Elizabeth A. Schriock

A method of predicting adult height and obesity in long-term survivors of childhood acute lymphoblastic leukemia.

PURPOSE Short stature and obesity have been reported among long-term survivors of childhood acute lymphocytic leukemia (ALL). We examined factors that contribute to these adverse sequelae. PATIENTS AND METHODS Serial height and weight measurements were analyzed for 91 long-term survivors who were treated for ALL between 1967 and 1975 at a single institution. These patients were all younger than 12 years at diagnosis, were in continuous complete remission, had reached final height, and had height and weight measurements within 1 year of age 18 years. They had received craniospinal (n = 33) or cranial irradiation (n = 58) to total doses of 24 Gy as CNS prophylaxis. Standard deviation scores (SDS) were used to reflect the deviation of height and weight measurements from population means, and the body mass index (BMI; weight divided by height squared) was used in assessing obesity at age 18 years. RESULTS Short stature (less than fifth percentile) was seen in 41 patients (45%), and obesity (BMI greater than or equal to 24 kg/m2) in 35 (38%). Regression formulae were developed that explain 65% and 62% of the variability in patient height and BMI, respectively. CONCLUSIONS Risk factors were identified for abnormally short stature, which was defined to be a decrease of 1.5 SDS in height from diagnosis to age 18 years. These factors include younger age and above-average height for age at diagnosis (height SDS greater than 0), craniospinal irradiation, and greater decrease in height SDS during antileukemic therapy. Risk factors for obesity at age 18 years include weight SDS greater than 0 and greater than height SDS at 1 year after the end of chemotherapy.

Studies with Growth Hormone-Releasing Factor (GRF) in the Human

In 1982 Guillemin et al. (1) isolated, characterized, and synthesized an amidated, 44-amino-acid growth hormone-releasing factor (GRF44) from the pancreatic tumor of a patient with acromegaly. GRF44 and two shorter, homologous peptides that lack a C-terminal amide, GRF37 and GRF40, stimulated GH release in cultured pituitary cells and in vivo. (1) An identical 40-residue peptide was isolated by Rivier et al. (2) from the tumor of a second acromegalic patient. There is increasing evidence that these tumor-derived growth hormone-releasing factors (GRF) are identical to human hypothalamic GRF and are encoded by a single GRF gene.(3–6) The availability of synthetic GRF provides an opportunity to define further the control and dynamics of GH secretion in the human in health and disease.

RESPONSE OF GROWTH HORMONE DEFICIENT PATIENTS TO SYNTHETIC GROWTH HORMONE-RELEASING FACTOR

We studied the growth hormone (GH) response to synthetic growth hormone-releasing factor (GRF (1-44)-NH2)† in non-GH deficient (NGHD) subjects (4-33 yrs) and GH deficient (GHD) patients (5-24 yrs). The mean (±SE) peak GH response of 6 NGHD children after 5 μg/kg IV of GRF (27.1±5.8 ng/ml) was similar to that of NGHD young men previously reported (JCEM 57:677). Of 20 patients with severe GH deficiency 17 responded to 5 μg/kg IV of GRF, but the mean peak plasma GH concentration was less than that of the NGHD group (5.0±1.2 vs 27.2±3.5 ng/ml; p<0.001). Patients with isolated GH deficiency had responses similar to those with multiple pituitary hormone deficiencies. The GH responses of the GHD children correlated negatively with chronologic age (r= −0.758, p<0.02). Six partially GHD children had a higher GH response to GRF than severely GHD children (13.1±1.8 vs 6.9±1.7 ng/ml; p<0.04) but lower than NGHD children (p<0.05). GRF induced higher plasma GH levels in GHD than did arginine, insulin, or L-dopa.The GRF test is useful for assessing GH reserve. Although a GH response to GRF in GHD patients suggests GRF deficiency, the lack of response does not exclude it. The GH response to synthetic GRF in many GHD children is consistent with a hypothalamic abnormality as the etiology of their GH deficiency and supports the potential therapeutic usefulness of GRF or an analog in these patients.

LONGITUDINAL STUDY OF GROWTH IN CHILDREN TREATED FOR ACUTE LYMPHOBLASTIC LEUKEMIA

To evaluate growth retardation and obesity in relation to treatment of ALL children in initial remission (7.3 to 17.5 yrs), the heights (HTs) and weights (WTs) of 74 boys and 83 girls diagnosed prior to 1976 were analyzed. Patients were treated with chemotherapy and cranial or craniospinal irradiation (2400 rads) after diagnosis at 0.2 to 16.6 yrs of age (median 4.4). HT and WT are expressed as SD from the mean of the normal population by age and sex (NCHS standards). A decrease in HT SD occurred between diagnosis and last HT recorded for boys (mean ± 1 SD: O.1±1.0 vs -0.9 ±1.2, p<0.001, Wilcoxon) and for girls (-0.1±1.1 vs -1.4±1.2, p<0.001). Boys lost -0.7±0.4 SD in HT during therapy; -0.1±0.6 SD from the end of therapy until age 11.6 yrs; and -0.3±0.7 SD thereafter. Girls lost -0.7±0.6 SD in HT during therapy; 0.0±0.6 SD from end of treatment until 2 yrs prior to menarche, and -0.7±0.9 SD thereafter. 35% of boys and 48% of girls lost ≥1.5 SD in HT during the study period. As an index of appropriateness of WT for HT, HT SD was subtracted from WT SD. For boys, last (WT-HT) SD was 1.3±1.3 compared to -0.2±0.9 at diagnosis; 76.2% had a final positive index of obesity. For girls, last (WT-HT) SD was 1.3±1.2 compared to - 0.2±0.8 SD at diagnosis: 85.7% had a final positive index. Last (WT-HT) SD inversely correlated to last HT SD among girls (r= -0.45, p<0.001 ) but not boys. Additional girls receiving 1800 rads (20) or no rads (23) were compared to girls receiving 2400 rads for change in HT during treatment and a difference was found (-0.4±0.6 vs -0.3±0.6 vs -0.7±0.6, respectively, p<0.02, ANOVA); no difference was found with similar groups of\boys. In conclusion, many long-term survivors of ALL develop linear growth retardation and obesity. Much of the height loss occurs during therapy and may be related to radiation; an adequate “catch-up” phase of growth does not occur.