Growth effects of methylphenidate among childhood cancer survivors: A 12-month case-matched open-label study

Abstract

Background

To investigate the effect of stimulant medication [methylphenidate (MPH)] on growth patterns among survivors of childhood cancer (acute lymphoblastic leukemia or brain tumor).

Procedure

Using a case-matched comparison design, childhood cancer survivors participating in a 12-month open-label MPH trial (n = 51) were compared with childhood cancer survivors not taking MPH (n = 51). Measures of body mass index (BMI), height, and weight were obtained at hospital visits and corrected for gender and age using Centers for Disease Control normative data.

Results

Significant deceleration of BMI and weight, but not height, was observed during the 12-month MPH trial for those children taking MPH.

Conclusions

Childhood cancer survivors taking MPH experience significant, though modest, deceleration of BMI and weight across the first year of MPH intervention. The absence of height deceleration, and the presence of only modest BMI and weight deceleration, suggests that MPH is reasonably well tolerated by childhood cancer survivors with respect to growth. Such findings are encouraging in light of increasing evidence that MPH mitigates some of the cognitive late-effects of cancer treatments. Nevertheless, on a case-by-case basis, clinicians should balance the intended benefits of MPH with potential growth effects in this vulnerable population. Pediatr Blood Cancer 2009;52:39–43. © 2008 Wiley-Liss, Inc.

INTRODUCTION

Children surviving acute lymphoblastic leukemia (ALL) and brain tumors (BT) are at significant risk for attention and learning problems 1–4. The Attention-Deficit/Hyperactivity Disorder (ADHD) 5 literature supports the therapeutic effectiveness of stimulant medications targeting attention and behavior symptoms 6–8 and recent research findings demonstrate the efficacy of stimulant medications in addressing untoward cognitive sequelae in childhood cancer survivors 9–11. Unfortunately, stimulant medications have also been associated with some negative side effects, including height and weight deceleration that may be of importance to childhood cancer patients 12–14.

Potential growth deceleration is of particular concern for the childhood cancer population given the already known risk for height deceleration secondary to disease and treatment related factors such as cancer burden in the central nervous system (CNS), concomitant endocrinopathies, chemotherapy, and radiation therapy—especially higher doses directed toward the hypothalamus-pituitary axis and spine 15–17. Findings of growth deceleration in the ADHD population cannot be generalized to childhood cancer survivors given that cancer survivors may respond differently to stimulant medications secondary to compromised neurological status and related endocrine changes 18. While growth changes associated with stimulant medications are relatively well established in the ADHD literature, we are not aware of studies of stimulant medication effects on growth in survivors of childhood cancer.

The main objective of this study is to evaluate changes in body mass index (BMI), height, and weight among childhood cancer survivors who participated in a 12-month open-label methylphenidate (MPH) trial. Our hypothesis was that MPH treatment would result in growth deceleration in our sample of childhood cancer survivors compared to a case-matched comparison group of cancer survivors not treated with MPH.

METHODS

Patients

The present study is part of a multiphase, multisite MPH trial in childhood cancer survivors for which participant selection criteria have been previously described 9–11. Study participants were treated for ALL or malignant BT with chemotherapy and/or CNS-directed radiation therapy, completed treatment at least 12 months prior to study enrollment with no evidence of recurrent disease, were between 6 and 18 years of age and were primary English speakers. Exclusion criteria included a premorbid ADHD diagnosis, uncontrolled seizures, uncorrected hypothyroidism, severe sensory loss, patient or family history of Tourette syndrome, glaucoma, substance abuse history or current use of psychotropic medication. The protocol was approved by the institutional review boards of the participating sites [St. Jude Children's Research Hospital, Duke University Medical Center, and Medical University of South Carolina]. Written informed consent was obtained from a legal guardian prior to participation. Patient enrollment began in January of 2000.

Patients that met preliminary inclusion and exclusion criteria were screened to ensure that they had adequate global cognitive functioning (IQ > 50) and had academic and attention difficulties that might be amenable to MPH therapy. Following the screening phase, qualifying participants took part in a 2-day, in-clinic, double-blind, cross-over trial during which they received MPH and placebo in randomly assigned order 9. Patients who did not have a significant adverse reaction during the 2-day trial went on to participate in a randomized, double-blind, placebo-controlled 3-week home cross-over trial consisting of placebo (inert substance; bid), low-dose MPH (0.3 mg/kg; maximum dose, 10 mg bid), and moderate-dose MPH (0.6 mg/kg; maximum dose, 20 mg bid) 10. Patients in the present study were selected for participation in the 12-month open-label MPH trial if they showed improvement (≥3 points) over placebo on parent and/or teacher Conners' Rating Scales 19 during the 3-week home cross-over trial. The 12-month open-label MPH trial included individually titrated MPH dosing to maximize clinical benefit, monthly monitoring of side effects, and regular acquisition of parent and teacher ratings of attention and behavior regulation.

We report on 51 patients who have completed the 12-month open-label MPH trial, as well as 51 case-matched comparison patients who were initially screened for participation in the multiphase MPH trial and either qualified but chose not to participate in the 12-month open-label trial (n = 3) or did not qualify [n = 48; IQ < 50 (n = 1), academic problems not demonstrated (n = 11), attention problems not demonstrated (n = 17), academic and attention problems not demonstrated (n = 17), and mood disorder diagnosis (n = 2)]. In an effort to control for extraneous variables that may relate to growth trajectories, we systematically case-matched by diagnosis (ALL/BT), tumor type, tumor location (e.g., posterior fossa, suprasellar region), cancer treatment intensity (based on dose of chemotherapy and CNS-directed radiation therapy), gender, and age at diagnosis. Systematic case-matching also ensured that case-matched comparison patients had adequate growth data within the same age range that their counterpart was receiving MPH treatment.

Clinical Variables

Average daily dose of MPH was measured in milligrams (mg) of MPH per kilogram (kg) of body weight and was adjusted for days missed during the 12-month trial (e.g., some patients chose not to take MPH during weekends and/or extended school breaks). The main outcome variables of BMI, height, and weight were converted to gender and age corrected standard scores (z-scores) prior to statistical analyses using Centers for Disease Control (CDC) normative data 20. Intensity of CNS treatment was categorized into three levels: Low (systemic and/or intrathecal chemotherapy only), Mild (≤24 Gy cranial radiation therapy with or without chemotherapy), and Moderate (>24 Gy cranial radiation therapy with or without chemotherapy).

Statistical Analyses

Data analyses were performed using SAS software 21. Demographic and medical data were compared between groups using chi-square analyses for categorical data and independent t-tests for continuous data. A linear mixed model was used to analyze the longitudinal growth data across two time periods: (1) before the MPH start date and (2) during the 12-month MPH trial. If the slope of the growth trajectory within the second period was significantly smaller than the slope of growth trajectory within the first period, then significant growth deceleration was concluded. Separate statistical modeling of growth trends for BMI, height, and weight were performed.

RESULTS

Case-Matching

Demographic, clinical, and medical data for the MPH and comparison groups are presented in Table I. Statistical findings revealed no significant differences between groups, indicating that the groups were well matched on relevant variables with respect to comparing growth trajectories.

Table I. Patient Demographic and Clinical Characteristics (N = 102)

Characteristic	MPH group (n = 51)	Comparison group (n = 51)	P-value
ALL, acute lymphoblastic leukemia; CNS, central nervous system; MPH, methylphenidate. Chi-square analyses were used for categorical data and independent t-tests for continuous data. a Dose was adjusted for days missed.
Gender (female, n)	24 (47%)	20 (39%)	0.43
Ethnicity
Caucasian (n)	44 (86%)	44 (86%)	0.31
African American (n)	7 (14%)	5 (10%)
Other/unknown (n)	0 (0%)	2 (4%)
Diagnosis
Brain tumor (n)	23 (45%)	23 (45%)	1.00
ALL (n)	28 (55%)	28 (55%)
CNS treatment intensity
Chemotherapy only (n)	20 (39%)	20 (39%)	0.96
≤24 Gy CRT ± chemotherapy (n)	7 (14%)	8 (16%)
>24 Gy CRT ± chemotherapy (n)	24 (47%)	23 (45%)
Tumor location
Infratentorial (n)	17 (74%)	17 (74%)	1.00
Supratentorial midline (n)	6 (26%)	6 (26%)
Spinal radiation	14 (27.5%)	10 (19.6%)	0.35
Thyroid replacement (yes, n)	12 (24%)	17 (33%)	0.27
Growth hormone Replacement (yes, n)	13 (25%)	14 (27%)	0.82
Age at cancer diagnosis (Mean ± SD)	5.24 ± 3.09	5.77 ± 3.37	0.41
	Mean ± SD (range)
MPH group (n = 51)
Age at MPH start date (years)	11.68 ± 3.12 (6.72–18.40)
Length of MPH trial (months)	11.80 ± 1.42 (8.77–15.15)
Days missed MPH	43 ± 43 (0–159)
Daily MPH dose (mg/kg)a	0.48 ± 0.18 (0.12–0.89)

Primary Growth Outcomes

Body mass index

BMI slope estimates, representing changes in z-score points per month, of the MPH and comparison groups before the MPH start date (up to 40 months) were significantly greater than zero (MPH group slope = 0.010, SE = 0.004, P < 0.01; comparison group slope = 0.015, SE = 0.004, P < 0.001), indicating that prior to the MPH start date both groups had gradually increasing BMI with age, with no significant difference between groups. The BMI intercept estimates, reported as z-scores (mean of 0 and standard deviation of 1), of the MPH and comparison groups at the MPH start date were significantly higher than CDC normative data (MPH group intercept = 0.972, SE = 0.153, P < 0.0001; comparison group intercept = 1.070, SE = 0.155, P < 0.0001) with no significant difference between study groups.

The BMI slope estimate during the MPH trial (up to 14 months) for the MPH group was significantly less than zero (MPH group slope = −0.038, SE = 0.007, P < 0.001), indicating a deceleration of BMI during the MPH trial. In contrast, the BMI slope estimate for the comparison group during the MPH trial was not significantly different from zero (comparison group slope = 0.001, SE = 0.008). As shown in Figure 1A, the change in estimated BMI slope from before the MPH trial to after the MPH start date revealed a significant change in slope for the MPH group (MPH group slope change = −0.049, SE = 0.008, P < 0.0001) but not for the comparison group (comparison group slope change = −0.014, SE = 0.010).

Figure 1. Growth patterns for cancer survivors participating in a year long MPH trial () relative to cancer survivors not participating in a year long MPH trial (). The vertical lines denote the MPH start date. A: Body mass index (BMI) growth trends over time (*P < 0.0001; significant slope differences between the groups during the MPH trial). B: Height growth trends over time (*P = 0.049; significant change in estimated height slope from before the MPH trial to after the MPH start date for the comparison group only). C: Weight growth trends over time (*P < 0.0001; significant slope differences between the groups during the MPH trial).

Height

Height slope estimates, representing changes in z-score points per month, of the MPH and comparison groups before the MPH start date were not significantly different than zero (MPH group slope = −0.004, SE = 0.004; comparison group slope = −0.007, SE = 0.004), indicating that prior to the MPH start date height over time is stable relative to CDC normative data, and there was no significant difference between groups. The height intercept estimate, reported as a z-score, was not significantly different than the CDC normative data for the MPH group at the MPH start date (MPH group intercept = −0.216, SE = 0.177) yet was significantly below CDC normative data for the comparison group (comparison group intercept = −0.393, SE = 0.178, P = 0.03); however, the difference between study groups was not statistically significant.

Height slope estimates during the MPH trial for the MPH and comparison groups were not significantly different than zero (MPH group slope = −0.008, SE = 0.005; comparison group slope = 0.008, SE = 0.006), indicating no change in height velocity during the MPH trial, with no significant difference between groups. As shown in Figure 1B, the change in estimated height slope from before the MPH trial to after the MPH start date was not significant for the MPH group (MPH group slope change = −0.004, SE = 0.007). However, the comparison group's change in slope just reached statistical significance (comparison group slope change = 0.015, SE = 0.008, P = 0.049).

Weight

Weight slope estimates, representing changes in z-score points per month, of the MPH and comparison groups before the MPH start date were marginally greater than zero (MPH group slope = 0.007, SE = 0.003, P = 0.06; comparison group slope = 0.008, SE = 0.004, P = 0.02), indicating slightly higher than age-appropriate weight gain while not on MPH, with no significant difference between groups. Weight intercept estimates, reported as z-scores, of the MPH and comparison groups at the MPH start date were significantly higher than CDC normative data (MPH group intercept = 0.669, SE = 0.198, P < 0.01; comparison group intercept = 0.721, SE = 0.199, P < 0.001), indicating that both groups were significantly heavier than age- and gender-based comparisons within the CDC normative data, with no significant difference between groups.

As shown in Figure 1C, the weight slope estimate during the MPH trial for the MPH group was significantly less than zero (MPH group slope = −0.031, SE = 0.007, P < 0.001), representing a significant deceleration of weight associated with the MPH trial. In contrast, the weight slope estimate for the comparison group during the MPH trial was not significantly different from zero (comparison group slope = 0.005, SE = 0.008). The change in estimated weight slope from before the MPH trial to after the MPH start date revealed a significant change in slope for the MPH group (MPH group slope change = −0.037, SE 0.009, P < 0.0001) but not for the comparison group (comparison group slope change = −0.003, SE = 0.009), indicating that only the MPH group showed significant change in weight associated with the MPH trial. The effect sizes for detecting a difference in slope between the MPH and comparison groups for BMI, weight, and height during the year of medication fell in the medium to large range (0.418–0.806).

DISCUSSION

Prior to the MPH trial, the MPH and comparison groups showed no significant growth differences; both groups showed significantly increasing BMI and weight relative to CDC normative data. Both groups also showed BMI and weight significantly above CDC normative data at the time of the MPH start date, that is, one standard deviation above and just over two-thirds of a standard deviation above for BMI and weight, respectively. These findings of elevated BMI and weight in our patient groups are consistent with previously known risk for obesity among childhood cancer survivors due to factors such as damage to the hypothalamus, high-dose glucocorticoid treatment, prolonged growth hormone deficiency, and reduced physical activity 22.

Regarding height prior to the MPH trial, MPH, and comparison groups were not significantly different from each other and neither group was significantly different from age- and gender-based CDC normative data with respect to height velocity. Both groups showed height slightly lower than CDC normative data at the MPH start date, yet only the comparison group's height reached statistical significance, that is, approximately one-third standard deviation below CDC normative data. Perhaps because our study patients were still relatively young we were not yet able to detect height deficits that have been previously observed when measuring final-height outcomes in the childhood cancer survivorship population 16.

During the 12-month MPH trial, the MPH group showed significant deceleration of BMI and weight, but not height, whereas the case-matched comparison group showed no significant growth deceleration. The rate of BMI and weight deceleration for the MPH group was modest relative to CDC normative data (i.e., nearly one-half standard deviation decline in BMI in 12-months and nearly one-third standard deviation decline in weight), suggesting that MPH was reasonably well tolerated with respect to short-term growth.

Our finding of stable height velocity during the 12-month MPH trial is encouraging. Negative findings concerning height deceleration and stimulant medications are not uncommon, especially when no comparison groups are employed, follow-up periods are short, and medication doses are low 14. Our study design, though limited by not being a randomized controlled trial, included a case-matched comparison group comprised entirely of childhood cancer survivors (rather than healthy controls), thereby greatly strengthening the generalizablity of our findings to the childhood cancer survivor population. Moderate doses of MPH were achieved through the standard clinical practice of dose titration individualized to patients' optimal therapeutic response. Nevertheless, our negative height finding should be embraced with some caution given the relatively short duration of the MPH trial. Future studies of MPH and growth in the cancer survivor population should ideally employ longer follow-up periods, as some studies of growth patterns in the ADHD population have found that weight changes are more pronounced than height changes during the first 1–2 years, and height changes tend to continue out into the third and fourth years 23, 24.

In summary, modest deceleration in BMI and weight change were found in this sample of childhood cancer survivors during a 12-month MPH trial. Perhaps fortunately in this regard, our sample tended to have a weight buffer consistent with the childhood cancer survivor population's propensity for obesity. Nevertheless, our findings of BMI and weight deceleration warrant careful clinical consideration when attempting to balance the benefits of MPH for mitigating cognitive late-effects of childhood cancer with the potential for growth deceleration.

Acknowledgements

This work was supported, in part, by the Cancer Center Support (CORE) Grant P30 CA21765, R01CA078957 (awarded to Ray Mulhern, Ph.D.), U01 CA81445 from the National Cancer Institute, and by the American Lebanese Syrian Associated Charities (ALSAC).