMD increased significantly as measured by Z score with and without HAW correction. The sequential results from DXA examinations are indicative of a short-term gain in LS aBMD from the administration of alendronate, blunting an otherwise downward trajectory. These changes occurred at an age when bone mass is increasing in healthy children, adolescents and young adults. In a population of healthy Caucasian subjects in Ohio, who were studied serially by DXA between the ages of 8-30 years,33 the rate of accumulation of bone mass began to slow in mid to late teens in females (N=343) with peak BMD attained in the early 20s, while in males (N=312) the corresponding rate began to slow in late teens to early 20s with peak BMD attained by mid to late 20s. Comparable Canadian data have been reported.34 Achieving a normal peak bone mass by age 30 is important in the prevention of osteoporosis in older adult life.35 Consequently, rendering the subjects in this study osteopenic from the treatment of ALL, despite a temporary and only partial reprieve by the administration of alendronate, has put them at risk of fragility fractures in later life. In a separate study from our centre,36 LS aBMD during maintenance/continuation therapy of ALL was shown to be predictive of later fractures in children who did not receive a bisphosphonate. The results in the small number of subjects (n=9) reported here, who were studied more than a decade after diagnosis, are intriguing, being suggestive of late improvement in LS aBMD, but longitudinal examination of a larger cohort will be required to evaluate this observation. A recent Dutch-Canadian collaborative study has demonstrated the predictive value of age and weight at diagnosis in relation to LS aBMD and to subsequent development of symptomatic fractures in children with ALL,37complementing our established practice of HAW correction of Z scores for aBMD of the lumbar spine.28
The results of bone morphometry and measures of bone strength, determined by peripheral quantitative computed tomography (pQCT)38 in our study of long term survivors of ALL (more than 10 years from diagnosis), have been reported separately.39 There were no statistically significant differences in any of 19 metrics between those who had received bisphosphonate (n=14, pamidronate in 5) and those who had not (n=58). This may represent adaptive restructuring of trabecular bone by thickening of a reduced number of trabeculae to maintain bone strength.40 Again, the improvement in LS aBMD following discontinuation of long-term steroid therapy has been reported in young adults with sarcoidosis.41
There have been few prior studies of bone by pQCT in children with ALL. Brennan et al. in the UK reported on 53 survivors on average almost 5 years after completion of treatment which did not include cranial radiation.42 There was no deficit in LS aBMD but reduced BMD in the trabecular bone of the distal radius was revealed. More recently, a group of 50 survivors in the US were studied within two years of completing treatment without cranial radiation, and again one year later.43 A large group of healthy subjects afforded the provision of Z scores. Initial deficits were shown in both trabecular and cortical BMD in the tibia. Subsequent changes in the cortical outcomes varied by the duration of the interval since completion of therapy. Interestingly, there were no associations of the outcomes with leukemia risk category, total glucocorticoid dose or antimetabolite therapy.
With respect to the short-term safety of oral alendronate, this has been assessed in detail in randomized clinical trials conducted in young people. In a study of 32 adolescents with anorexia nervosa who received the drug or placebo daily for one year, the authors stated ”We found alendronate to be well tolerated and safe”.44 A more recent trial in 139 children with osteogenesis imperfecta, involving daily oral alendronate for two years, prompted the authors’ observation ”Importantly, ALN was associated with few adverse events. In particular gastrointestinal symptoms were not more common in patients receiving ALN than in those receiving placebo”.45 In a systematic review of bisphosphonates in children and adolescents with secondary osteoporosis Ward and colleagues noted that oral alendronate appeared to be well tolerated for intervals up to 3 years.46 A report of the use of oral alendronate in post-menopausal osteoporosis stated that the drug was efficacious and well tolerated over a 10 year period.47 In our non-randomized study gastrointestinal symptoms were infrequent and minor in degree, with one exception, and no other adverse sequelae were recorded. The retrospective study by Nasomyont and colleagues32 determined that “In 468 patient years of bisphosphonate exposure there were no reports of ONJ (osteonecrosis of the jaw) or AFF (atypical femoral fracture) in medical record.” In a systematic review published in 2020 no cases of ONJ in children were identified.48
The other safety issue addressed in our study is that of the relationship of bisphosphonates to the risk of cancer, in this instance relapse of ALL. We are not aware of this having been studied previously in children. By contrast, it has been a topic of considerable interest in the context of adults. Pre-clinical data and some, but not all, clinical trials suggest an anti-neoplastic effect of bisphosphonates.49 A recent study using murine models of ALL showed a significantly shorter survival in animals receiving zoledronic acid, a third generation bisphosphonate, as well as chemotherapy compared to those receiving chemotherapy alone, but this was not seen using xenografts of human B precursor and T cell ALL.50
A large study in the UK, using data extracted from the General Practice Research Database, examined 41,826 subjects who had been exposed to oral bisphosphonates and an equal number of matched controls,51 investigating the risk of developing cancer overall as well as specific neoplasms. It was reported that there was a significant protective effect in general and specifically in relation to breast and colorectal cancers. No such effect was evident for leukemias and lymphomas, but in no instance was the risk of cancer increased. In comparison, the results of a population-based cohort study in South Korea, using the National Health Insurance Services database, showed no association between exposure to oral alendronate or residronate and the risk of breast, ovarian or cervical cancer.52 This study involved 14,847 users and 204,525 unmatched controls.
A systematic literature review of the prognostic effect of bisphosphonate exposure in adult patients with a solid tumor found improved overall survival as well as cancer-specific and recurrence-free survival.53 For individual cancer types there was an apparent benefit in overall survival for patients with gastro-esophageal cancer and longer cancer-specific survival in those with breast cancer. In a separate study, exposure to alendronate was associated with a lower risk of bone metastases in osteoporotic women with breast cancer.54
Against this background our finding that there was no increased frequency of relapse in children receiving oral alendronate while undergoing active therapy for ALL may offer a measure of reassurance. Whether there was a protective effect is intriguing – the proportion of patients who relapsed after receiving alendronate was less than 50% of that in those who did not receive the drug, and more patients who received alendronate had high risk disease - but cannot be addressed with confidence. However, in a recent report,55 B cell ALL cells were shown to mediate resorption of trabecular bone by a RANK-RANKL (receptor activator of nuclear factor kB ligand) mechanism through activation of osteoclasts. The authors postulated that combining an anti-resorptive agent with chemotherapy may reduce the risk of relapse by disrupting this B cell activity.
This was a retrospective cohort study with a small number of patients. Other limitations are the variability in the duration of alendronate administration, although this reflects variation in the timing of responsiveness to this medication, and the lack of reliable information on fractures. A much larger longitudinal study with well-matched controls, perhaps in the form of a randomized trial to examine the efficacy of a bisphosphonate in osteopenic children with ALL, would be valuable.
Conflict of Interest statement. The authors have no conflicts of interest to declare
Acknowledgements. We acknowledge the contributions of Jo-Ann Fowler BScN and Sarah Beasley BScN to the collection of data from clinical records.
Figure legend. Sequential lumbar spine bone mineral density Z scores in children and adolescents with acute lymphoblastic leukemia who received alendronate.