Hematopoietic cell transplantation (HCT)-specific-comorbidity index: A new tool for risk assessment before allogeneic HCT

We have reported on the use of Charlson comorbidity index (CCI) to predict non-relapse mortality (NRM) and overall survival (OS) for patients (pts) given nonablative or ablative HCT1. However, the sample size of pts with scores of ≥1, captured by the CCI, did not exceed 35%. Further, some of comorbidities were rarely found among pts given HCT. Therefore, we sought to develop an HCT-specific-comorbidity index aimed at a) better defining previously identified comorbidities, i.e. adding pulmonary functions tests to pulmonary, liver function tests to hepatic, and ejection fraction ≤50% to cardiac comorbidities and b) investigating additional HCT-related comorbidities. To this end, we retrospectively reviewed comorbidities of 1055 pts given HCT at our center between 1997–2003 after nonablative (n=294) or ablative (n=761) conditioning. Pts were randomly divided into training (n=708) and validation sets (n=347). In the training set, the unadjusted hazard ratios (HR) for 2-year NRM were calculated for each comorbidity and then adjusted for other comorbidities, disease risk, and conditioning type. The adjusted HRs were employed as weights for individual comorbidities. Differences encountered compared to the original CCI were: a) two new comorbidities were added (obesity= score 1 and peritransplantation infection= score 2), b) age ≥50 years acquired a score of 2, and c) hypertension and asthma each acquired a score of 1 instead of 0, moderate pulmonary, peptic ulcer, and rheumatologic each acquired a score of 2 instead of 1, valvular heart disease a score of 2 instead of 0, and severe pulmonary comorbidity a score of 3 instead of 1. In the training set, HR for NRM for scores 0, 1, 2, 3, 4, ≥5 were 1, 1.2, 3.5, 6.1, 7.1, and 10.8, respectively. The modified index was then validated using the validation set. This index had the advantage of 1) capturing more pts with high scores and 2) distinguishing pts with low scores who had lower NRM and better OS when compared to the original CCI (Table). Applying the scores to nonablative and ablative pts, respectively, NRM of 5 vs 10% (p=0.4) and OS of 85% vs 75% (p=0.1) were seen for scores of 0–1, 17 vs 27% (p=0.04) and 61 vs 59% (p=0.2) for scores of 2–3, and 33 vs 54% (p=0.03) and 43 vs 30% (p=0.006) for scores of ≥4. This HCT-specific comorbidity index provides a simple, readily applicable and valid method of estimating NRM and OS among pts given nonablative or ablative allogeneic HCT. (1. Sorror et al. Blood. 2004;104:961.)TablePercentages of pts and Two-Year Rates of NRM and OS in the Validation Set (n = 346) as Scored by the New HCT-Specific Comorbidity Index Compared to the Original CCIHCT-CIpts, %Two-Year RateOriginal CCIpts, %Two-Year RateNRM, %OS, %NRM, %OS, %0–137137205817642–33622581273351≥4274035≥2153141NRM indicates non-relapse mortality; OS, overall survival; CCI, Charlson comorbidity index; and HCT-CI, hematopoietic cell transplantation-specific comorbidity index. Open table in a new tab NRM indicates non-relapse mortality; OS, overall survival; CCI, Charlson comorbidity index; and HCT-CI, hematopoietic cell transplantation-specific comorbidity index.