High incidence of invasive aspergillosis associated with intestinal graft-versus-host disease following nonmyeloablative transplantation

Abstract

Invasive aspergillosis (IA) remains a major complication following allogeneic hematopoietic stem cell transplant (HSCT). In contrast to conventional HSCT, few investigators have examined risk factors of IA associated with nonmyeloablative (NMA) regimens characterized by outpatient administration, immunosuppression rather than cytoreduction, and short duration of neutropenia posttransplant. We report our results on a cohort of 125 patients treated homogenously who received a 6/6 matched sibling NMA HSCT designed to be performed on an outpatient basis. Conditioning regimen included fludarabine (30 mg/m(2) x 5 days) and cyclophosphamide (300 mg/m(2) x 5 days) followed by reinfusion of a minimum of 4 x 10(6) CD34(+) cells/kg. Acute graft-versus-host disease (aGVHD) prophylaxis consisted of tacrolimus and mycophenolate mofetil (MMF). Overall, 13 patients developed IA (5 proved, 6 probable, 2 possible) 44-791 days (median 229) after NMA HSCT, with a risk of 7% at 1, 11% at 2, and 15% at 3 years. Patients who suffered from IA had poorer overall survival (crude hazard ratio 2.3; 95% confidence interval [CI] 1.0-5.4; P = .045). Intestinal aGVHD or chronic GVHD (cGVHD) was significantly associated with IA at 1 (27% versus 3%, P = .003), 2 (27% versus 8%, P = .01), and 3 years (37% versus 10%, P = .005). The use of daclizumab was also significantly associated with IA at 3 years (47% versus 12%, P = .02). Age, sex, diagnosis, previous autologous transplant, duration of neutropenia, occurrence of cytomegalovirus viremia, duration of steroids or MMF intake, aGVHD, cGVHD, and cumulative number of days spent in hospital were not associated with IA. After multivariate analysis, intestinal GVHD remained the only statistically significant risk factor for IA at 1 (P = .003), 2 (P = .01), and 3 years (P = .005). We conclude that in NMA HSCT, the risk of IA increases over time and is significantly associated with intestinal GVHD. Because there is currently no surrogate in vitro markers of immunocompetence following NMA HSCT, this clinical finding is of particular importance to identify a population at higher risk who should be targeted for antimold prophylaxis.