Changes in body mass index, weight, and height in children with acute myeloid leukemia and the associations with outcome

Abstract

Little is known about body composition changes in patients with acute myeloid leukemia (AML) during and after treatment or their associations with outcomes. Z-scores for body mass index (BMI), weight, and height at diagnosis, their longitudinal changes from diagnosis to 5 years off therapy, and their associations with adverse effects and outcomes were evaluated in 227 pediatric patients with AML enrolled in the AML02 and AML08 trials at St. Jude Children's Research Hospital between 2002-2017. The median Z-scores for baseline weight, height, and BMI were 0.193, 0.209, and 0.170, respectively, and those for weight and height decreased significantly during therapy to -0.038 and -0.163, respectively, at off-therapy (P < .001 for both). At 5 years off therapy, the Z-scores for weight and BMI had increased significantly to 0.492 (P = .003) and 0.911 (P < .001), respectively, whereas the height Z-score remained significantly lower at -0.066 (P < .001) compared with baseline. The height Z-score of transplant recipients decreased further from -0.211 at transplant to -0.617 12 months later (P < .001). Baseline BMI category and Z-score were not associated with outcomes, but higher weight Z-scores were associated with lower incidences of refractory or relapsed disease (hazard ratio [HR], 0.82; 95% confidence interval [CI], 0.67-0.99) and higher incidences of death in remission (HR, 1.31; 95% CI, 1.01-1.70). Furthermore, weight Z-score decrease during induction therapy was associated with gastrointestinal, hepatic, and infection toxicities during subsequent therapy and with death in remission (HR, 2.66; 95% CI, 1.11-6.45). Multidisciplinary monitoring for weight changes and short stature is required from diagnosis to the off-therapy period.

Graphical abstract

Figure 1.

Longitudinal changes in the percentage of patients in each body mass index category. Con, consolidation; E, end of each chemotherapy course; Ind, induction; m, month; S, start of each chemotherapy course; y, year.

Figure 2.

Longitudinal changes in the median Z-scores for body mass index (A), weight (B), and height of patients during therapy and in the off-therapy period (C). The box corresponds to the interval from the first quartile to the third quartile. The line dividing the box indicates the median Z-score. The whiskers represent 1.5 times the interquartile range values, and values outside these ranges are shown as green dots (•). Significant differences from baseline values are represented by asterisks (*). Asterisks and letters in red represent increases in Z-scores from the baseline values; asterisks and letters in blue represent decreases in Z-scores from the baseline values. For panel A, blue, orange, and red lines show the Z-score cutoffs for underweight (<−1.6499, [5th percentile]), overweight (≥1.0364 [85th percentile] to <1.6499 [95th percentile]), and obese (≥1.6499 [95th percentile]), respectively. Healthy weight Z-scores range from ≥−1.6499 (5th percentile) to <1.0364 (85th percentile). To obtain smooth body composition Z-scores for toddlers (aged 0-3 years), we interpolated the Z-scores for BMI and weight-for-length for children aged 2 to 3 years; for patients aged >20 years, the Z-score was calculated based on reference data for individuals aged 20 years, as described in “Patients and methods.” Con, consolidation; E, end of each chemotherapy course; Ind, induction; m, months; S, start of each chemotherapy course; y, years.

Figure 3.

Outcomes of patients based on body mass index category. (A) Event-free survival, (B) overall survival, (C) cumulative incidence of refractory/relapsed disease, and (D) cumulative incidence of death in first remission according to body mass index category at diagnosis in 227 children with acute myeloid leukemia. Data were analyzed with the patients categorized by body mass index as underweight, healthy weight, or overweight/obese.

Figure 4.

The association of Z-scores for body mass index and weight at diagnosis and decreases in those Z-scores during induction therapy with treatment outcomes.