Fidaxomicin versus Conventional Antimicrobial Therapy in 59 Recipients of Solid Organ and Hematopoietic Stem Cell Transplantation with Clostridium difficile-Associated Diarrhea

Abstract

The feasibility of fidaxomicin versus vancomycin and metronidazole (conventional therapy) was assessed in 59 transplant recipients with 61 episodes of Clostridium difficile-associated diarrhea (CDAD). Overall clinical cure was achieved in 86% of episodes, and in 7% of episodes, infection recurred. Fidaxomicin was well tolerated. Clinical cures were not significantly different compared with conventional therapy (67% versus 89%, respectively; P = 0.06). Univariate analysis of predictors for lack of clinical cure included continued use of broad-spectrum systemic antibiotics (P = 0.026) and prior diagnosis of CDAD (95% confidence interval, 1.113 to 19.569; odds ratio, 4.667; P = 0.041). New-onset vancomycin-resistant Enterococcus (VRE) colonization was not noted after fidaxomicin therapy alone. However, this occurred in 10 of 28 patients (36%) following conventional therapy, and 2 of 3 patients with subsequent bacteremia died.

INTRODUCTION

Patients undergoing transplantation have an increased risk for Clostridium difficile-associated diarrhea (CDAD), and the response to conventional anti-CDAD therapy with oral vancomycin or metronidazole remains suboptimal (1–6). Fidaxomicin, a macrocyclic antibiotic with a limited spectrum of antimicrobial activity, became available in the United States in May 2011 (6–9). In transplant recipients, the clinical relevance of this narrow spectrum of antimicrobial activity and its influence on the preservation of the intestinal microbiota are uncertain (9). To help address this issue, we assessed the safety, clinical response, and intestinal and urinary bacterial and yeast colonization and its impact on subsequent invasive disease in transplant recipients treated with fidaxomicin or conventional therapy at our tertiary care university medical center.

(This study will be presented in part at the 53rd Interscience Conference on Antimicrobial Agents and Chemotherapy, Denver, CO, 10 to 13 September 2013.)

MATERIALS AND METHODS

All adult patients aged 18 years or older with CDAD who had undergone hematopoietic stem cell, liver, or kidney transplantation at the New York University Langone Medical Center between August 2007 and March 2013 were assessed retrospectively. The study was undertaken after obtaining approval from the institutional review board, and requirement for informed consent was waived. Patients were identified from the microbiology laboratory database for confirmed cases of C. difficile infection and from a registry of patients treated with fidaxomicin. Data were collected from electronic medical records. Since September 2011, fidaxomicin has been approved at our institution as the preferred treatment of primary CDAD episodes in patients undergoing transplantation. For all other patients, it was approved for secondary and salvage therapy.

Microbiology.

Only unformed stool samples were evaluated for the presence of C. difficile toxins. An enzyme immunoassay (EIA) for detection of C. difficile toxins A and B (Meridian Bioscience, Inc., Cincinnati, OH) was changed to a PCR assay (Xpert C. difficile; Cepheid, Sunnyvale, CA) in April 2010. Since June 2011, all stool samples with a positive C. difficile PCR had hypervirulent (North American pulsed-field type 1 [NAP1]) strain genotyping performed.

Definitions.

A diagnosis of CDAD required a positive C. difficile toxin assay and the presence of diarrheal illness. Each CDAD episode was considered mild to moderate if the white blood cell (WBC) count was <15,000 cells/mm³ and the serum creatinine (SCr) level was <1.5× the baseline value. Severe disease was defined as a WBC count of ≥15,000 cells/mm³ or an SCr level that was >1.5× the baseline value.

Patients were not included if they had septic shock, toxic megacolon, or a history of inflammatory bowel disease. Resolution of diarrhea was defined as well-formed or ≤3 unformed bowel movements per day for a 48-h period. CDAD recurrence was defined as diarrhea within 8 weeks of the end of therapy accompanied by a positive C. difficile toxin test.

Antimicrobial therapy.

Patients were given 200-mg fidaxomicin tablets twice daily. Patients who received 500 mg oral metronidazole three times daily, 125 mg oral vancomycin four times daily, or both were included in the control group.

Clinical outcomes.

The primary efficacy endpoint was clinical cure, defined as resolution of diarrhea, and no further need for anti-CDAD therapy. Secondary endpoints included (i) recurrence of CDAD; (ii) subsequent colonization and/or infection due to vancomycin-resistant Enterococcus (VRE); (iii) adverse events; (iv) drug-drug interactions, including the serum levels of the immunosuppressive agents and dose adjustment during fidaxomicin therapy; and (v) impact on graft-versus-host disease (GVHD) or graft rejection while on anti-CDAD therapy.

Statistical analysis.

Chi-square or Fisher exact tests were used to analyze categorical variables, and 2-tailed Student t tests and Mann-Whitney U tests were used to analyze continuous variables. Predictors of clinical failure were determined by univariate regression analysis.

RESULTS

Fifteen of 59 patients (25.4%) received fidaxomicin; for 7 patients, it was given as a salvage therapy after failure to achieve clinical cure following therapy with vancomycin, metronidazole, or both. Of patients who were treated with fidaxomicin, 5 underwent orthotropic liver transplantation (OLT), 4 had cadaveric kidney grafts, and 1 of 6 hematopoietic stem cell transplant (HSCT) recipients received an allogeneic donor graft. Forty-six treatment courses with conventional anti-CDAD therapy were given to 44 patients following OLT (n = 19), autologous HSCT (n = 22), and renal transplantation (n = 3). Metronidazole and oral vancomycin were given alone to 21 (46%) and 24 (52%) patients, respectively; a single patient (2%) received combination therapy. The mean duration of therapy for fidaxomicin was 11 days (range, 5 to 17 days), compared with 16 days (range, 6 to 35 days) among the conventional therapy group (P = 0.04). A single patient received <10 days of fidaxomicin therapy due to a lack of clinical response and was treated successfully with high-dose oral vancomycin with dose tapering.

Patient and disease characteristics are shown in Table 1. Conventional therapy was given to 83% of patients with a first CDAD episode, whereas fidaxomicin was used in this setting for 60% of episodes (P = 0.087). One patient had graft-versus-host disease. Patients who received fidaxomicin versus conventional therapy had more severe CDAD (27% versus 13%; P = 0.2), required fewer critical care unit stays (13% versus 24%; P = 0.49), had more hypervirulent (NAP1) C. difficile infection (43% versus 10%; P = 0.17), and continued to receive broad-spectrum systemic antimicrobial therapy (89% versus 67%; P = 0.1); however, these differences were not statistically significant.

Table 1.

Patient and disease characteristics and response to therapy in transplant recipients treated with fidaxomicin compared with controls given conventional vancomycin and metronidazole

Characteristic	Value for group			P value
	All (n = 59)	Fidaxomicin (n = 15)	Control (n = 44)
Mean age (yr) (SD)	56 (11)	57 (14)	57 (10)	0.98
No. (%) of male patients	31 (53)	6 (40)	25 (57)	0.26
No. (%) of patients with comorbidities	59 (100)
Cardiovascular disease	34 (58)	9 (60)	25 (57)	0.83
Chronic kidney disease	25 (42)	7 (47)	18 (41)	0.7
End-stage liver disease	25 (42)	5 (33)	20 (46)	0.41
Diabetes mellitus	20 (34)	4 (27)	16 (36)	0.49
Hematologic malignancy	25 (42)	6 (40)	19 (43)	0.83
Solid tumor	14 (24)	3 (20)	11 (25)	1
CDAD treatment courses	n = 61	n = 15	n = 46
No. (%) of patients on immunosuppressive therapya	45 (76)	11 (73)	34 (74)	1
No. (%) of patients on systemic antibiotics
Prior to CDAD diagnosis (30 days)	43 (71)	7 (47)	36 (78)	0.03
Fluoroquinolone	16 (26)	4 (57)	12 (33)	0.39
Cephalosporin	22 (36)	4 (57)	18 (50)	1
Penicillin	11 (18)	1 (14)	10 (28)	0.66
Carbapenem	4 (7)	1 (14)	3 (8.3)	0.52
During CDAD therapy	43 (71)	13 (87)	30 (65)	0.19
Cephalosporin	22 (36)	6 (46)	16 (53)	0.67
Penicillin	19 (31)	7 (54)	12 (40)	0.4
Carbapenem	10 (23)	4 (31)	6 (20)	0.46
Fluoroquinolone	10 (17)	3 (23)	7 (23)	1
No. (%) of patients tested by diagnostic assay (PCR)	40 (67)	14 (93)	26 (57)	0.002
No. of samples with NAP1 strain isolated/no. of NAP1 genotyping assays performed (%)	7/24 (29)	6/14 (43)	1/10 (10)	0.17
No. (%) of patients experiencing first episode of CDAD	47 (80)	9 (60)	38 (83)	0.087
No. (%) of patients with CDAD diagnosis after transplantb
Early (≤30 days)	28 (47)	6 (43)	22 (48)	0.74
Intermediate (≤1 yr)	16 (27)	3 (21)	13 (28)	0.74
Late (>1 yr)	16 (20)	5 (36)	11 (24)	0.49
Median MELD score prior to OLT (range)e	17 (6–34)	22 (12–27)	15 (6–34)	0.58
No. of cases of neutropenia in autologous HSCT recipients/total no. of HSCT recipients (%)c	12/30 (43)	1/6 (16)	11/24 (45)	0.36
Mean leukocyte count (cells per mm3) (range)	6 (0–60)	10 (0.9–31)	7.0 (0–60)	0.17
Median serum creatinine level (mg per dl) (range)	1.4 (0.4–7)	1.3 (0.7–4.6)	1.5 (0.4–7.0)	0.82
No. (%) of patients with feverd	28 (46)	6 (40)	22 (48)	0.6
Median no. of bowel movements during 24 h (range)	4 (1–10)	3 (2–6)	5 (1–10)	0.33
No. (%) of patients with CDAD severity graded as:
Mild to moderate	51 (84)	11 (73)	40 (87)	0.24
Severe	10 (16)	4 (27)	6 (13)	0.20
No. (%) of patients with critical care unit admission	13 (21)	2 (13)	11 (24)	0.49
Median length of intensive care unit stay (days) (range)	9 (3–133)	8 (6–9)	25 (3–133)	0.51
Median total length of hospitalization (days) (range)	18 (2–71)	19 (3–68)	17(2–71)	0.69
Mean time to resolution of diarrhea following CDAD therapy (days) (SD)	4 (1–25)	4 (1–10)	5 (1–25)	0.71
No. (%) of patients with clinical cure	51 (86)	10 (67)	41 (89)	0.06
No. (%) of patients with CDAD recurrence	4 (7)	1 (7)	3 (7)	1
No. (%) of patients with in-hospital mortality	3 (5)	0	3 (7)	0.57

^aImmunosuppressive therapy included systemic corticosteroids, tacrolimus, mycophenolate, cyclosporine, and interferon.

^bOne patient started fidaxomicin in the pre-OLT period, which was carried over after transplantation.

^cNeutropenia was defined as an absolute neutrophil count of <1,000 neutrophils/mm³.

^dFever was defined as an oral temperature of ≥100.4°F.

^eMELD, model for end-stage liver disease.

All 59 patients tolerated therapy without moderate-to-severe (grades II to IV) adverse events. The impact of fidaxomicin therapy on concomitant tacrolimus serum levels and variations in immunosuppressive drug dosing during CDAD therapy are shown in Fig. 1A and B.

Fig 1.

(A) Tacrolimus serum levels in 5 patients following orthotropic liver transplant and in a single cadaveric renal transplant recipient treated with fidaxomicin for CDAD. Shown are plasma concentrations of tacrolimus in serum samples of patients in the fidaxomicin group in relation to days of fidaxomicin treatment. The five patients were all solid organ transplant recipients (patient 1, kidney; patients 2 to 6, liver) with tacrolimus data available after fidaxomicin initiation. Values below the detectable threshold were reported as the value of the threshold during the relevant collection period: 2 ng/ml. Each data series ends when the patient was discharged or completed fidaxomicin therapy. Regarding dose adjustments during therapy, patient 1 was on experimental dosing of tacrolimus, and adjustment data are unavailable. The tacrolimus dosing was increased once for patient 2 and patient 5, increased twice for patient 6, decreased once for patient 3, and decreased twice for patient 4. (B) Tacrolimus serum levels for 7 orthotropic liver transplant recipients with CDAD who had received vancomycin and/or metronidazole. Shown are plasma concentrations of tacrolimus in relation to days of anti-CDI treatment for patients in the control group. The 13 patients are all OLT patients with tacrolimus data available after anti-CDI treatment initiation. The lower limit of tacrolimus detection during the relevant data collection period, which ranged between 2 and 4 ng/ml, was used as the tacrolimus level for data points reported to be below the detectable level. Each data series ends when the patient was discharged or completed anti-CDI treatment. Regarding dose adjustments during therapy, tacrolimus dosing was increased once and decreased once for patient 1; unchanged for patients 2, 4, 6, and 13; increased once for patients 3, 7, and 12; and increased twice for patient 5. Dose adjustment information was not available for patients 8 to 11.

Table 2 provides a univariate analysis of characteristics associated with a lack of clinical cure. Three of five patients for whom fidaxomicin therapy failed to produce clinical cure had severe orointestinal mucositis following HSCT, had prior diagnosis of CDAD, and had received fidaxomicin as salvage therapy. For all 5 patients, broad-spectrum antimicrobial therapy was continued during fidaxomicin therapy.

Table 2.

Univariate analysis of characteristics associated with probability of treatment failure in transplant patients with CDAD

Characteristic	No. (%) of patients with result		95% CI; OR (P value) determined by univariate analysis
	Clinical cure (n = 51)	Clinical failure (n = 10)
Systemic antibiotics during CDAD therapy	33 (65)	10 (100)	(0.026)a
Prior diagnosis of CDAD	9 (18)	5 (50)	1.113–19.569; 4.667 (0.041)
Late onset of disease	11 (22)	5 (50)	0.867–14.502; 3.545 (0.112)
Receipt of metronidazole therapy	20 (39)	1 (10)	0.020–1.465; 0.172 (0.143)

^aThe odds ratio cannot be determined, as the clinical failure rate was 100%.

The intestinal and urinary microbiota was assessed in 32 patients for up to 3 months after conventional therapy alone (n = 24), initial fidaxomicin therapy (n = 4), and initial salvage fidaxomicin therapy after extensive treatment with conventional therapy failed to achieve cure (n = 4). The reason for repeat stool cultures was persistent or recurrent diarrhea. Surveillance rectal or stool samples were not routinely obtained to assess VRE colonization. VRE colonization occurred only in patients exposed to conventional therapy (P = 0.28). Three patients developed bacteremia following VRE colonization, and of these patients, 2 OLT recipients died due to VRE bacteremia and sepsis. Candida species colonization was observed for 25% of patients in both groups (P > 0.5).

DISCUSSION

Fidaxomicin was well tolerated in transplant recipients at our center. Clinical responses and CDAD recurrences were comparable in patients given conventional therapy and those given fidaxomicin therapy. This was despite the presence of a higher but not significant representation of factors associated with CDAD treatment failure in the fidaxomicin treatment group (Table 1). Importantly, unlike patients treated with conventional therapy, those given fidaxomicin alone did not develop VRE colonization or invasive disease (10). Furthermore, in solid organ transplant recipients, fidaxomicin had no impact on tacrolimus dose and serum levels (Fig. 1A and B).

In two phase III trials, fidaxomicin resulted in a clinical response comparable to that of 10-day oral vancomycin therapy; however, the investigational drug was associated with significantly less risk for diseases recurrence 3 weeks after therapy was completed (11, 12). In a post hoc analysis of 1,164 patients enrolled in these two trials, persistent diarrhea, diarrhea recurrence, and deaths were reduced by 40% in the fidaxomicin-treated group (P < 0.0001) compared with the oral vancomycin group (13). In another meta-analysis, diarrhea due to C. difficile infection was an independent predictor of disease recurrence in the following 3 months (P < 0.01) (13). In our patients, CDAD recurrence rates were similar among patients treated with fidaxomicin and those treated with conventional therapy, although patients who received fidaxomicin had received this therapy more commonly in the setting of prior CDAD diagnosis than those given conventional therapy (Table 1). In another meta-analysis based on the two phase III trials, fidaxomicin and oral vancomycin resulted in comparable responses in patients with first disease recurrences, although patients with first CDAD recurrences treated with fidaxomicin had less risk for a second recurrence than did patients treated with oral vancomycin (14). On an extended follow-up, we observed no such difference, and this may reflect the limited number of patients with second recurrences treated in the present study.

Continued use of systemic broad-spectrum antibiotics during CDAD treatment is a well-recognized risk for clinical treatment failure and disease recurrence (1–15). In the 59 transplant recipients presented, the prevalence of exposure to systemic antibiotics was high and significantly associated with a lack of clinical treatment response (Tables 1 and 2). In a post hoc analysis of the 2 phase III trials (11, 12), antibiotics were continued in 27.5% of subjects during CDAD treatment, and compared with patients with no concurrent antibiotics, this alone was associated with a significantly low cure rate (84% versus 93%, respectively) and high rates of disease recurrence (24.8% versus 17.7%, respectively) (15). However, there was a difference in the responses of these patients: the cure rate was 90% among the fidaxomicin-treated group, whereas those treated with oral vancomycin had a significantly lower (79.4%) CDAD cure rate (P = 0.04) (15). There was also a 12% difference in recurrences favoring fidaxomicin over vancomycin therapy (15). In the two groups presented in Table 1, only one patient with a NAP1 CDAD infection developed recurrent disease after fidaxomicin therapy despite continued use of broad-spectrum antibiotics during CDAD therapy.

Despite the limited number of transplant patients in the fidaxomicin group and a smaller number of patients with follow-up intestinal colonization treated with fidaxomicin, this analysis provides potentially important information regarding current features of disease state, response to therapy, and outcome plus a possible reflection of the alteration in the microbiota among this highly susceptible population.

In this preliminary report, fidaxomicin was well tolerated, and we did not observe any unexpected adverse events or drug-drug interactions, including changes in tacrolimus dosing requirements or serum levels. The absence of intestinal tract colonization due to VRE following fidaxomicin therapy was cautiously reassuring, and larger, comprehensive, prospective microbiota assessment trials are needed and are under way.