Fluconazole step-down therapy versus echinocandins for the treatment of Candida glabrata invasive candidiasis with candidaemia

Abstract

Objectives

Candida glabrata is the second most common species responsible for invasive candidiasis, including candidaemia. Echinocandins are typically the first-line therapy for C. glabrata candidaemia, with the option to transition to oral fluconazole. Studies are needed to evaluate clinical outcomes in patients initially treated with echinocandins then transitioned to fluconazole.

Methods

This was a retrospective, single-centre cohort study of patients with C. glabrata candidaemia from November 2011 to August 2023. Inpatients aged 18–89 years who received an echinocandin within 24 h of the initial positive blood culture were included. Patients were excluded if they received antifungal treatment less than 48 h, combination therapy, or fluconazole as initial therapy. The primary composite outcome was 30-day clinical failure.

Results

A total of 186 patients were included (n = 153 echinocandin only; n = 33 fluconazole step-down). The most common source of candidaemia was line-associated in both groups with the majority having source control (43% echinocandin versus 58% fluconazole; P = 0.32). Compared to fluconazole, patients in the echinocandin group had a higher rate of concomitant bacteraemia (45% versus 24%; P = 0.03) and endovascular complications (11% versus 0%; P = 0.05). There was no significant difference in treatment duration between echinocandin and fluconazole (16 versus 19 days; P = 0.46), incidence of persistent candidaemia (22% versus 24%; P = 0.7), or 30-day clinical failure (15% versus 9%; P = 0.58).

Conclusions

Fluconazole appears to be a safe and reasonable step-down therapy in the management of C. glabrata candidaemia.

Introduction

Candidaemia stands as the second most common healthcare-associated bloodstream infection, with an estimated 25 000 cases annually in the USA.^1,2 Among these, Candida glabrata has emerged as the second predominant cause of candidaemia.³

Invasive candidiasis encompasses both candidaemia and deep-seated tissue candidiasis.^4,5 Echinocandins are typically the first-line treatment for C. glabrata candidaemia, owing to their tolerability, minimal drug-drug interactions and potent fungicidal properties.⁴ Nonetheless, the option to transition to oral fluconazole when clinically appropriate is appealing because of its tolerability, significant pharmacokinetic advantages over echinocandins and convenient oral formulation.^4,6 Although the susceptible-dose-dependent (SDD) minimum inhibitory concentrations (MICs) established by the Clinical and Laboratory Standards Institute necessitate higher fluconazole dosages, its high oral bioavailability highlights its pharmacokinetic advantages.^4,6

Prior studies comparing clinical outcomes among patients with C. glabrata candidaemia receiving echinocandin versus azole therapy focused on uncomplicated candidaemia, featured small sample sizes and yielded mixed results.^7–16 This study compared clinical outcomes between echinocandins with fluconazole step-down therapy and echinocandins alone for C. glabrata invasive candidiasis with candidaemia.

Materials and methods

This was a retrospective review conducted at a multi-site academic medical centre of patients with a positive C. glabrata blood culture from November 1, 2011 to August 31, 2023. Data were collected via the electronic medical record Integrated Health Information System (IHIS). Eligible patients were adults aged 18–89 years, with C. glabrata growth in one or more blood cultures, who received an echinocandin within 24 h of a positive blood culture Gram stain or blood culture identification panel (BCID).

Patients were excluded for any of the following: prisoner status, receipt of antifungal therapy for less than 48 h, definitive fluconazole therapy for C. glabrata strains with a fluconazole MIC > 32 mg/L,⁶ fluconazole as initial therapy, and definitive combination therapy with fluconazole and an echinocandin.

Patients meeting inclusion criteria were categorized into two treatment arms according to their definitive therapies: (i) those who received echinocandin therapy for at least 70% of their total treatment duration, and (ii) those who were initiated on echinocandin therapy that was subsequently transitioned to fluconazole for more than 30% of their total treatment duration.

The primary outcome was a composite endpoint of 30-day clinical failure, encompassing C. glabrata candidaemia recurrence, escalation in therapy, infection-related readmission or infection-related mortality. Candidaemia recurrence was defined as a positive blood culture for C. glabrata after more than 72 h of previously demonstrated clearance. Infection-related readmission and mortality were determined by an Infectious Diseases physician. Escalation in therapy followed Infectious Diseases Society of America guidelines, involving transitions from fluconazole to an echinocandin or amphotericin B formulation, and from echinocandin to an amphotericin B formulation.⁴

Secondary outcomes included adverse drug events [such as hepatic toxicity (defined as liver function tests > five times the upper limit of normal), infusion-related toxicity or reaction (such as rigours and hypotension.)], length of stay (LOS) (hospital and intensive care unit or ICU) (days) and persistent candidaemia (defined as any positive blood culture with C. glabrata collected ≥5 days from the index blood culture).¹⁷ Duration of candidaemia was evaluated and defined as the number of days during which blood cultures remained positive for C. glabrata. The Charlson comorbidity index (CCI) was utilized to compare baseline demographics, given its association with mortality rates in patients with candidaemia.¹⁵

Definitive therapy was defined as fluconazole or echinocandin therapy selected upon availability of susceptibility results. Descriptive statistics were calculated for baseline demographics, clinical data and outcomes, with continuous variables assessed via the Student’s t-test or Wilcoxon rank sum test, and categorical variables analysed using the Chi-square or Fisher’s exact test where appropriate. A power calculation was performed estimating a 15% difference in clinical failure between fluconazole and echinocandin therapies based on previous literature and an expected mortality rate of 25%.² Establishing power at 80% and a clinically significant effect size of 15%, 200 patients were required (100 per group).

All statistical tests were performed using SAS Version 9.4 (SAS Institute, Inc., Cary, NC) with significance defined as P < 0.05.

Ethics

The study received approval from The Ohio State University institutional review board (study number 2023H0341). Given its retrospective design, written informed consent was not required.

Results

Clinical characteristics

Out of the 212 patients initially screened, 186 met inclusion criteria. The most common reason for exclusion was antifungal initiation greater than 24 h after the index blood culture (46%). Baseline characteristics were similar between the two groups (Table 1). All isolates in the echinocandin group demonstrated susceptibility to either caspofungin or micafungin, both of which were reported due to an institutional conversion from caspofungin to micafungin during the study period. All patients in the fluconazole group exhibited an SDD MIC to fluconazole. Blood culture clearance was achieved in 142 (93%) and 33 (100%) patients in the echinocandin and fluconazole groups, respectively (P = 0.22), with most patients receiving echinocandins at the time of clearance (82% echinocandin; 85% fluconazole; P = 0.19). Four patients in the echinocandin group received fluconazole during their treatment course, however, they each received more than 70% of their treatment duration with an echinocandin. The median duration to clearance of cultures from the initiation of active therapy was 1 day for both groups (P = 0.13). Overall, the median duration of candidaemia was 3 days [IQR 2–4] for both groups.

Table 1.

Baseline characteristics of echinocandin- versus fluconazole-treated patients with C. glabrata invasive candidiasis with candidaemia

Characteristic	ECH (n = 153)	FLU (n = 33)	P value
Age, years	59 [49–68]	58 [42–65]	0.33
Male	81 (53)	18 (55)	0.87
Actual body weight, kg	73.7 [62.1–89.4]	83.1 [61.8–92.2]	0.53
Charlson comorbidity index	5 [3–8]	4 [2–9]	0.71
Infectious diseases consultation	146 (95%)	29 (88%)	0.11
Abdominal candidaemia source			0.25
Intra-abdominal abscess	11 (7)	4 (12)
Gastrointestinal translocation	33 (22)	6 (18)
Other candidaemia source			0.19
Line-associated	51 (33)	13 (39)
Skin	3 (2)	3 (9)
Genitourinary	6 (4)	0
Surgical	1 (1)	0
Unknown	48 (31)	7 (21)
Source control, yes	66 (43)	19 (58)	0.32
Complications, yes	23 (15)	2 (6)	0.26
Infective endocarditis or cardiac device involvement	17 (11)	0	0.05
Endophthalmitis	0	2 (6)	0.03
Other	6 (4)	0	0.59
Concomitant bacteraemia	69 (45)	8 (24)	0.03
Concomitant non-glabrata candidaemia	19 (12)	3 (9)	0.77
Fluconazole MIC distribution, μg/mL			0.0003
≤0.5	0	1 (3)
1	1 (1)	3 (9)
2	9 (6)	6 (18)
4	39 (25)	12 (36)
8	51 (33)	9 (27)
16	23 (15)	2 (6)
32	10 (7)	0
>32	20 (13)	0
Fluconazole dose/MIC ratio			0.33
50	—	8 (24)
≥100	—	25 (76)
Fluconazole mg/kg dosing using actual body weight	—	6 [5–8]
Total treatment duration, days	16 [14–25]	19 [15–29]	0.46
Duration of echinocandin prior to fluconazole, days	18 [12.5–28.5] (n = 4)	7 [5–9]	0.002
Duration of fluconazole post-echinocandin, days	3 [2.5–4.5] (n = 4)	13 [9–19]	<0.0001

Data are presented as number (%) or median [IQR] as appropriate.

ECH, echinocandin; FLU, fluconazole.

Outcomes

No significant difference was observed between the echinocandin (15%) and fluconazole groups (9%) with regard to 30-day clinical failure (P = 0.58) (Table 2). No differences were observed in secondary outcomes (Table 2).

Table 2.

Treatment outcomes with echinocandin and fluconazole for C. glabrata invasive candidiasis with candidaemia

Outcome	ECH (n = 153)	FLU (n = 33)	P value
30-Day clinical failure	23 (15)	3 (9)	0.58
Recurrence	9 (6)	0	0.37
Infection-related readmission	2 (1)	0	1
Infection-related mortality	11 (7)	1 (3)	0.7
Escalation in therapy required	6 (4)	2 (6)	0.63
Persistent candidaemia	33 (22)	8 (24)	0.7
Treatment-related adverse events	0	0	—
Hospital LOS, days	26 [14–43]	28 [11–38]	0.29
ICU LOS, days	16 [8–29]	11 [7–25]	0.39

Data are presented as number (%) or median [IQR] as appropriate.

ECH, echinocandin; FLU, fluconazole; LOS, length of stay.

Discussion

These findings support fluconazole step-down therapy in managing C. glabrata invasive candidiasis with candidaemia. Baseline demographics, source of infection and achievement of source control were similar across both treatment cohorts. However, increased fluconazole use in endophthalmitis patients and higher echinocandin use in those with infective endocarditis reflected azoles’ optimal eye penetration and echinocandins’ potent biofilm activity.⁴ Interestingly, the echinocandin group exhibited higher concomitant bacteraemia rates, suggesting a potentially more complex patient profile despite similar ICU admission and LOS between the groups. Alternatively, this may indicate a preference for intravenous echinocandins in patients with bacteraemia already receiving intravenous antimicrobials. However, despite no differences in hospital or ICU LOS, oral fluconazole remains a desirable option to minimize intravenous therapy and related complications, such as secondary line infections.

Previous literature indicates a preference for echinocandins as initial therapy in invasive candidiasis after adjusting for confounders.^8,10 However, studies comparing definitive antifungals for invasive candidiasis, including candidaemia, yielded inconsistent findings, with few investigating oral azoles for step-down treatment.^7,9,11–15 A Phase IV study evaluated patients receiving intravenous anidulafungin and patients transitioned to oral azoles for uncomplicated invasive candidiasis treatment (majority candidaemia). Similar response rates were observed between the modified intent-to-treat population and those switched to oral azoles by day 7, however, only 20 C. glabrata isolates were in the early step-down group. Unlike the previous study, which used oral voriconazole for C. glabrata and excluded complicated cases, the present study used oral fluconazole and included complications such as endocarditis and endophthalmitis.¹⁶ Furthermore, this study sheds light on the significance of fluconazole dosing in Candida infections, an aspect less explored previously. While there is not a guideline-recommended goal PK/PD fluconazole dosing target, existing literature suggests a fluconazole dose to MIC (dose/MIC) ratio of ≥50 for candidaemia, while EUCAST recommends a dose/MIC ratio of ≥100.^18,19 Additionally, it is unclear whether a target dose/MIC ratio should differ for initial therapy versus step-down therapy, especially if source control is achieved. In this study, 76% of fluconazole-treated patients received a fluconazole dose/MIC ratio of ≥100 (Table 1), potentially contributing to lower clinical failure rates than earlier studies that reported clinical failure rates ranging from 28% to 50% and mortality rates from 3% to 52%.^8–11,14,15 Other contributing factors likely include high infectious diseases consultation rates,²⁰ rigorous blood culture monitoring and proactive intervention protocols, rather than ICU admission or illness severity.

This study had limitations as the retrospective design inherently limited data collection. Patients without follow-up blood cultures at the specified time frames were presumed to be negative unless additional information suggested otherwise, possibly underestimating true recurrence and persistence rates. While ICU admission rates and CCI scores were similar between the two groups, other factors may have influenced outcomes. None of the considered variables met criteria for inclusion as confounding factors in a multivariable logistic regression model, possibly due to a limited sample size that limited statistical power and may have failed to detect differences in clinical failure rates. Caution is warranted in interpreting these findings, particularly for complicated candidiasis. Future studies should consider clinical severity and fluconazole dosing as potential confounders influencing candidaemia outcomes.

In conclusion, this study found no difference in outcomes between echinocandins followed by fluconazole step-down and echinocandins monotherapy for managing C. glabrata invasive candidiasis with candidaemia. Fluconazole step-down therapy appears safe and effective when appropriately dosed, particularly with documented candidaemia clearance and ideally with source control. Future studies are warranted to evaluate its utility in complicated C. glabrata invasive candidiasis, as fluconazole may not be the optimal therapy option in certain clinical situations.

Funding

The project described was supported by award number ULT1TR002733 from the National Center for Advancing Translational Sciences. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Advancing Translational Sciences or the National Institutes of Health.

Transparency declarations

All authors have no conflicts of interest to disclose.