Absolute lymphocyte count recovery after allogeneic hematopoietic stem cell transplantation predicts clinical outcome.

Immune reconstitution is critical for clinical outcome after allogeneic hematopoietic stem cell transplantation (HSCT). To determine the impact of absolute lymphocyte count (ALC) recovery on clinical outcomes, we conducted a retrospective study of 1109 adult patients who underwent a first allogeneic HSCT from 2003 through 2009, excluding patients who died or relapsed before day 30. The median age was 51 years (range, 18 to 74) with 52% undergoing reduced-intensity conditioning and 48% undergoing myeloablative conditioning HSCT with T cell–replete peripheral blood stem cells (93.7%) or marrow (6.4%) grafts. The median follow-up time was 6 years. To determine the threshold value of ALC for survival, the entire cohort was randomly split into a training set and a validation set in a 1:1 ratio, and then a restricted cubic spline smoothing method was applied to obtain relative hazard estimates of the relationship between ALC at 1 month and log hazard of progression-free survival (PFS). Based on this approach, ALC was categorized as ≤.2 × 109 cells/L (low) or >.2 × 109 cells/L. For patients with low ALC at 1, 2, or 3 months after HSCT, the overall survival (OS) (P ≤ .0001) and PFS (P ≤ .0002) were significantly lower and nonrelapse mortality (NRM) (P ≤ .002) was significantly higher compared with patients with ALC > .2 × 109 cells/L at each time point. When patients who had low ALC at 1, 2, or 3 months after HSCT were grouped together and compared, their outcomes were inferior to those of patients who had ALC > .2 × 109 cells/L at 1, 2, and 3 months after HSCT: the 5-year OS for patients with low ALC was 28% versus 46% for patients with ALC > .2 × 109 cells/L, P < .0001; the 5-year PFS was 21% versus 39%, P < .0001, respectively and 5-year NRM was 40% versus 18%, P < .0001, respectively. This result remained consistent when other prognostic factors, including occurrence of grade II to IV acute graft-versus-host disease (GVHD), were adjusted for in multivariable Cox models stratified by conditioning intensity: hazard ratio (HR) for OS: 1.52; P ≤ .0001; for PFS: 1.42; P = .0008; and for NRM: 2.4 P < .0001 for patients with low ALC. Low ALC was not significantly associated with relapse (HR, 1.01; P = .92) in the multivariable model. Low ALC early after HSCT is an independent risk factor for increased NRM and poor survival independent of grade II to IV acute GVHD.