Impact of Self-Reported Pre-Transplant Quality of Life (QOL) on Long-Term Survival after Allogeneic or Autologous Hematopoietic Cell Transplantation (HCT)

The role of pre-HCT self-reported QOL on long-term survival in autologous and allogeneic HCT recipients is not well understood. Although pre-HCT comorbidity is now a well-established indicator of post-HCT short-term outcome, patient-reported pre-HCT QOL may provide additional information for predicting long-term survival after HCT.

This study examined the relationship between pre-HCT QOL scores and overall survival (OS), non-relapse mortality (NRM), and disease-related mortality (DRM) in patients who received their first allogeneic or autologous HCT at City of Hope between 2001 and 2005. Four domains of QOL (physical, psychological, social, and spiritual well-being; scale of 0 [worst] to 10 [best]) were assessed prior to HCT using the City of Hope-QOL instrument. Cox regression analysis was used to identify QOL domains associated with long-term survival, adjusted for demographic, clinical, and transplant-related variables. HCT-Comorbidity-Age Index (HCT-CAI) (Sorror et al. 2014) was also assessed in allogeneic HCT recipients. Vital status was determined using medical records and linkage with the National Death Index. OS probabilities were estimated using the Kaplan-Meier estimator. Analyses were conducted by stem cell source.

The cohort comprised 103 allogeneic HCT recipients (median age at HCT, 42 years; 51% males; median follow-up since HCT, 7.6 years; 55 deaths) and 141 autologous HCT recipients (median age at HCT, 50 years; 56% males; median follow-up, 7.1 years; 59 deaths). The diagnoses included acute leukemia (n=45), chronic leukemia (n=20), lymphoma (n=14) and other hematologic conditions (n=24) in allogeneic HCT recipients and acute leukemia (n=24), lymphoma (n=76), myeloma (n=35), and other hematologic conditions (n=6) in autologous HCT recipients. All autologous and 81.5% of allogeneic HCT patients received myeloablative conditioning agents. OS: In allogeneic HCT recipients, OS adjusted for significant covariables increased with higher pre-HCT physical well-being score (Hazard Ratio [HR]=0.76 per point increase, p=0.003). Of note, HCT-CAI was not associated with OS in the multivariate model (HR=1.1, p=0.39). The hazard of death for physical well-being scores below the 75^th percentile was 3.7 times that above this percentile (p=0.002). Survival rates at 5 and 10 years after HCT for physical well-being scores below the 75^th percentile were 51% and 38%, respectively, compared to 72% (p=0.062) and 68% (p=0.018), respectively, above the 75^th percentile. OS was also independently worse in males, unrelated donor HCT recipients, patients with acute leukemia vs. other diagnoses, absence of chronic GvHD, and pre-HCT income of $20K to $100K vs. <$20K or >$100K. In autologous HCT recipients, adjusted OS varied non-linearly with pre-HCT physical well-being score (p=0.03), with OS being significantly lower in the lowest (worst) quartile of pre-HCT physical well-being score (HR=2.4, p=0.004) compared to those with better scores. Survival rates at 5 and 10 years after HCT in the lowest quartile were 53% and 47%, respectively, compared to 70% (p=0.048) and 62% (p=0.065) for those with scores above the lowest quartile. Older age at HCT, other diagnoses vs. lymphoma, high risk of relapse at pre-HCT, cyclophosphamide as conditioning agent vs. no cyclophosphamide were also independently associated with worse OS. Other QOL domains were not significant for either HCT types. NRM: In allogeneic HCT recipients, NRM was not associated with HCT-CAI (HR=1.18, p=0.11) adjusted for chronic GvHD and stem cell donor relatedness. However, adjusted OS increased for higher pre-HCT physical well-being score (HR=0.8, p=0.06). NRM was higher for those with physical well-being score below the 75^th percentile (HR=7.3, p=0.007) compared to those with scores above the 75th^th percentile. In autologous HCT recipients, no QOL domains were significant. DRM: None of the QOL domains for either HCT types, or HCT-CAI in allogeneic HCT recipients were significant after accounting for salient clinical and demographic factors.

This study shows that patient-reported pre-HCT physical well-being scores provide information not captured by medical factors that can aid in predicting long-term OS and NRM in allogeneic and OS in autologous HCT patients. Effort to incorporate such evaluation as part of the routine assessment preceding HCT may be useful in developing patient selection and intervention strategies.