Biomarkers To Predict Outcome after Allogeneic Hematopoietic Cell Transplant (HCT).

Two tools have recently been developed to predict allogeneic HCT tolerance: the HCT-comorbidity index (HCT-CI) (Sorror, Blood 2005) and “risk group” (Artz BBMT, 2006). The HCT-CI focuses on a comprehensive list of comorbidities whereas the “risk group” combines a simple comorbidity tool and performance status to identify patients at high risk for transplant-related mortality (TRM) and inferior survival. However, clinical parameters have limitations of reproducibility and accuracy. Biomarkers represent another promising method of risk stratification. Thus, we analyzed whether biomarkers independently predicted HCT tolerance. Among 112 consecutive transplants on a single protocol, 81 patients with pre-HCT cryopreserved sera underwent analysis of C-reactive protein (CRP) and 79 of interleukin-6 (IL-6) levels. AML and MDS represented the most common diagnosis (57%). All patients underwent HCT using fludarabine (125 mg/m2 IV total), melphalan (140 mg/m2 IV total) and alemtuzumab (100 mg IV total). The median age was 52 yrs and 46% had active disease at HCT. The median follow-up was 42 months. To determine tolerance, we evaluated initial hospital duration, aGVHD, and TRM. Median duration of initial hospitalization for HCT was 23 days and Grade II–IV aGVHD developed in 30 pts. Day 100 and day 180 TRM were 17% and 23% respectively. The median pre-HCT CRP level was 18.5 mg/L (mean, 40.5 mg/L; range, 0.17 to 180). The median IL-6 was 78.3 mg/L (range, 10 to 2258). CRP and IL-6 above the median were tested as adverse risk factors. High CRP was strongly associated with prolonged hospitalization (P=0.007) whereas increased IL-6 was not (P=0.30). HCT-comorbidity index (HCT-CI) ≥ 3(P=0.126) and “risk-group” (P=0.091) did not confer an increased risk of prolonged hospitalization. In univariate analysis, CRP above the median also predicted for grade II–IV aGVHD (P= 0.003). TRM was predicted by CRP (P=0.013) but not IL-6 (P=0.22). Multivariate analysis showed CRP retained independent predictive value for TRM when considering adverse risk factors of age ≥50, HCT-CI ≥ 3, active disease, or “risk group”. The impact of these biomarkers was limited to HCT tolerance as CRP and IL-6 did not predict for increased relapse (P=0.42 for both). Finally, inferior overall survival was associated with pre-HCT CRP (P= 0.029) but not IL-6 (P= 0.48). CRP holds promise as a reproducible biomarker to predict HCT related morbidity and mortality independent of standard measures. IL-6 may be less useful. These findings require validation, but because of the ready availability and reproducibility of CRP, this biomarker could be rapidly integrated into other HCT risk-assessment tools.