Impact of high-dose granulocyte transfusions in patients with cancer with candidemia: retrospective case-control analysis of 491 episodes of Candida species bloodstream infections.

BACKGROUND

The efficacy and feasibility of donor granulocyte transfusion therapy (GTX) have changed considerably over the past four decades. The authors sought to determine the impact of high-dose (approximately 5.5 x 10(10) cells) GTX in patients with candidemia.

METHODS

The authors' case-control retrospective analysis comprised 491 consecutive patients treated at The University of Texas M. D. Anderson Cancer Center (Houston,TX) from 1993 to 2000. The cohort included 29 patients with Candida species bloodstream infection who had received GTX and 462 who had not.

RESULTS

Both groups were comparable in age, gender, APACHE II score, recent chemotherapy received, broad-spectrum antibiotics, systemic corticosteroids, radiotherapy, intravascular catheter, and concordant antifungal therapy (P > or = 0.1). The patients who received GTX compared with those who did not had a higher incidence of underlying leukemia (86% vs. 29%, P <0.001), persistent neutropenia (59% vs. 18%, P <0.001), non-Candida albicans candidemia (Candida glabrata, 35%; Candida krusei, 31%: 90% vs. 67%, P=0.01), and breakthrough invasive mycosis (62% vs. 23%, P <0.001). Neutropenia was more prolonged in patients who received GTX (28 vs. 10 days, P <0.001). Also, more of the patients who received GTX had received hematopoietic stem cell transplantations (28% vs. 13%, P = 0.03), exposure (within 4 weeks) to antifungals (79% vs. 38%, P <0.001), and stays in critical care units (62% vs. 40%, P=0.02). The overall attributable mortality rate for 25 evaluable recipients of GTX was 48% (n=12), compared with 45% (n=115) of 254 evaluable patients in the control group (P=0.5). Of the 158 patients with leukemia, 25 (16%) had received GTX. In patients with leukemia, more of those who had received GTX experienced disseminated candidiasis (44% vs. 26%; P <0.07) and persistent neutropenia (68% vs. 43%, P <0.02), had candidemia that was more prolonged (> 72 hours, P <0.02), and had more stays in critical care units (68% vs. 44%, P <0.03). On the bases of a reduced multivariate model, a significantly increased risk of death was found for patients with hematopoietic stem cell transplantation (odds ratio [OR]=2.51; 95% confidence interval [95% CI], 0.99-6.31; P <0.05), for patients with persistent neutropenia (OR=4.57; 95% CI, 1.99-10.47; P <0.0003), and for patients with leukemia who also had prolonged candidemia (OR=3.59; 95% CI, 1.61-7.98; P <0.002), disseminated candidiasis (OR=5.19; 95% CI, 2.17-12.42; P <0.0002), or non-C. albicans candidemia (OR=5.02; 95% CI, 1.07-23.64; P <0.04). In patients with leukemia, death was attributable to candidemia in 50% of the GTX recipients, compared with 59% of the non-GTX patients who had received antifungal therapy alone (P=0.4).

CONCLUSIONS

Despite the presence of multiple predictors of increased mortality, high-dose GTX therapy in these high-risk patients with cancer was associated with better than expected survival rates.