A Subgroup Analysis of Patients Undergoing Autologous and Allogeneic Hematopoietic Stem Cell Transplantation (HSCT) in Deflect-1: A Double-Blind, Randomized Study of Fidaxomicin Vs. Placebo for Prophylaxis Against C. Difficile Infection

Introduction:  Clostridium difficile-associated diarrhea (CDAD) is a common complication during HSCT that has been associated with increased morbidity and mortality.  CDAD has been reported to occur in 6 to 30% of HSCT patients.  An ideal agent for prevention of CDAD has not been established.  Fidaxomicin (FDX) is bactericidal against C. difficile, has low systemic absorption, provides a prolonged post-antibiotic effect, and is currently approved for the treatment of CDAD. The efficacy and safety of low dose FDX (200mg once daily) for prevention of CDAD were evaluated in both autologous (AUTO) and allogenic (ALLO) HSCT patients receiving fluoroquinolone (FQ) prophylaxis. 

Methods:  HSCT patients were randomized to receive FDX 200 mg once daily (QD) or placebo (PLA) from start of conditioning therapy or FQ prophylaxis until 7 days after post-transplant engraftment (ANC >500/mm³) or completion of FQ prophylaxis.   Randomization was stratified by type of HSCT (AUTO vs. ALLO). Patients were contacted twice weekly for 30 days and then weekly through 60 days post-prophylaxis for symptoms; CDAD was confirmed by toxin testing or PCR.  The incidence of CDAD from start of prophylaxis up to 30 days and 60 days post-treatment were primary and secondary endpoints, respectively.   For the primary analysis, failures were defined as patients with confirmed CDAD, as well as those who had missing data for the primary endpoint assessment (due to death or discontinuations resulting from adverse events, protocol non-compliance, lost follow-up, or any other reasons) and those who received any CDAD-effective medication (without confirmed CDAD). Sensitivity analyses, including an analysis that defined failure as confirmed CDAD, were prospectively defined for the overall population.  These same analyses were performed for the ALLO and AUTO subgroups.

Results:  Of 611 enrolled patients, 600 were evaluable:  352 (59%) in AUTO and 248 (41%) in ALLO subgroups.  The primary analysis for the primary endpoint did not show a significant reduction in incidence of CDAD with FDX prophylaxis (28.6% FDX vs. 30.8% PLA, p = 0.28). Secondary endpoint results were similar.  However, the incidence of confirmed CDAD for up to 30 days was significantly lower in AUTO, ALLO, and overall in HSCT patients receiving prophylactic FDX versus PLA (see table). There were similar trends for 60 day post-treatment results.  Overall, serious adverse events (AEs) and AEs resulting in death were more frequent in the ALLO vs. AUTO subgroup.  In both subgroups, no differences in drug-related serious and non-serious AEs were seen between FDX and PLA.

Incidence of Confirmed CDAD, 30d Post-treatment
	AUTO	ALLO	Total
FDX (%)	5/176 (2.8%)	8/125 (6.4%)	13/301 (4.3%)
PLA (%)	14/176 (8.0%)	18/123 (14.6%)	32/299 (10.7%)
PLA-FDX, %	5.1	8.2	6.4
Wald p-value	0.0163	0.0166	0.0014
Incidence of Confirmed CDAD, 60d Post-treatment
FDX (%)	6/176 (3.4%)	11/125 (8.8%)	17/301 (5.6%)
PLA (%)	14/176 (8.0%)	18/123 (14.6%)	32/299 (10.7%)
PLA-FDX, %	4.5	5.8	5.1
Wald p-value	0.0321	0.0759	0.0117

Conclusions: For the primary endpoint (failure of prophylaxis due to confirmed CDAD or non-CDAD events), the efficacy of FDX and PLA were similar. However, in both AUTO and ALLO HSCT patients, prophylactic FDX significantly reduced the incidence of confirmed CDAD and was not associated with drug-related adverse events.