Longitudinal growth retardation in a prepuberal girl with chronic myeloid leukemia on long-term treatment with imatinib

Recently Jonsson et al. reported on increased cortical bone mineralization as a side effect in imatinib treated adult patients with chronic myelogenous leukemia (CML).1 Also for pediatric CML targeted therapy by this tyrosine kinase inhibitor has been replacing stem cell transplantation (SCT) as front-line treatment.2 Thus, although CML is rare in childhood, the number of children treated by imatinib is constantly growing. However, side effects of imatinib on the immature skeleton might differ from adults. We here report on massive growth retardation while on imatinib treatment.

The patient was born as one of triplets at the 34 week of gestation (weight: 2540 g; length: 47 cm). Her further development was uneventful until the age of 5½ years when Philadelphia CML was diagnosed. Treatment with imatinib (Glivec) starting in October 2005 was tolerated without specific side effects. Regular follow-up examinations confirmed ongoing remission and at 40 months on treatment the BCR-ABL rearrangement was undetectable. However, growth retardation had been observed already six months after treatment with imatinib had been initiated. While at diagnosis her body height was equivalent to the 74 percentile it fell to the 9 percentile after three years of treatment (Figure 1). Regular monitoring of bone biochemical markers exhibited increased excretion of urinary Ca, pyridinoline (PYR) and desoxypyridinoline (DPYR), reduced serum N-terminal propeptide of type I collagen (PINP) and increased serum osteocalcin (Table 1). X-ray of the left hand demonstrated skeletal age retardation of 22 months at the age of eight years. Further diagnostics showed partial growth hormone (GH) deficiency (IGF1: 43 ng/mL and 60 ng/mL at the age of seven years and eight years, respectively: normal values: 64–345 ng/mL). No other drugs beside imatinib had been taken by the patient. The two heterozygous triplet siblings showed normal hematologic parameters and their growth followed the 82 and 55 percentiles, respectively.

Also in children, rapid control of Philadelphia CML has been achieved by imatinib.6 However, in adults side-effects affecting bone metabolism were described in 2006 hypothesizing that unspecific inhibition of tyrosine kinases (c-KIT, PDGF-α, -β, c-FMS) which are either expressed by osteoblasts and osteoclasts may impair bone remodeling.7 This mechanism was demonstrated to be active in adult mice.8 Further reports providing evidence that imatinib treatment does affect bone metabolism followed rapidly.9,10 Also in a model of juvenile mice, long-term imatinib treatment had an anti-resorptive effect on osteoclasts and impaired the length of tubular bone. These effects on the growing skeleton were more pronounced in prepubertal animals.11

Table 1.Analysis of biochemical markers of bone metabolism. (N) depicts normal values while pathological values are indicated by (↑) if increased, and by (↓) if decreased.

Mariani et al. observed a 9-year old boy who developed impaired growth shortly after the start of imatinib treatment which resolved at the onset of puberty resulting in no deficiency of his prospective final height.12 GH deficiency was excluded in this case but endocrinological abnormality in the form of a reduced inhibin B/FSH ratio was present. However, in contrast to other reports on findings in adults and animal data he exhibited reduced bone mineral density of his lumbar vertebrae as determined by DXA following six years of imatinib treatment.

Skeletal growth impairment is of major concern in pediatric patients. In the girl described here cumulative data of biomarkers of mineral and skeletal homeostasis showed a slight increase in bone turnover as the plasma osteocalcin level was elevated while bone formation was reduced as indicated by lowered plasma PINP. Urinary Ca excretion was increased as were the urinary concentrations of the collagen cross-links PYR and DPYR which are indicative of bone resorption. These alterations may not be attributed to the partial GH deficiency as in children with this disorder urinary concentration of PYR and DPYR is lower than in healthy controls.13 Thus, these data suggest a prevalence of bone resorption over formation by which skeletal growth is affected resulting in the impressive cutting score of the length percentiles noticed. It is noteworthy that the growth of her two triplet siblings was unaffected. Whether catch-up growth at the onset of puberty as reported by Marani et al. will also occur in the girl reported here has to be studied in the future as she is still prepubertal.9 The underlying partial GH-deficiency may represent a major obstacle to this hope. However, without standardized testing random GH measurements may be inconclusive.

Figure 1.Decrease in body height under imatinib therapy (percentiles according to Hesse, Jaeger, Vogel et al. 1997).3

As the number of children with CML on long-term imatinib treatment is constantly increasing, so far unobserved and/or unreported side effects of this small molecule inhibitor may play a more important role in the future. As SCT may also result in growth impairment, the choice of the optimal treatment for pediatric CML remains an ongoing challenge.

Footnotes

• HS and BAS J contributed equally to this work.