ESPE Yearbook of Paediatric Endocrinology

Brief Summary: This single-center, retrospective cohort study compared the final adult height (FAH) of 92 male childhood cancer survivors (CCS) with growth hormone deficiency GHD) treated with growth hormone alone (monotherapy) or in combination with an aromatase inhibitor. The addition of AI to GH therapy did not improve FAH.

This study from the Children’s Hospital of Philadelphia is the most extensive study on the role of AI associated with GH in improving final height of CCS with growth hormone deficiency. A few previous studies had reported a slight benefit of AI in augmenting predicted adult height in boys with short stature.

70 CCS received GH monotherapy and 22 received GH plus aromatase inhibitors (GH+AI). There were no baseline differences in age at GH start, height Z-score, mid-parental Z-score, pubertal maturation (73% were prepubertal at initiation in GH group vs 50% in GH+AI group). But median age at cancer diagnosis was lower in the GH+AI group (3.2 vs 5.5 years). The most common diagnosis was neuroblastoma in the GH+AI group and CNS tumor in the GH group. More patients in the AI+GH group received stem cell transplantation, abdominal radiation, total body irradiation, and cis-retinoic acid. Mean FAH and FAH z-score were similar in the two groups (FAH 161 cm in GH+AI group vs 167 cm in GH group; FAH z-score -2.1 in AI/GH group vs -1.3 in GH group).

Increase in height z-score was higher in the GH group than the GH+AI group (mean 0.63 vs -0.30). However, the difference between FAH and MPH was higher in AI+GH group (median -19.1 vs -7.3 cm). Lower FAH z-score was associated with spinal radiation, lower height z-score at therapy start and greater difference between bone age and chronological age. No side effects were reported, including differences in glycated hemoglobin pre- and post-AI treatment.

Study limitations include the retrospective design, the small sample size, and differences in cancer treatments between the GH and GH+AI groups. Furthermore, selection bias is very likely, as AI was more likely indicated in patients with severe short stature and worse predicted final height. More patients in the GH+AI group received spinal radiation or cis-retinoic acid, which are known to negatively affect linear growth. Larger studies are needed to evaluate predicted height variation before and after AI and final adult height. At the same time, the AI side effect profile must be carefully considered in this population with a well-known increased risk of impaired glucose metabolism, cardiovascular events, hepatotoxicity and osteoporosis.