Patient-Reported Outcomes of BMT CTN 1703: A Randomized Phase III Study for Gvhd Prophylaxis - a Quality of Life Evaluation

Abstract

Introduction A randomized phase III trial of post-transplantation cyclophosphamide (PTCy), tacrolimus, (Tac), and mycophenolate mofetil (MMF) versus tacrolimus/methotrexate (Tac/MTX) for graft-versus-host disease (GVHD) prophylaxis showed superior graft-versus-host disease-free, relapse-free survival (GRFS) in the PTCy-containing arm (NEJM, 2023). Here, we report the effect of GVHD prophylaxis on patient-reported outcomes and impact on quality of life (QOL). Methods The QOL study component of BMT CTN 1703 employed four primary endpoints: Lee Chronic GVHD Symptom Score and the PROMIS subscales of physical function, gastrointestinal symptoms, and satisfaction with participation in social roles. Secondary endpoints included hemorrhagic cystitis symptoms and Lee symptom subscale differences, with questionnaires to English and Spanish-speaking participants at enrollment and on Days 100, 180, and 365 post-transplant. Given the risk of missing assessments, an inverse probability weighted-independent estimating equations (IPW-IEE) model was applied for score comparison between arms, adjusting for baseline score, assessment time, age, donor type, disease risk index (DRI), planned conditioning regimen, and planned use of post-transplant maintenance therapy. Significance was declared at a 1.25% level for score comparisons between arms after Bonferroni correction for multiple testing. The analysis approach classified patients according to the intention-to-treat principle. Results QOL survey completion rates for patients at each timepoint were similar, exceeding 70% in each study arm. With the Lee Chronic GVHD Symptom Scale, higher scores indicate more severe symptoms. The IPW-IEE models indicated that the Lee Chronic GVHD Symptom Scale total scores were significantly lower in the PTCy arm compared to Tac/MTX at day 100 (mean difference [MD] -2.22, 99% confidence interval [CI] -5.05 to 0.61) and day 365 (MD -1.84, 99% CI -5.13 to 1.45, p=0.01 for overall treatment effect across all timepoints, Figure 1). Nutrition subscores were significantly better with PTCy at day 100 (MD -2.41, 99% CI -5.06 to 0.24, p=0.0014). Mouth scores were significantly better in the PTCy arm at day 365 (MD -7.09, 99% CI -13.21 to -0.98, p=0.0035). Patients over 65 years of age had better psychological subscores across the study population (MD -4.80, 99% CI -8.91 to -0.69, p=0.0026); no interaction was found between age and study treatment. Patient-reported hemorrhagic cystitis symptoms showed no significant difference between the study arms at any post-transplant time point. There was no significant treatment effect observed on the 3 PROMIS subscales. Across the entire study population, both relapse and acute GVHD were associated with worse scores on the Lee Symptom Scale, and PROMIS physical function, gastrointestinal symptoms, and social functions subscales. Chronic GVHD was associated with worse scores on the Lee Symptom Scale only, not the PROMIS subscales. No associations were found between patient-reported QOL at baseline and 1-year GRFS, disease-free survival, or overall survival. Conclusions The PTCy arm of BMT CTN 1703 showed superior GRFS, and this study confirms better patient-reported outcomes with PTCy. Patients who received PTCy showed better nutrition and mouth subscores on the Lee Chronic GVHD Symptom Scale compared to those who received Tac/MTX, possibly due to lower observed rates of severe acute and chronic GVHD in the PTCy arm. No association was found between patient-reported baseline quality of life and survival endpoints, suggesting the necessity of further research to elucidate the interplay of treatment strategies, patient factors, and their impact on long-term outcomes.