Methods: We analyzed consecutive patients who received melphalan 200 mg/m2 followed by an AHPCT for MM from 2008 to 2014. ES was diagnosed using a minor modification of the Maiolino criteria (MC) and Spitzer criteria (SC). ES using MC was defined as a noninfectious fever ›38°C plus rash or pulmonary infiltrates within 96 hours of neutrophil recovery. ES using SC was defined as the presence of 3 major or 2 major plus 1 or more minor criteria within 96 hours of neutrophil recovery (major criteria: noninfectious fever ›38°C, rash, pulmonary infiltrates; minor criteria: liver/kidney dysfunction, weight gain or encephalopathy). We analyzed the incidence of ES using exact logistic regression, and time to engraftment and hospital length of stay (LOS) using Kaplan-Meier, logrank test and Cox proportional hazards regression.
Results: A total of 111 patients were included (68 received G-CSF and 43 received GM-CSF starting day +3 of AHPCT). There was no significant difference in patient age, gender and race. Peripheral blood stem cell mobilization was achieved using a plerixafor-containing regimen in 45% of patients in the G-CSF group vs. 95% of patients in the GM-CSF group (p=0.001). A cyclophosphamide-containing regimen was used for mobilization in 45% of patients in the G-CSF group vs. 9% of patients in the GM-CSF group (p=0.001). The incidence of ES by MC and SC were 24% and 13%, respectively. In unadjusted analysis, the GM-CSF group was significantly more likely to develop ES than the G-CSF group using SC (28% vs. 3%; odds ratio, OR=12.5; p=0.001) and MC (47% vs. 10%; OR=7.42; p=0.001). After adjusting for age, gender, race and CD34+ cell dose, the GM-CSF group remained at higher risk for ES than the G-CSF group (OR=10.4, p=0.069 for SC; OR=5.91, p=0.029 for MC). The GM-CSF group also had a significantly longer time to neutrophil engraftment than the G-CSF group by 1 day (p=0.001); the difference remained after multivariable analysis (hazard ratio, HR=0.22; 95% CI: 0.12 to 0.43; p=0.001). There was no significant difference between the two groups with respect to time to platelet engraftment (p=0.891). The median LOS of the G-CSF and GM-CSF group was 15 and 16 days, respectively, although not statistically significant (p=0.279).
Conclusion: GM-CSF use after AHPCT for MM may be associated with a higher risk of ES and longer time to neutrophil recovery as compared to G-CSF. Our observations should prompt further investigation.