Effectiveness of Anidulafungin in Eradicating Candida Species in Invasive Candidiasis

Abstract

In a phase 2 open-label, dose-ranging study in patients with candidemia, anidulafungin was effective in eradicating Candida albicans and other species of Candida. The anidulafungin MIC distribution showed that Candida albicans and C. glabrata were the most susceptible species and C. parapsilosis was the least susceptible species.

Anidulafungin is a semisynthetic echinocandin under development for the treatment of fungal infections, including invasive candidiasis. The most common species associated with candidemia remains Candida albicans (50 to 60% of cases), but non-C. albicans species, especially C. glabrata (approximately 20% of cases) followed by C. parapsilosis, C. tropicalis, and C. krusei, have emerged as significant pathogens (5, 7). The toxicity of polyenes and resistance to azoles have led to increased interest in alternative classes of antifungal agents, particularly those which target fungus-specific processes and promise to be well tolerated. Anidulafungin, like other echinocandins, targets the fungal cell wall, is fungicidal for all species of Candida, and is active against strains resistant to amphotericin B and azoles (4, 11, 13, 15, 17).

Anidulafungin was recently evaluated in an open-label, noncomparative, dose-ranging, phase 2 study in 123 adults with invasive candidiasis (8). We now report on the Candida isolates from the study, their susceptibilities, and the pathogen eradication rates obtained for different species and dosage regimens.

Study design.

Adult patients with candidemia (94% of total) and/or candidiasis at another normally sterile site (seven patients) were randomized to receive 50, 75, or 100 mg of intravenous anidulafungin daily (after a loading dose consisting of twice the daily dose on day 1) until 2 weeks after clinical cure or improvement (and eradication or presumed eradication of Candida) or treatment failure but not to exceed 42 days of treatment. A follow-up evaluation occurred 2 weeks after the end of therapy or at the time of treatment failure. In each group, 40 patients received anidulafungin, and a total of 68 patients (respectively, 18, 26, and 24 in the three groups) completed the study and were evaluable at follow-up.

The modified intent-to-treat population consisted of those patients who received at least one dose of anidulafungin and had Candida isolated at baseline (i.e., immediately preceding start of treatment). Patients with fungal infections other than those with Candida or with yeast fungemia not identified to species level were not included in this population. The primary endpoint of the study was response at follow-up in the evaluable population. This was the subset of the modified intent-to-treat patients who either completed the study with a sufficient length of treatment (≥10 doses) and no protocol violations that affected assessment of efficacy or were identified as failures by the investigator after at least 5 doses of anidulafungin.

Candida isolates.

Blood was cultured at baseline (before treatment), during treatment (on days 1, 2, and 6 and then once weekly), at the follow-up visit, and as clinically indicated. Cultures of other normally sterile sites were performed as clinically indicated. The majority of patients (94%) had candidemia, and only 10% had Candida identified at another normally sterile site (five patients had both blood and tissue sites positive). Isolates were presumptively identified at the local laboratories (30 participating centers in the United States) and then sent to a reference laboratory at the University of Iowa (Iowa City) for definitive identification to species level and MIC determinations by the Clinical and Laboratory Standards Institute (formerly NCCLS) broth microdilution method (12). All baseline isolates of Candida spp. (127 isolates) from all patients who received anidulafungin were tested against anidulafungin and four comparator antifungal agents: fluconazole, voriconazole, amphotericin B, and caspofungin. The MIC endpoint utilized for anidulafungin and caspofungin in this study was complete inhibition (MIC₀) after 48 h of incubation, a conservative parameter, which was considered standard at the time that the study was initiated (2). The MIC endpoint criteria for fluconazole, voriconazole, and amphotericin B were those recommended by the Clinical and Laboratory Standards Institute (12).

The distribution of baseline pathogens among species in the three treatment groups and the anidulafungin MIC statistics are shown in Table 1. Although this was a relatively small study, the frequency of isolation of different Candida species was similar to what has been reported in the literature for bloodstream infections (5, 7). The overall median anidulafungin MIC was 0.25 μg/ml. As seen with other echinocandins (6, 9, 10, 13), most MICs above 1 μg/ml were attributable to C. parapsilosis. MIC ranges and MICs at which 90% of the isolates tested are inhibited (MIC₉₀s) similar to those determined in this clinical trial have been reported from other in vitro studies when the same methodology has been used (1, 3, 6, 9, 13, 14, 19). MIC₉₀s generated with the more recently adopted prominent inhibition endpoint tend to be 1 to 2 dilutions lower, as seen in a recent survey of Candida bloodstream isolates (Table 1) (10).

TABLE 1.

Distribution of baseline Candida species and anidulafungin MICs

Species [no. of isolates (%)a]	nb	MICc range	MIC50	MIC90	MIC90 (n) in surveyd
All species [127 (100)]	114	0.03-8	0.25	2
C. albicans [62 (49)]	59	0.03-4	0.12	0.25	0.06 (500)
C. glabrata [36 (28)]	31	0.06-0.5	0.25	0.25	0.12 (105)
C. parapsilosis [11 (9.5)]	11	4-8	4	8	4 (106)
C. tropicalis [10 (8.6)]	7	0.12-2	0.5		0.06 (106)
C. krusei [5 (4.3)]	4	0.12-0.25			0.06 (23)
Othere [3 (2.4)]	2	0.12-0.5

^aAll isolates from the three treatment groups combined; microbiological intent-to-treat population (all patients who received anidulafungin and had Candida isolated at baseline).

^bBaseline isolates for which MICs were available; microbiological intent-to-treat population.

^cMICs in μg/ml; 100% inhibition for study isolates.

^dData from survey of >800 Candida bloodstream isolates (10); MICs in μg/ml, prominent inhibition endpoint.

^eOne each of C. dubliniensis, C. famata, and Candida sp.

MICs were also determined for comparators, including azoles, amphotericin B, and caspofungin (data not shown). Considering baseline isolates from all patients, all 4 C. krusei isolates, 16 of the 31 C. glabrata isolates, and 2 of 59 C. albicans isolates were nonsusceptible to fluconazole (MIC, >8 μg/ml). Additionally, one isolate of C. tropicalis had a caspofungin MIC of 16 μg/ml (2 μg/ml for anidulafungin); this strain was twice reisolated early in therapy but was then eradicated.

Eradication rates of Candida spp. in the 68 evaluable patients showed a dose-related trend, with 74, 85, and 89% success at 50, 75, and 100 mg/day of anidulafungin, respectively (Table 2). Among all three treatment arms, 61 of 73 baseline pathogens were eradicated or presumed to be eradicated (follow-up culture was unavailable for a patient with a clinical evaluation of success), and the majority of these (49 isolates) were documented by negative blood cultures. Eradication rates for the more commonly isolated individual species were comparable to the eradication rate for all species combined (Table 2). The patients' clinical and global response rates were similar to the eradication rates (8). Success was not related to the anidulafungin MIC for the isolates, and in particular, six of seven C. parapsilosis isolates, for which anidulafungin MICs were 4 to 8 μg/ml, were eradicated among the three study arms (Table 3). This lack of in vitro-in vivo correlation is not unusual in clinical trials of a new antifungal agent, largely due to the fact that the overall number of isolates tested, and specifically the number with true resistance to the new agent, is usually too small to demonstrate such a correlation (18). These results should be approached with caution, since evidence is pointing to the fact that the reading endpoints chosen for this study may not be optimal (16). Results and correlation with clinical response may be different if the 24-h prominent growth reduction endpoint is used (10, 16).

TABLE 2.

Eradication rates of the more frequently isolated Candida species

Species [no. of isolates (%)a]	No. eradicated/no. total (%)
	Daily dose (mg)			All doses
	50	75	100
All species [73 (100)]	14/19 (74)	23/27 (85)	24/27 (89)	61/73 (84)
C. albicans [34 (47)]	5/6 (83)	12/14 (86)	11/14 (79)	28/34 (82)
C. glabrata [21 (29)]	6/7 (86)	4/4 (100)	10/10 (100)	20/21 (95)
C. parapsilosis [7 (9.6)]	2/3 (67)	3/3 (100)	1/1 (100)	6/7 (86)
C. tropicalis [6 (8.2)]	0/1 (0)	3/3 (100)	2/2 (100)	5/6 (83)

^aEvaluable population at follow-up.

TABLE 3.

Response to therapy by MIC and daily dose of anidulafungin for baseline isolates of Candida spp.

MIC (μg/ml)	No. of isolates tested (n) and % success (%S) by anidulafungin dose (mg/day)
	50		75		100		All
	n	%S	n	%S	n	%S	n	%S
0.03			1	0			1	0
0.06			2	100	2	100	4	100
0.12	3	66	7	71	7	100	17	82
0.25	10	70	8	88	13	77	31	77
0.5	1	100	1	100	1	0	3	66
1
2			1	100	1	100	2	100
4	2	50	2	100	2	100	6	83
8	1	100					1	100
Unknowna	2	100	5	100	2	100	9	100

^aUnknown, MIC not determined.

In conclusion, in a phase 2 open-label, noncomparative, dose-ranging study, anidulafungin was effective in eradicating Candida spp. from the bloodstream and from other normally sterile sites. The eradication rate was high for different frequently encountered species including C. parapsilosis. A trend toward dose dependence was observed, and the highest daily dose tested (100 mg/day) is being evaluated in a phase 3 comparative study.