The impact of CHOP versus bendamustine on bone mineral density in patients with follicular lymphoma enrolled in the GALLIUM study

Abstract

Introduction: A standard treatment with CHOP or CVP includes a cumulative dose of 3,000–4,000 mg prednisone. Observational studies have shown increased risk of fractures in patients treated with R-CHOP/R-CVP. However, it is not clear whether reductions in bone mineral density (BMD) are mainly caused by glucocorticoids or the combined impact of chemotherapy and glucocorticoids. The objective of this study was to investigate the impact of G/R-bendamustine versus G/R-CHOP on BMD in patients with follicular lymphoma (FL) enrolled in the GALLIUM study. Methods: Patients enrolled in the GALLIUM study, an international randomized phase 3 trial of first-line treatment for FL, were included if they were ≥60 years at inclusion and in complete remission following treatment with G/R-bendamustine or G/R-CHOP. Exclusion criteria were treatment with anabolic or antiresorptive therapies prior to baseline, co-existing medical conditions associated with low BMD, and new anti-lymphoma treatment during the initial 5 years of follow-up. CT scans were performed at baseline, induction treatment completion (ITC), and annually in five years after inclusion. Hounsfield units (HU) measured in an ovoid region in the anterior of L1 were used as surrogate for BMD, which has been shown to correlate with BMD measurements from conventional dual-energy X-ray absorptiometry (DXA). Low HU values represent lower bone density while high values represent more dense bone. Vertebral compression fractures were defined as vertebral height loss of ≥20%. Results: At ITC, 244 patients ≥60 years old were in CR. Of those, 155 fulfilled the inclusion criteria, 55 (35%) received G/R-CHOP and 100 (65%) received G/R-bendamustine. CT scans with contrast were performed at baseline and ITC in 143 patients. Baseline characteristics, including age and gender, were balanced between groups. The mean baseline HU was 143.8 in the CHOP group and 141.5 in the bendamustine group (P = 0.74). The mean HU decrease from baseline to ITC was 27.8 after CHOP versus 17.1 after bendamustine, corresponding to a difference of 12.5 (95% CI: 5.5–19.5) after adjusting for age, sex, B-symptoms, and baseline HU. BMD remained at the ITC level, and thus below baseline, during the first 3 years of follow-up. During treatment and 5 years follow up, a total of nine new compression fractures were recorded in five patients. One patient in the bendamustine group and four patients in the CHOP group developed new vertebral fractures. Keywords: Indolent non-Hodgkin lymphoma, Late Effects in Lymphoma Survivors No conflicts of interests pertinent to the abstract.