Incorporating patient‐reported outcome data into a predictive calculator for allogeneic hematopoietic cell transplantation recipients

• Published in: Cancer 2024-05, Vol.130 (10), p.1826-1835
• Main Authors: Shaw, Bronwen E., Flynn, Kathryn E., He, Naya, Cusatis, Rachel, D'Souza, Anita, Hamilton, Betty K., Horowitz, Mary M., Mattila, Deborah, Phelan, Rachel, Lee, Stephanie J., Brazauskas, Ruta
• Format: Article
• Language: English
• Subjects: Adult; Aged; Allografts; Blood; Bone cancer; Bone marrow; Bone marrow transplantation; Calculators; Cancer therapies; Clinical trials; Female; hematopoietic cell transplantation (HCT); Hematopoietic Stem Cell Transplantation; Hematopoietic stem cells; Humans; Male; Middle Aged; Patient Reported Outcome Measures; patient‐reported outcome (PRO); Quality of Life; Questions; single‐item health question; Stem cell transplantation; Survival; Transplantation; Transplantation, Homologous; Young Adult
• ISSN: 0008-543X; 1097-0142; 1097-0142
• DOI: 10.1002/cncr.35189

Background The Center for International Blood and Marrow Transplant Research (CIBMTR) provides a 1‐year overall survival calculator to estimate outcomes for individual patients before they undergo allogeneic hematopoietic cell transplantation (HCT) to inform risk. The calculator considers pre‐HCT clinical and demographic characteristics, but not patient‐reported outcomes (PROs). Because pre‐HCT PRO scores have been associated with post‐HCT outcomes, the authors hypothesized that adding PRO scores to the calculator would enhance its predictive power. Methods Clinical data were obtained from the CIBMTR and the Blood and Marrow Transplant Clinical Trials Network. The PRO measures used were the 36‐Item Short Form Survey (SF‐36) and the Functional Assessment of Cancer Therapy‐Bone Marrow Transplantation. One thousand thirty‐three adult patients were included. Results When adjusted for clinical characteristics, the SF‐36 physical component score was significantly predictive of 1‐year survival (hazard ratio [HR], 0.88; 95% confidence interval [CI], 0.81–0.95; p = .0015), whereas the mental component score was not (HR, 1.02; 95% CI, 0.95–1.10; p = 0.6396). The baseline single general health question on the SF‐36 was also significantly associated with mortality (HR, 1.91 for those reporting fair/poor health vs. good, very good, or excellent health; 95% CI, 1.33–2.76; p = .0005). The addition of PRO scores to the calculator did not result in a significant change in the model’s predictive ability. Self‐reported pre‐HCT scores were strongly predictive of self‐reported health status (odds ratio, 3.35; 95% CI, 1.66–6.75; p = .0007) and quality of life (odds ratio, 3.24; 95% CI, 1.93–5.41; p