Previous Article | Next Article ![]()
Antimicrobial Agents and Chemotherapy, January 2003, p. 34-38, Vol. 47, No. 1
0066-4804/03/$08.00+0 DOI: 10.1128/AAC.47.1.34-38.2003
Copyright © 2003, American Society for Microbiology. All Rights Reserved.
Department of Internal Medicine, Wayne State University, Detroit, Michigan,1 Department of Internal Medicine, William Beaumont Hospital, Royal Oak, Mich.,2 Department of Family Medicine, University of Michigan, Ann Arbor, Michigan,3 Department of Internal Medicine, University of Washington, Seattle, Washington,4 Department of Obstetrics and Gynecology, Medical College of South Carolina, Charleston, South Carolina,5 Department of Obstetrics and Gynecology, Jefferson Medical College, Philadelphia, Pennsylvania,6 Department of Obstetrics and Gynecology, University of Arizona, Tucson, Arizona,7 Scripps Clinic, La Jolla, California,8 Pfizer Pharmaceutical Trials TAC, New York, New York9
Received 22 February 2002/ Returned for modification 15 May 2002/ Accepted 3 October 2002
|
|
|---|
64 µg/ml). Following fluconazole therapy, one case of in vitro resistance developed during 6 weeks of monitoring. In accordance with the NCCLS definition, in vitro fluconazole resistance correlated poorly with the clinical response, although a trend of a higher mycological failure rate was found (41 versus 19.6% on day 14). By using an alternative breakpoint of 1 µg/ml, based upon the concentrations of fluconazole achievable in vaginal tissue, no significant differences in the clinical and mycological responses were observed when isolates (n = 250) for which MICs were
1 µg/ml were compared with isolates (n = 30) for which MICs were >1 µg/ml, although a trend toward an improved clinical outcome was noted on day 14 (odds ratio, >2.7; 95% confidence interval, 0.91, 8.30). Although clinical failure was uncommon, symptomatic recurrence or mycological relapse almost invariably occurred with highly sensitive strains (MICs, <1.0 µg/ml). In vitro fluconazole resistance developed in 2 of 18 initially susceptible C. glabrata isolates following fluconazole exposure. Susceptibility testing for women with complicated Candida vaginitis appears to be unjustified. |
|
|---|
|
|
|---|
Statistical methods.
The clinical and mycological responses at days 14 and 35 between patients for whose isolates the baseline MIC was
1 µg/ml and patients for whose isolates the baseline MIC was >1 µg/ml dose were compared statistically by the logistic regression method. A logistic regression model controlling for the severity of vaginitis and the treatment was used to assess the association of pretreatment MICs with the clinical and mycological outcomes at days 14 and 35.
All tests were two tailed, and P values of <0.05 were considered statistically significant.
|
|
|---|
MICs for C. albicans. (i) Baseline isolates.
A total of 393 baseline isolates of C. albicans were tested (Table 1). By using the susceptibility and resistance definitions widely accepted for both systemic and mucosal infections (9), 377 (96%) of the C. albicans isolates were susceptible to fluconazole (MICs,
8 µg/ml), 2 (0.4%) were considered dose-dependent susceptible (DD-S; MICs, 16 to 32 µg/ml), and 14 (3.6%) were resistant (MICs,
64 µg/ml). The MIC at which 50% of isolates are inhibited (MIC50) and the MIC90 for all the baseline C. albicans isolates were 0.25 and 1.0 µg/ml, respectively. Given the pharmacokinetics of fluconazole, which achieves concentrations in vaginal tissues and secretions considerably lower than those achieved in serum (2), we also selected
1 µg/ml and not
8 µg/ml as another potentially clinically relevant susceptibility breakpoint. Accordingly, MICs were
1 µg/ml for 365 of 393 isolates (92.8%) and in vitro MICs were >1 µg/ml for 28 isolates (7.2%); the latter concentration exceeds the predicted peak fluconazole concentration in the vagina.
|
View this table: [in a new window] |
TABLE 1. Comparative susceptibilities of C. albicans vaginal isolates at baseline and follow-up visits
|
64 µg/ml) and 1 DD-S isolate. The remaining 68 C. albicans isolates were susceptible (MICs,
8 µg/ml). The MIC50 and MIC90 for the 73 isolates of C. albicans obtained at the day 14 follow-up were unchanged at 0.25 and 1.0 µg/ml, respectively.
By using the alternative susceptibility breakpoint of
1 µg/ml, 62 isolates (84.9%) were determined to be fluconazole susceptible. Thus, the vast majority of early fluconazole clinical and mycological failures occurred in patients infected with susceptible strains of C. albicans (93.1%); in fact, most failures (84.9%) occurred with extremely susceptible strains for which the fluconazole MICs were
1 µg/ml.
(iii) Fluconazole susceptibility at the day 35 follow-up.
The vaginal isolates obtained on the first follow-up (day 14) visit reflected the isolates from all patients with clinical and mycological failures. Patients who failed clinically at day 14 were dropped from the study, and hence, no cultures were later performed for these patients. However, isolates of C. albicans were obtained from 101 patients on day 35. Some of the patients had become symptomatic once more, but the majority were asymptomatic. Among the 101 isolates, 3 were fluconazole resistant (MICs,
64 µg/ml). No DD-S strains were identified. Among the remaining 98 isolates, fluconazole MICs were
1 µg/ml for 96 of them, indicating once more that clinical and mycological relapses occurred in patients infected with strains highly susceptible to fluconazole. The MIC50 and MIC90 for the 101 C. albicans isolates at the day 35 visit were 0.25 and 0.5 µg/ml, respectively.
Correlation of baseline fluconazole susceptibility of C. albicans with clinical and mycological outcomes.
By using the baseline in vitro fluconazole susceptibility data for 280 C. albicans isolates obtained from patients evaluable for efficacy, pretreatment MICs were correlated with the clinical and mycological outcomes following fluconazole therapy (Table 2). The baseline MIC was >1 µg/ml for only 30 C. albicans isolates, and these isolates were compared with 250 isolates of C. albicans for which the MIC was
1 µg/ml. Rates of clinical success at day 14 were 93.2% (233 of 250 isolates) and 83.3% (25 of 30 isolates) for the groups of patients infected with isolates for which the MICs were low and high, respectively (no significant difference). At day 35, clinical success rates in the groups of patients infected with isolates for which the MICs were low and high were 76.8% (192 of 250 isolates) and 73.3% (22 of 30 isolates), respectively (no significant difference). By using the logistic regression model and controlling for the severity of vaginitis and the treatment regimen (one versus two doses of fluconazole), the odds of having clinical success in patients for whose isolates the baseline MICs were
1 µg/ml were >2.7 times that in patients for whose isolates the baseline MICs were >1 µg/ml at day 14 (95% confidence interval [CI], 0.91, 8.30). By day 35, the odds of having clinical success in patients for whose isolates the baseline MICs were
1 µg/ml were 1.2 times those in patients for whose isolates the baseline MICs were >1 µg/ml (95% CI, 0.49, 2.78). These odds ratios were not statistically significant.
|
View this table: [in a new window] |
TABLE 2. Correlation of in-vitro susceptibility and treatment outcome
|
1 µg/ml were twice those in patients for whose isolates the baseline MICs were >1 µg/ml at day 14 (95% CI, 0.87, 4.58) and 1.6 times at day 35 (95% C1, 0.72, 3.34), but statistical significance was not achieved.
Finally, despite the small numbers of patients, we compared the clinical and mycological responses for patients infected with susceptible C. albicans isolates (MICs,
1 µg/ml; n = 250) with the responses observed in the 14 patients for whose isolates the MICs were
64 µg/ml. Among the 12 patients available for follow-up, all 12 patients infected with resistant C. albicans strains improved or responded clinically by the time of the first follow-up visit and only 2 patients failed clinically by day 35. These patients fared less well mycologically, with persistent culture positivity found for 5 of 12 (41%) patients at day 14, which was double the rate for patients infected with susceptible strains (19.6%) but only slightly higher than the rate at day 35: 5 of 11 (45%) versus 97 of 250 (38.8%), respectively (Table 2).
Among patients infected with susceptible C. albicans isolates at the baseline, longitudinal comparison of in vitro susceptibilities revealed the development of fluconazole resistance in an isolate from only one patient. The baseline fluconazole MIC for the isolate from this patient was 0.125 µg/ml, and the MIC for the isolate recovered at the first follow-up visit was >64 µg/ml. Nevertheless, clinical improvement continued, and at the final visit, the patient was clinically cured and culture negative.
MICs for C. glabrata.
A total of 44 isolates of C. glabrata were isolated at any time during the study, and 2 (4.5%) were resistant to fluconazole (
64 µg/ml). No isolates for which MICs were 16 to 32 µg/ml were encountered. Thus, for 42 of 44 (95%) isolates MICs were 8 µg/ml or less and, hence, could be considered fluconazole susceptible. However, by using the susceptibility breakpoint of
1 µg/ml, only four isolates (9.1%) were susceptible to fluconazole, with the fluconazole MIC being 2 to 8 µg/ml for 86% of the C. glabrata isolates, concentrations not achievable in the vagina. The MIC50 and MIC90 for all C. glabrata isolates were 4 and 8 µg/ml, respectively.
Due to the paucity of C. glabrata isolates (n = 44), detailed comparison of longitudinally obtained isolates was not possible. The mean and median MICs remained unchanged. No fluconazole-resistant (MIC,
64 µg/ml) C. glabrata isolates were detected at the baseline. Two isolates of C. glabrata resistant to fluconazole (MICs,
64 µg/ml) emerged following exposure to fluconazole.
|
|
|---|
Recently, in vitro azole susceptibility data describing vaginal isolates obtained from a large cohort of HIV-positive women and risk behavior-matched HIV-negative women were published (18). However, the vaginal isolates were obtained from asymptomatic women colonized with Candida and not symptomatic women with Candida vaginitis. In that study of 647 vaginal isolates of C. albicans obtained over 2 years, only 2 isolates (<1%) were resistant to fluconazole and a similar number showed dose-dependent susceptibility to fluconazole (18) Similarly, a large multicenter prospective study of Candida vaginitis in women with AIDS failed to identify fluconazole-resistant vaginitis (13). Accordingly, the finding in the present study of 14 baseline isolates of C. albicans with fluconazole resistance (3.6%) should be viewed with concern and emphasizes the need for continued surveillance.
The short follow-up and duration of therapy in the present study precluded the drawing of any major conclusion regarding the acquisition of fluconazole resistance. Nevertheless, no change in the MIC50s or the MIC90s for the C. albicans vaginal isolates occurred when baseline and follow-up isolates were compared. C. glabrata strains from two patients developed resistance to fluconazole. A progressive reduction in fluconazole susceptibility in C. glabrata vaginal isolates following exposure to fluconazole has been reported previously (18). In the same longitudinal study, no corresponding increase in fluconazole MICs was seen for C. albicans isolates (18).
Clinically relevant susceptibility breakpoints have been suggested by Rex et al. (9) on the basis of a correlation of the clinical responses of patients infected with both blood and oral isolates of C. albicans and in vitro susceptibility data. We also analyzed the clinical responses of patients and the mycological eradication rates using a breakpoint which took into consideration the vaginal pharmacokinetic characteristics of fluconazole, although that breakpoint has not been validated as being clinically relevant by studies with humans or animals. Houang et al. (2) measured the fluconazole levels in vaginal secretions of patients following administration of a single 150-mg dose of fluconazole. The peak concentrations of fluconazole in vaginal secretions rarely exceeded 2.0 µg/ml; hence, we correlated the response to fluconazole therapy using a fluconazole MIC of
1 µg/ml as the susceptibility breakpoint. By use of these more stringent criteria, 93% of the baseline C. albicans isolates were still highly susceptible and would theoretically be inhibited by the predicted fluconazole concentrations. In fact, >90% of these patients responded extremely well clinically. Surprisingly, patients infected with microorganisms for which the MICs were higher (>1 µg/ml), including those infected with strains for which the MICs were
64 µg/ml, appeared to respond clinically, as did those patients infected with highly susceptible strains. However, a trend toward reduced mycological eradication rates was observed in patients infected with less susceptible strains. Hence, patients infected with C. albicans strains demonstrating reduced in vitro susceptibility to fluconazole could be expected to improve clinically and respond to fluconazole treatment but would be more likely to remain colonized and would be colonized with a population of organisms sufficiently large to be detected by routine culture of vaginal specimens. These individuals would therefore remain at risk for a subsequent recurrence of symptomatic disease when conditions associated with relapse develop. Nevertheless, the reason why patients infected with resistant and relatively resistant C. albicans strains did so well clinically is unclear. Perhaps a reduction in the vaginal population below a critical threshold is all that is needed to facilitate clinical improvement.
The most striking observation is that the overwhelming majority of women who failed clinically at both follow-up visits did so as a result of infection with C. albicans strains highly susceptible to fluconazole. Similarly, most persistently culture-positive women were infected with fluconazole-susceptible strains, emphasizing the limitation of azoles as fungistatic agents in the management of Candida vaginitis. Both systemic and topical azole agents appear to be similarly limited in their abilities to eradicate Candida from the vagina, despite impressive clinical success rates (17). The roles of the vaginal environment and local defense mechanisms in contributing to drug-associated vaginal yeast eradication or persistence have not been adequately defined and may be underestimated. Poorly controlled diabetic patients represent an example of individuals with a vaginal microenvironment that facilitates Candida persistence after azole therapy. Similarly, C. glabrata and other but not all non-C. albicans Candida species for which MICs are higher are more likely to persist (16, 19).
Several studies have demonstrated the clinical usefulness of in vitro susceptibility testing of yeasts by use of the NCCLS reference method (6, 10, 12, 20). In principle, standards applied to oropharyngeal candidiasis are such that one should anticipate a favorable response in 60 to 100% (usually 90%) of patients treated with antifungals to which the organism is susceptible in vitro, whereas one would anticipate a favorable response in 0 to 60% of patients infected with organisms exhibiting in vitro resistance (the 90/60 rule). Although the present study was biased by the inclusion of only patients with complicated Candida vaginitis and not those with mild and uncomplicated diseases, the results can still be used to validate the usefulness of in vitro susceptibility in predicting the outcome of therapy. The logistic regressive model that adjusted for several patient variables revealed that women infected with C. albicans isolates highly susceptible to fluconazole (MICs,
1 µg/ml) were 2.7 times more likely to respond clinically than women infected with isolates for which the MICs were above the concentration of fluconazole achievable in vaginal tissue. Moreover, the advantage of low MICs was even more apparent with regard to the mycological eradication of vaginal C. albicans strains. Finally, this principle is supported by the poor clinical and mycological outcomes seen for patients infected with C. glabrata, for which, predictably, fluconazole MICs are above the concentrations of fluconazole achievable in vaginal tissue (16, 19).
On the basis of the results of this study, we do not recommend in vitro susceptibility testing for all patients with severe or recurrent Candida vaginitis. It is reasonable to perform susceptibility tests for women who have poor clinical and mycological responses in individual episodes or for those receiving maintenance suppressive azole regimens who have breakthrough infections. Studies to validate these recommendations are in progress.
In summary, complicated Candida vaginitis caused by fluconazole-resistant C. albicans remains uncommon. Why women infected with highly susceptible strains of C. albicans occasionally fail to respond clinically and why many more improve clinically but remain colonized with C. albicans following fluconazole therapy remain unresolved but may be due to the fungistatic nature of azole agents. Organisms for which MICs are higher showed a trend toward vaginal persistence. Routine susceptibility testing of vaginal C. albicans isolates from women with complicated (severe or recurrent) vaginitis is not justified.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»