Diagnostic Microbiology and Infectious Disease 39 (2001) 2531

www.elsevier.com/locate/diagmicrobio

Mycology

Clinical applicability of antifungal susceptibility testing on non-Candida albicans species in hospitalized patients
Annie Wong-Beringera,*, Janet Hindlerb, Lilliana Brankovicb, Lisa Muehlbauerc, Lynn Steele-Moorec
a

Western University of Health Sciences, College of Pharmacy, 309 East Second Street, Pomona, CA 91766 b UCLA Clinical Microbiology, 11633 San Vicente Blvd., 4th oor, Los Angeles, CA 90049 c Christiana Care Infectious Disease Laboratory, 501 West 14th Street, Wilmington, DE 19801

Abstract We assessed the distribution, antifungal susceptibility, and treatment associated with 161 non-Candida albicans isolates recovered from hospitalized patients over a 6-month period. The three most prevalent species were C. glabrata (100), C. tropicalis (28), and C. krusei (15). Resistance of C. glabrata to uconazole and itraconazole were 6% and 17% respectively; 80% of the uconazole-resistant isolates were cross-resistant to itraconazole. Prior azole exposure signicantly reduced azole susceptibility in C. glabrata and also affected its subsequent selection among colonized patients. Only 21% of the patients had clinical infections. Patients with fungemia were more likely to be treated with amphotericin versus an azole. Overall treatment success was higher in patients treated with amphotericin versus an azole (56% vs 31%). Routine susceptibility testing on all Candida species does not appear necessary except where therapy with an azole is being considered to detect resistant isolates or for epidemiologic surveillance purposes. Further studies are needed to delineate the relationship between azole MICs and treatment outcomes of invasive candidiasis due to non-C. albicans species. 2001 Elsevier Science Inc. All rights reserved.

1. Introduction Nosocomial infections due to Candida species have increased dramatically since the 1980s. Data based on a national surveillance program on bloodstream infections (BSI) conducted between April 1995 and June 1996 indicates that Candida is the fourth leading cause of nosocomial BSI (Pfaller et al., 1996). While C. albicans continues to be the predominant species responsible for BSI, non-C. albicans species are increasing in proportion (Abi-Said et al., 1997; Wingard et al., 1991). One multicenter prospective study of candidemia observed a steady rise in non-C. albicans as the cause from 40% to 53% over a 3.5 year period (Nguyen et al., 1996). Whether this observed shift is attributable in part to the widespread use of imidazoles, in particular uconazole, remains controversial (White, 1997). While the incidence of invasive disease is increasing and the epidemiology appears to be shifting, diagnostic approaches for early detection of invasive candidiasis are less than optimal at present. Hence, prophylactic and empirical
* Corresponding author. Tel.: 1-909-469-5580; fax: 1-626-6283024. E-mail address: [email protected] (A. Wong-Beringer).

antifungal treatment strategies are commonly employed in oncology or surgical patients at risk for invasive candidiasis. The availability of uconazole and itraconazole in the early 1990s has provided attractive therapeutic options in antifungal therapy. The favorable pharmacokinetic and safety proles of uconazole and itraconazole are appealing when compared to amphotericin B. However, the activities of uconazole and itraconazole are highly variable against Candida species. In one study, uconazole demonstrated the following rank order by MIC value against Candida species: C. albicans C. parapsilosis C. glabrata C. krusei (Rex et al., 1995). Candida tropicalis was noted to have bimodal distribution in that study with MICs at 0.5 g/mL and 64 g/ml. Superinfections with C. glabrata and C. krusei have been demonstrated in neutropenic patients in whom uconazole was used for antifungal prophylaxis (Wingard et al., 1991; Winston et al., 1993). However, despite reports of clinical failures with uconazole in the treatment of infections due to C. glabrata, some clinicians are reluctant to use amphotericin B for fear of its nephrotoxicity, particularly in patients with pre-existing renal dysfunction. In addition, despite the lower risk of nephrotoxicity associated with the lipid-formulated amphotericin B products, some clinicians are reluctant to use these agents

0732-8893/01/$ see front matter 2001 Elsevier Science Inc. All rights reserved. PII: S 0 7 3 2 - 8 8 9 3 ( 0 0 ) 0 0 2 0 9 - 1

26

A. Wong-Beringer et al. / Diagnostic Microbiology and Infectious Disease 39 (2001) 2531

given the high drug acquisition costs compared to the conventional formulation. Unlike the practice with antibacterial agents, antifungal susceptibility testing has not been used as a guide for antifungal therapy in the past due to the lack of standardization of testing methodology. Signicant work done in this area has resulted in a standard methodology (document M27-A) adopted by the National Committee on Clinical Laboratory Standards (NCCLS) recently for the testing of uconazole, itraconazole, amphotericin B, and 5-ucytosine against Candida species and Cryptococcus neoformans (NCCLS 1997). In addition, NCCLS has established and adopted susceptibility breakpoints in the categories of susceptible, susceptible-dose-dependent, and resistant for uconazole and itraconazole against Candida species. Outcome evaluation with uconazole MIC for non-C. albicans infections and for invasive candidiasis are limited at this point while none are available for itraconazole (Rex et al., 1997). As for amphotericin B, clinically useful susceptibility breakpoints continue to be the subject of ongoing studies (Clancy et al., 1999; Nguyen et al., 1998). We performed a prospective observational study to evaluate the clinical applicability of antifungal susceptibility testing at a large tertiary care medical center over a 6-month period. The primary study objectives were to 1) identify the distribution and the antifungal susceptibility prole of each species of Candida other than C. albicans and 2) examine the effect of antecedent azole exposure on subsequent colonization or infection with non-C. albicans and susceptibility to the azoles in those isolates. In addition, a secondary objective of the study was to evaluate the therapeutic interventions and resultant outcomes in the subset of patients with clinical infections due to a non-C. albicans species.

tested with the API 20 C AUX yeast identication kit (bioMerieux, Hazelwood, MO) following the manufacturers directions to determine its identity. 2.2. Susceptibility testing Susceptibility testing was performed at the Christiana Care Infectious Disease Laboratory by a broth microdilution method based on the NCCLS guideline (NCCLS 1997). Sterile, plastic 96-well microtiter trays (Falcon) with roundbottom wells were used to prepare panels in house for susceptibility testing. The microtiter trays contained drug dilutions in RPMI 1640 with MOPS buffer (BioWhittaker, Walkersville, MD) at the following concentration ranges: amphotericin B (Bristol-Myers Squibb, Princeton, NJ) at 0.032.0 g/ml; uconazole (Pzer, New York, NY) at 0.12 64.0 g/ml; itraconazole (Janssen Pharmaceutica, Titusville, NJ) at 0.0316.0 g/ml. Organisms received in the Christiana Care Infectious Disease Laboratory were subcultured onto Sabouraud dextrose agar three times to ensure purity and viability. Plates were incubated at 35C. The test inoculum was prepared by selecting ve well isolated colonies from 24 h old subcultures and suspending them in 5ml of sterile 0.145 mol/l saline. The resulting suspension was vortexed 15 seconds and turbidity adjusted spectrophotometrically to reach a 0.5 McFarland Standard (530nm wavelength) which yielded a stock suspension of 1 1065 106 cells/mL. This was then diluted from 1:50 to 1:20 with RPMI to obtain two times test inoculum (1 1035 103 CFU/ml). This was further diluted 1:2 when the wells were inoculated to achieve the desired nal inoculum size of 0.5 103 to 2.5 103 CFU/ml. Quality control strains C. parapsilosis ATCC 22019 and C. krusei ATCC 6258 were included with each batch of MIC testing and results were within acceptable ranges (NCCLS M27-A). Inoculated microdilution trays were incubated at 35C for 24 hrs. Microdilution wells were scored with the aid of a reading mirror, the growth in each well was compared with that of the growth in the control well. Trays were then reincubated and read again at 48 hrs. The 48 h MIC provided the best agreement with the reference broth macrodilution method (NCCLS M27-A) (NCCLS 1997). The MIC for amphotericin was dened as the lowest concentration of the tested drug that completely inhibited growth at 48 hrs. The MIC for uconazole and itraconazole was dened as the lowest concentration in which a score of 2 (prominent reduction of growth) was observed at 48 hrs as compared to the growth control well. NCCLS approved interpretive criteria were used to dene susceptible, susceptible-dosedependent, and resistant for all strains (Table 1). Isolates which were resistant to one azole were compared to the other to determine the pattern of cross-resistance for each isolate.

2. Materials and methods 2.1. Isolates A prospective 6-month culture surveillance for non-C. albicans species isolated from any site was conducted at UCLA Medical Center on all patients hospitalized during the period of December 1997 to May 1998. Only the rst isolate per patient was used for susceptibility testing. One hundred sixty-one yeast strains were recovered from 161 patients. Yeasts were identied by the UCLA Clinical Microbiology Laboratory using the following methods. The germ tube test was used as the rst step and if positive, the isolate was called Candida albicans. Any isolate that was germ tube negative was inoculated to Cornmeal Tween 80 agar (Hardy Diagnostics, Santa Maria, CA). Growth was examined microscopically to note the appearance of any pseudohyphae and blastoconidia that would be characteristic for Candida tropicalis, Candida parapsilosis, or Candida krusei. If any observation was equivocal or not characteristic of the aforementioned species, the isolate was

A. Wong-Beringer et al. / Diagnostic Microbiology and Infectious Disease 39 (2001) 2531 Table 1 In vitro susceptibility of non-Candida albicans species to uconazole, itraconazole, and amphotericin B. Candida species No. isolates tested 100 28 15 8 6 3 1 Fluconazole MIC ( g/ml) 90 16 1.0 NA 1.0 1.0 4.0 0.25 Range 0.25128 0.12128 0.121.0 0.121.0 1.04.0 % Fluc R (No. of isolates) 6 (6) 3.6 (1) NA 0 0 0 0 Itraconazole MIC ( g/ml) 90 1.0 0.25 0.5 0.06 0.12 1.0 0.12 Range 0.034.0 0.030.5 0.031.0 0.030.06 0.030.12 0.061.0 % Itrac R (No. of isolates) 17 (17) 0 13 (2) 0 0 33 (1) 0

27

Amphotericin MIC ( g/ml) 90 1.0 1.0 2.0 1.0 2.0 1.0 2.0 Range 0.252.0 0.51.0 1.02.0 0.51.0 0.252.0 0.51.0

C. C. C. C. C. C. C.

glabrata tropicalis krusei parapsilosis lusitaniae guillermondii kefyr

R resistance dened based on the interpretive criteria (M27-A, NCCLS 1997); NA susceptibility testing is not applicable due to inherent resistance to uconazole. NCCLS breakpoints are not dened for amphotericin B. NCCLS Interpretive breakpoints ( g/mL) for Candida species against the indicated agents: Antifungal Agent Fluconazole Itraconazole Susceptible (S) 8 0.125 Susceptible-dose dependent (S-DD) 1632 0.250.5 Intermediate (I) Resistant (R) 64 1

2.3. Clinical Correlation Patients in whom a non-C. albicans species was isolated from sterile sites dened as sites which were not commonly colonized were selected for chart reviews. Those sites included blood, other sterile sites (i.e. peritoneal uid), and urine. Despite the fact that urine of hospitalized patients are often colonized by yeast in the presence of urinary catheters, infection vs colonization is a clinical dilemma due to the difculty in diagnosis; thus, those patients were included in the sterile sites for further evaluation via chart review. Notably, multiple Candida sp. were isolated from some of the urine cultures which could be a signicant confounding problem in the diagnosis of infection. Ninety-three isolates of non-C. albicans were obtained from sterile sites from 93 patients; medical charts were available for review in 73 patients. The following information was obtained from the medical charts: patient demographics, underlying disease, antecedent use of uconazole or itraconazole within 30 days prior to the isolation of non-C. albicans species, presence of infection, antifungal therapy and treatment outcome for those infected with non-C. albicans, length of stay, and mortality. 2.4. Denition Infection was dened as the isolation of non-C. albicans species from sterile body sites PLUS local and systemic signs and symptoms of infection (leukocystosis, fever). Other sterile sites where yeast colonization or infection was equally likely depending upon various host and environmental factors had the following denitions for infection: 1) oropharyngeal or esophageal candidiasis: white plaques on oral or endoscopic examination PLUS positive

cultures, 2) pneumonia: presence of non-albicans Candida species in respiratory specimens without the presence of bacterial ora PLUS histopathologic evidence of disease as demonstrated by tissue biopsy, inltrates on chest x-ray, and systemic signs and symptoms of infection (leukocytosis, fever), and 3) candiduria: presence of non-albicans Candida species in the urine PLUS either local signs and symptoms of infection (dysuria, urgency) OR an indwelling urinary catheter and unexplained fever. We did not include a colony count threshold in our denition for candiduria since the organism burden did not consistently predict symptomatic disease (Wong-Beringer et al., 1992). Treatment outcome was determined by clinical response as dened by the following: 1) complete response: resolution of fever, leukocytosis, and local signs of infection, 2) partial response: improvement of fever, leukocystosis, and local signs of infection without complete resolution, 3) relapse: recurrence of infection with the same organism at any body site within 1 month of discontinuation of therapy, and 4) failure: absence of resolution or worsening of signs and symptoms of infection or death. 2.5. Statistical Analysis Chi-square, Fishers exact, and unpaired t-test were used where applicable. Statistical signicance was dened by a P value of less than or equal to 0.05. Relative risk was calculated to determine the effect of prior azole exposure on subsequent colonization or infection with non-C. albicans isolates. All statistical analyses were performed using GraphPad Prism version 3.00 for Windows, GraphPad Software, San Diego California USA, www.graphpad.com.

28

A. Wong-Beringer et al. / Diagnostic Microbiology and Infectious Disease 39 (2001) 2531 Table 2 Demographics for Patients with Sterile Sites Positive for Non-C. albicans Characteristics Group 1 Colonized N 39 32/7 58 (1483) 19 (2244) 8 (21%)* 4 (10%)** 17 (44%) 10 (26%) 6 (15%) 6 (15%) 0.0022, Chi-square. Group 2 Infected N 34 17/17 55 (0.582) 26 (1158) 15 (44%)* 14 (41%)** 8 (24%) 10 (29%) 2 (6%) 14 (41%)

Fig. 1. Frequency Distribution of Fluconazole MICs for C. glabrata (n 100). Median MIC 2.0 g/ml.

Gender (M/F) Age, median years (range) LOS, median days (range) Expired Sepsis-related Underlying disease Transplant recipients (1 BMT, 16 solid-organ) Malignancy GI/GU surgery Others * p 0.03, Chi-square; ** p LOS length of stay

3. Results 3.1. Distribution and Antifungal Susceptibility One hundred sixty-one isolates of non-C. albicans species were recovered from 161 patients hospitalized during the six month study period. The distribution of species were in the order of frequency: C. glabrata (100), C. tropicalis (28), C. krusei (15), C. parapsilosis (8), C. lusitaniae (6), C. guillermondii (3), and C. kefyr (1). The three most frequently isolated non-C. albicans species accounted for 89% of all the isolates recovered. Table 1 shows the susceptibility prole for each non-C. albicans species against amphotericin B, uconazole, and itraconazole. Figs. 1 and 2 depict the frequency distribution of uconazole and itraconazole MICs for C. glabrata isolates. Using the NCCLS approved MIC interpretive criteria, Candida glabrata was the second most resistant species to the azoles following C. krusei; however, more isolates were resistant to itraconazole than with uconazole (17% vs 6%). Of note, C. krusei is innately resistant to uconazole and should not be tested in vitro against this agent. Thus, NCCLS uconazole breakpoints do

not apply to C. krusei. Only one isolate of C. tropicalis was resistant to uconazole at MIC of 128 mcg/ml; the isolate was obtained from a patient who had received itraconazole prophylaxis for over 45 days. Data on cross-resistance differed between the azole agents (data not shown). Notably for C. glabrata, 80% (4/5) of the uconazole-resistant isolates were cross-resistant to itraconazole whereas only 24% (4/17) of itraconazole-resistant isolates were cross-resistant to uconazole. As for C. tropicalis and other non-C. albicans species, the limited number of resistant isolates precluded any meaningful pattern of cross-resistance between the azoles. 3.2. Effect of Antecedent Azole Exposure Medical charts were available for review for 73 patients in whom a non-C. albicans species was isolated from sterile sites. Table 2 depicts the patient demographics. According to study denitions, 39 patients were colonized while 34 were clinically infected with a non-C. albicans species. The median age was similar in both groups (58 vs 55 years). Colonized patients were predominantly male while the infected patients had an equal number of male versus female. Notably, infected patients had a longer length of hospital stay (26 vs 19 days, p 0.05) when compared to the colonized patients. In addition, the rates of overall and sepsis-related mortality were signicantly higher among infected versus colonized patients. Organ or bone marrow transplantation and malignancy were the major underlying conditions in those patients; no statistically signicant difference was found for colonized versus infected patients with respect to underlying conditions. Prior azole exposure was documented in 30% (22/73) of all patients colonized or infected with non-C. albicans species. Itraconazole was twice as frequently prescribed as uconazole primarily for prophylaxis. The mean daily dose

Fig. 2. Frequency Distribution of Itraconazole MICs for C. glabrata (n 100). Median MIC 0.25 g/ml.

A. Wong-Beringer et al. / Diagnostic Microbiology and Infectious Disease 39 (2001) 2531 Table 3 Antecedent azole exposure on subsequent non-C. albicans species isolation Species Colonized Prior Azole n 12 C. glabrata Non-C. glabrata C. tropicalis C. parapsilosis C. krusei Others** 11 (92%)* 1 (8%) 1 Non prior n 27 15 (55%)* 12 (45%) 6 2 1 3 Total N 39 26 (67%) 13 (33%) 7 2 1 3 Infected Prior Azole n 10 5 (50%) 5 (50%) 2 2 1 None prior n 24 11 (46%) 13 (54%) 7 3 1 2

29

Total N 34 16 18 9 3 3 3

* p 0.027, Chi-square ** Others include C. lusitaniae, C. kefyr, and C. guillermondii

and duration for itraconazole and uconazole were 6.5 mg/kg and 66 days versus 5.1 mg/kg and 17 days, respectively. Recipients of organ or bone marrow transplantation were more likely to have had received prior azole therapy compared to the overall patient population (18/22, 82% vs 22/73, 30%, respectively). In addition, they were more likely to be colonized than infected. Prior exposure to an azole appeared to have an effect on the subsequent isolation of C. glabrata among the colonized patients but not the infected patients. (Table 3) Patients who had had prior azole exposure had a 5.5 times risk of being colonized with C. glabrata compared to those without azole exposure (condence interval, 0.79 38.13). In addition, the susceptibility of C. glabrata to uconazole was signicantly reduced in patients with prior azole exposure compared to patients without (mean MIC 18 vs 3.5 mcg/ml, p 0.0045); the effect was less pronounced with itraconazole (mean MIC 0.89 vs 0.33 mcg/ml, p 0.02). (Table 4) As for the other non-C. albicans species, the small number of isolates and the number of patients who received prior azole therapy precluded meaningful comparison between groups. 3.3. Treatment Outcome Of those 34 patients who had clinical infections due to non-C. albicans species, 29 were deemed evaluable for treatment outcome; the remaining 5 patients either received topical antifungal therapy or none at all. Table 5 describes the drugs used, indications, and outcomes of therapy. Patients with fungemia were signicantly more likely to have

been treated initially with amphotericin B compared to an azole (11/16, 68% vs 4/13, 30%, p 0.04); two other patients with fungemia who were initially treated with an azole were switched to amphotericin B after 72 hrs. Thus, the majority of patients who were prescribed an azole agent either had focal organ involvement or had bacteria isolated from the site of infection confounding the etiologic role of the fungi. Treatment success was observed in 56% (9/16) vs 31% (4/13) of patients receiving amphotericin versus an azole, respectively. It is noteworthy that nearly a third of the azole-treated patients were considered non-evaluable due to intercurrent bacterial infections. All amphotericin B failure patients (n 7) expired secondary to sepsis while only 22% (2/9) in the successfully treated group expired (both without evidence of fungal infection at the time of death). There was no apparent relationship between amphotericin B or azole MIC and treatment outcomes. Of note, amphotericin MICs were between 0.5 and 1.0 g/ml in all but 2 patients; those patients failed treatment and had isolates with MICs of 2.0 g/mL to amphotericin B. The MICs for all isolates obtained from the azole-treated patients were within the susceptible range of the treating agent according to the NCCLS interpretive criteria.

4. Discussion According to one large epidemiologic surveillance study, the distribution of Candida bloodstream isolates in the United States between 1992 to 1998 indicates that the proportion of non-C. albicans species ranged between 40 55% (Pfaller et al., 1998). Among the non-C. albicans isolates, the following species were found in order of decreased frequency: C. glabrata, C. parapsilosis, C. tropicalis, C. krusei, and others (Pfaller, et al., 1998, 1999). While the distribution and susceptibility of non-Candida species causing nonsocomial bloodstream infections have been evaluated, few reported in vitro susceptibility of non-C. albicans isolated from other sites. The distribution of non-C. albicans species isolated from various clinical specimens over a 6-month period in this study showed a similar pattern to that

Table 4 Effect of prior azole use on the susceptibility of Candida glabrata to the azoles Species/ Azole C. glabrata Fluconazole Itraconazole Prior Exposure mean MIC, g/ml (range) N 15 18 (2.0 64) 0.89 (0.124.0) None prior mean MIC, g/ml (range) N 26 3.5 (1.08.0) 0.35 (0.032.0) P value, Unpaired t-test p p 0.0045 0.02

30

A. Wong-Beringer et al. / Diagnostic Microbiology and Infectious Disease 39 (2001) 2531

Table 5 Treatment indication and outcome Drug No. of patients Indication Outcomes Success Amphotericin 16 Fungemia (11) Focal organ (5) Abscess; peritonitis; pneumonia Azole 13 Fungemia (4) Focal organ (9)* UTI; necrotizing wound; T-tube abscess; tracheobronchitis; pneumonia Outcomes: success complete or partial response; non-evaluable included patients with intercurrent bacterial infections. * 3 patients had bacteria isolated from the same site and had received concurrent antibacterial therapy and 1 had bronchopleural aspergillosis 4 (31%) 5 (38%) 4 (31%) 9 (56%) Failure 7 (44%) Non Evaluable

reported in the literature. C. glabrata was the leading species followed by C. tropicalis, C. krusei, C. parapsilosis, and others. Using the microbroth dilution method for antifungal susceptibility testing and interpretive criteria as recommended by NCCLS, we found that C. glabrata was the second most resistant Candida species to the azoles following C. krusei; more isolates were resistant to itraconazole compared to uconazole. In addition, the pattern of crossresistance for C. glabrata indicated that the majority (80%) of uconazole-resistant isolates were also resistant to itraconazole whereas only 24% of the itraconazole resistant isolates were cross-resistant to uconazole. Considering the above, it would be important for clinicians to recognize that a non-Candida albicans species that tested resistant to uconazole is highly likely to be resistant to itraconazole as well. Conversely, uconazole may remain a viable treatment option for patients in whom an itraconazole-resistant Candida species was isolated. A second objective of our study was to examine the effect of azole exposure on subsequent colonization and infection with non-C. albicans. We evaluated a subset of 73 patients in whom Candida was isolated from sterile sites. It appeared that prior exposure to azole alone could not account for the selection of non-C. albicans isolates in our patients; less than a third (30%, 22/73) of the patients received an azole agent within 30 days prior to the isolation of Candida. However, patients who did receive an azole (mostly transplant recipients for prophylaxis) were more likely to be colonized than infected suggesting a protective effect of antifungal prophylaxis on infection. C. glabrata was the predominant non-C. albicans species isolated from colonized patients who had prior azole exposure. This phenomenon could be explained in part by our observation that antecedent azole exposure resulted in signicantly reduced susceptibility of C. glabrata to uconazole and itraconazole. Among the 161 patients in whom a non-C. albicans species was isolated, only 21% (34/161) had clinical evidence of infections. In those patients, it appeared that clinicians were reluctant to use an azole for the treatment of

disseminated candidiasis or fungemia due to a non-C. albicans species. Patients who were treated with an azole were primarily those who had focal organ involvement or had intercurrent bacterial infections where the causative role of fungi was uncertain. As for the patients who were colonized with non-C. albicans species (n 39), one third of those patients continued to have isolations of the same Candida species for weeks to months after the initial culture of the urine or bronchoalveolar lavage. Our study had several limitations. The large number of non-Candida albicans isolates with which antifungal susceptibility testing was performed in this study was obtained from a single-center; thus, results of the susceptibility testing may not be applicable in all settings. While we were able to determine the effect of azole exposure on the subsequent colonization and susceptibility of C. glabrata relative to other non-C. albicans species, we could not state with certainty the relationship between azole exposure and subsequent selection for non-C. albicans species without a control group of patients who had had prior azole exposure and were subsequently colonized or infected with C. albicans. In addition, the number of patients treated with an azole agent for disseminated candidiasis or fungemia due to non-C. albicans species was limited and thus, precluded meaningful evaluation of treatment outcome. Despite variable activity of the azole compounds against non-C. albicans species, routine antifungal susceptibility testing on all clinical specimens is unnecessary given the low overall rate of resistance and the low level of clinical signicance when isolated. However, it is important for individual institutions to determine the distribution of Candida species and susceptibility patterns periodically, particularly where the use of azoles for antifungal prophylaxis is common practice. Amphotericin remains the treatment of choice for invasive candidiasis due to non-C. albicans until further studies can delineate the relationship between azole MIC and treatment outcome. For now, the use of an azole compound for the treatment of candidiasis due to C. glabrata should be guided by results from antifungal suscep-

A. Wong-Beringer et al. / Diagnostic Microbiology and Infectious Disease 39 (2001) 2531

31

tibility testing particularly on individuals who have had prior azole exposure.

References
Abi-Said, D., Anaissie, E., & Uzun, O., et al. (1997). The epidemiology of hematogenous candidiasis caused by different Candida species. Clin Infect Dis 24, 1122 8. Clancy, C. J., & Nguyen, M. H. (1999). Correlation between in vitro susceptibility determined by E test and response to therapy with amphotericin B: results from a multicenter prospective study of candidemia. Antimicrob Agents Chemother 43, 1289 90. National Committee for Clinical Laboratory Standards. (1997). Reference method for broth dilution antifungal susceptibility testing of yeast: approved standard. NCCLS document M27-A. Wayne, PA: National Committee for Clinical Laboratory Standards. Nguyen, M. H., Peacock, J. E., & Morris, A. J., et al. (1996). The changing face of Candidemia: emergence of non-Candida albicans species and antifungal resistance. Am J Med 100, 61723. Nguyen, M. H., Clancy, C. J., & Yu, V. L., et al. (1998). Do in vitro susceptibility data predict microbiologic response to amphotericin B? Results of a prospective study of patients with Candida fungemia. J Infect Dis 177, 42530. Pfaller, M. A., Jones, R. N., & Messer, S. A., et al. (1998). National surveillance of nosocomial bloodstream infection due to species of Candida other than Candida albicans: frequency of occurrence and

antifungal susceptibility in the SCOPE program. Diagn Microbiol Infect Dis 30, 1219. Pfaller, M. A., Messer, S. A., & Hollis, R. J., et al. (1999). Trends in species distribution and susceptibility to uconazole among blood stream isolates of Candida species in the United States. Diagn Microbiol Infect Dis 33, 21722. Rex, J. H., Pfaller, M. A., & Barry, A. L., et al. (1995). Antifungal susceptibility testing of isolates from a randomized multicenter trial of uconazole versus amphotericin B as treatment of non-neutropenic patients with candidemia. Antimicrob Agents Chemother 39, 40 4. Rex, J. H., Pfaller, M. A., & Galgiani, J. N., et al. (1997). Development of interpretive breakpoints for antifungal susceptibility testing: conceptual framework and analysis of in vitro-in vivo correlation data for uconazole, itraconazole, and Candida infections. Clin Infect Dis 24, 235 47. White, M. H. (1997). The contribution of uconazole to the changing epidemiology of invasive candidal infections. Clin Infect Dis 24, 1129 30. Wingard, J., Merz, W. G., & Rinaldi, M. G., et al. (1991). Increase in Candida krusei infection among patients with bone marrow transplantation and neutropenia treated prophylactically with uconazole. N Engl J Med 325, 1274 7. Winston, D. J., Chandrasekar, P. H., & Lazarus, H. M., et al. (1993). Fluconazole prophylaxis of fungal infections in patients with acute leukemia. Results of a randomized placebo-controlled, double-blind multicenter trial. Ann Intern Med 118, 495503. Wong-Beringer, A., Jacobs, R. A., & Guglielmo, B. J. (1992). Treatment of funguria. JAMA 267, 2780 2785.