Symptom Burden of Busulfan + Melphalan Versus Melphalan Alone for Multiple Myeloma

Background: High-dose melphalan at 200 mg/m² (Mel) is the standard for autologous hematopoietic stem cell transplantation (autoHSCT) for multiple myeloma. Two recent retrospective analyses suggested that a combination of busulfan and melphalan (Bu-Mel) may be associated with a longer progression-free survival (PFS). A secondary aim of a randomized phase III trial that compared the safety and efficacy of Bu-Mel vs Mel was to compare the symptom severity and symptom burden of the two regimens. Symptom burden is defined as the combined impact of disease- and therapy-related symptoms on patient functional ability.

Methods: Patients were randomized to Bu-Mel or Mel. In the Bu-Mel arm, Bu 130 mg/m² was infused daily for 4 days, either as a fixed dose or to target an average daily area under the curve (AUC) of 5000 μmol-min, followed by 2 daily doses of Mel at 70 mg/m². A subset of patients completed the 20 symptom severity and 6 interference items of MD Anderson Symptom Inventory for Multiple Myeloma (MDASI-MM) prior to the start of the treatment regimen and weekly for 5 weeks post autoHSCT. Symptoms and interference are rated on 0-10 scales (0 = none or no interference, 10 = worst imaginable or complete interference). Differences in individual symptom severity and interference and symptom burden (the composite mean scores of the 5 most severe symptoms and the 6 interference items) between the two arms were assessed by t-tests and mixed modeling at each time point.

Results: As previously reported, 92 patients (Bu-Mel: 49, Mel: 43) were enrolled between October 2011 and August 2013. Grade 3-4 non-hematologic toxicity was seen in 41 (84%) and 20 (47%) patients in Bu-Mel and Mel, respectively (p=0.0003). At day 90, 8 (18%) and 15 (35%) patients had achieved a CR (p=0.05), and 13 (27%) and 20 (47%) patients had achieved a CR + nCR (p=0.05) in Bu-Mel and Mel, respectively. One-year PFS was 93% and 82% (p=0.26), and 1-year OS was 100% and 100% in Bu-Mel and Mel arms, respectively. Fifty-six of the patients (Bu-Mel: 29, Mel: 27) completed at least one MDASI-MM assessment between October 2011 and July 2013. Median ages at autoHSCT were 55.5 and 55.1 years in Bu-Mel and Mel groups, respectively (p=0.862). At baseline, t-tests showed significantly higher mean severity for bone aches (3.00, standard deviation [SD] = 2.85 vs 1.40, SD = 1.94; p=0.022), drowsiness (2.96, SD = 3.12 vs 1.44, SD = 1.87; p=0.034), muscle weakness (2.98, SD = 2.68 vs 0.79, SD = 1.14; p=0.001, appetite (1.93, SD = 2.60 vs 0.44, SD = 1.19; p=0.009), and overall interference (3.46, SD = 2.89 vs 1.03, SD = 1.04; p<0.001) in the Bu-Mel arm than the Mel arm. Longitudinal analysis using mixed modeling showed a significant difference in the combined mean of the 5 most severe symptoms (fatigue, pain, drowsiness, disturbed sleep, and lack of appetite) at baseline (estimated difference [ED] = 1.16, p=0.016) between the arms and a significant increase in the top 5 symptoms for both arms between baseline and Weeks 1 (ED = 3.30, p<0.0001), 2 (ED = 2.06, p=0.002), and 3 (ED = 2.28, p=0.003). The Bu-Mel arm had a significantly lower mean severity of disturbed sleep at Week 4 (ED = -3.69, p=0.022) and significantly lower mean severity of lack of appetite at Week 2 (ED = -3.08, p=0.016) and 3 (ED = –3.15, p=0.033) than the Mel arm, but a significantly higher mean severity of sore mouth at Week 1 (ED = 2.35, p=0.018) and 2 (ED = 2.51, p=0.009).

Conclusions: Patients with MM undergoing autoHSCT report a significant increase in symptom severity and interference during Weeks 1 to 3 post autoHSCT, with the greatest increase occurring during the first week. In addition, patients receiving Bu-Mel vs Mel report differences in individual symptom severity, with Mel patients having significantly more problems with disturbed sleep and lack of appetite and Bu-Mel patients reporting significantly more severe sore mouth and throat. The effect of the significant differences prior to the start of autoHSCT between these two groups, despite randomization, in the severity of the top 5 symptoms and interference on symptom report during autoHSCT is unclear and will be further explored. Additional differences in symptom severity may have gone undetected in this study because of initial differences. Systematic measurement of symptom burden during clinical trials can provide useful information for clinicians and patients in evaluating the full impact of different treatment regimens.