Candidemia is a severe and life-threatening infection caused by different Candida species. In Germany, candidemia ranks as the 6th most common bloodstream infection after Escherichia coli, Staphylococcus aureus, coagulase-negative staphylococci, Klebsiella pneumoniae, and enterococci (Schöneweck et al., 2021). From the bloodstream, Candida can disseminate to multiple organs involving most commonly the liver, spleen, kidney, myocardium, or eyes and less frequently the brain (Pilmis et al., 2017). Invasive candidiasis is associated with a high mortality rate of about 40%. While many different Candida species can be responsible for invasive diseases, 95% of the infections are caused by the five species Candida albicans, Candida glabrata, Candida krusei, Candida parapsilosis, and Candida tropicalis (Yapar, 2014). Although C. albicans is still responsible for the majority of infections in Germany (Mohr et al., 2020; Schroeder et al., 2020), non-albicans Candida species can represent up to 60% of the cases of candidemia in other parts of the world like France (Vannini et al., 2022), Italy (Montagna et al., 2013), or the United States (Pfaller et al., 2012). Compiled data from the ARTEMIS DISK registry including data from 147 medical centers in 41 countries all over the world for the years 1997–2007 indicated that despite a decrease in frequency, C. albicans remained the most frequent species worldwide, while the frequency of C. glabrata and C. krusei was stable, and the frequency of C. parapsilosis and C. tropicalis increased (Pfaller et al., 2010). Echinocandins are recommended as first-line therapies for invasive candidiasis, and azoles can be used, as step-down therapies (Cornely et al., 2012), but decreased susceptibility and even resistance to antifungals can occur. Decreased susceptibility or drug resistance in Candida species can be intrinsic, like decreased caspofungin susceptibility in C. parapsilosis (Pfaller et al., 2006) or fluconazole resistance in C. krusei (Orozco et al., 1998), or can be acquired like echinocandin resistance in C. glabrata (Aldejohann et al., 2021). The high mortality rate, the shift towards more difficult to treat Candida species, and the lack of efficacy in monotherapy for some difficult-to-treat infections make alternative approaches necessary (Vitale, 2021). It has been shown in vitro (Schwarz et al., 2007), in vivo (Schwarz et al., 2006b), and in patients (Day et al., 2013) that a combination of two antifungals can increase their potency and decrease resistance. However, against invasive candidiasis, no favorable combination has been found (Johnson et al., 2004). Apart from the combination of two antifungals, combinations of antifungal and other molecules can also lead to favorable interactions (Schwarz et al., 2019). The antibiotic colistin has shown in vitro synergy in combination with amphotericin B against Candida species (Schwarz et al., 2022), but routine use in critically ill patients is limited due to the nephrotoxicity of the molecules. Isavuconazole is a broad-spectrum azole with favorable tolerability (Ledoux et al., 2018), which makes it an interesting partner for combinations of molecules. In vitro combination of isavuconazole with colistin has been shown to be synergistic against Aspergillus nidulans, Aspergillus niger (Schwarz et al., 2020b), and Candida auris (Schwarz et al., 2020a) and could thus be also an interesting combination against Candida species. We therefore explored the combination of isavuconazole with colistin against common Candida species by a checkerboard technique based on the European Committee on Antimicrobial Susceptibility Testing (EUCAST) methodology for antifungal susceptibility testing.

The interactions of isavuconazole with colistin were evaluated by checkerboard and interpretation of the results by FICI or by response surface analysis against strains from six common Candida species as presented in Table 1 . A summary of the results is presented in Table 2 . Figure 1 shows the synergy distributions for the combination of isavuconazole with colistin against a representative isolate of each tested species.

The 57 Candida strains exhibited MICs for isavuconazole ranging from 0.0001 to 0.5 μg/ml ( Table 1 ) with MIC50, MIC90, and geometric mean MIC of 0.008, 0.125, and 0.018 μg/ml, respectively. Isavuconazole MICs ranged from 0.0005 to 0.004, 0.125 to 0.5, 0.03 to 0.25, 0.008 to 0.03, 0.008 to 0.06, and 0.0001 to 0.008 µg/ml for C. albicans, C. glabrata, C. krusei, C. parapsilosis, C. tropicalis, and C. kefyr, respectively. When tested alone, colistin exhibited MICs ranging from 16 to 128 µg/ml for the different species (128 μg/ml being the high-off scale MIC) with MIC50, MIC90, and geometric mean MIC of 64, 128, and 63.23 μg/ml, respectively. The best activity of colistin was seen against C. tropicalis with MICs ranging from 16 to 32 µg/ml. Significant activity was also seen for C. kefyr with a geometric mean MIC of 32 µg/ml. Against C. krusei, colistin MICs were high with a geometric mean MIC of 59.71 µg/ml. Colistin showed almost no activity against C. albicans, C. glabrata, and C. parapsilosis with geometric mean MICs of 119.43, 119.43, and 109.73 µg/ml, respectively. The geometric mean MIC for colistin in combination with the synergistic strains was 9.93 µg/ml. Between experiments, isavuconazole and colistin MICs were within ±1 log₂ dilutions in 99.12% of the cases for all Candida species tested (data not shown). The interaction was synergistic for 58% of the strains with FICIs ranging from 0.25 to 0.5 with a geometric mean FICI of 0.4. Synergy was obtained for 0%, 33%, 50%, 80%, 90%, and 90% of C. albicans, C. parapsilosis, C. kefyr, C. krusei, C. glabrata, and C. tropicalis, respectively ( Table 2 ). All other interactions were indifferent. The geometric mean FICI for all strains was 0.48.

Analysis of the checkerboard data of the 57 Candida strains by the response surface approach led to similar results compared to the FICI results. Although synergy was less frequently obtained than by FICI analysis, the tendency if the combination is synergistic, or for a certain species indifferent (at least 50% of the strains for both interpretation models), was correct in 5/6 cases ( Table 2 ). The SUM-SYN-ANT metric for the synergistic strains ranged from 59.37 to 122.67, with a mean of 78.22. For C. glabrata, the mean of the SUM-SYN-ANT metric of all strains was 40.42. Synergy was obtained for 0%, 0%, 10%, 50%, 50%, and 67% of C. albicans, C. parapsilosis, C. glabrata, C. krusei, C. tropicalis, and C. kefyr, respectively. When comparing the results of the FICI with the response surface approach, synergy was obtained for at least 50% of the strains for both interpretation models for C. kefyr, C. krusei, and C. tropicalis. One difference between the interpretation techniques was that interaction against C. glabrata was synergistic (9 of 10 strains) by FICI and indifferent by response surface analysis (9 of 10 strains).

Advantages of drug repurposing are the use of de‐risked compounds, potentially lower development costs, and shorter development timelines (Pushpakom et al., 2019). Repurposed drugs against Candida belong to the antihelmintics (quinacrine), cytotoxic agents (doxorubicin and daunorubicin), or 5α-reductase inhibitors (finasteride) and inhibit filamentation (Miro-Canturri et al., 2019). Colistin is an antibiotic with activity against gram-negative bacteria (Lim et al., 2010), but it has been shown to be able to damage the membrane in C. albicans (Yousfi et al., 2019) and increase its permeability (Bibi et al., 2021), which makes the molecule an interesting partner to test combinations with antifungals.

Compared to other studies, the MICs of isavuconazole for the different Candida species in this study were in the same range as previously reported (Desnos-Ollivier et al., 2019; Jorgensen et al., 2019). MICs for colistin were comparable to those of another study, but apart from C. albicans, only one strain of each species has been tested (Zeidler et al., 2013). Compared to our recently published study on the combination of amphotericin B and colistin, colistin MICs were similar (Schwarz et al., 2022).

Combination MICs of colistin ranged from 1 to 32 μg/ml, with a geometric mean MIC for the synergistic strains of 9.93 μg/ml. Peak serum levels of 13 to 32 have been reported in patients with cystic fibrosis (Reed et al., 2001), which would be sufficient against the strains tested in this study, but colistin use in patients is limited due to its nephrotoxicity (Javan et al., 2015). However, in patients with cryptococcosis, it has been shown that synergy can be achieved with lower serum levels than those tested in vitro (Kontoyiannis et al., 2008).

Against common Candida species, in vitro synergy of combinations including colistin has been reported for echinocandins (Zeidler et al., 2013) and for amphotericin B (Schwarz et al., 2022). Combinations of colistin with azoles were evaluated by two studies. One study found indifference for the combinations of colistin with either fluconazole or itraconazole against C. albicans, but both combinations were only tested against one strain (Yousfi et al., 2019). Our study is in accordance with these results. We found indifference for the combination of isavuconazole with colistin by checkerboard and interpretation of the results by FICI and response surface analysis. Another study also evaluated the combination of colistin with fluconazole by checkerboard against one C. albicans strain and found synergy in vitro and in vivo in a Galleria mellonella model of invasive candidiasis (Bibi et al., 2021). This discrepancy compared to our results might be strain specific, as only one strain was tested or might be related to the different azole used.

For the other species tested in this study, we found synergy for the majority, or at least for half of the tested strains, for C. glabrata, C. krusei, C. tropicalis, and C. kefyr. Only C. parapsilosis indifference was found for the majority of the strains. Although synergy was less frequently seen, when the checkboard data were evaluated by response surface analysis, at least 50% of the strains exhibited synergy regardless of the interpretation model used for C. kefyr, C. krusei, and C. tropicalis. For C. parapsilosis, both interpretation models evaluated indifference. Between the two interpretation techniques, one difference was found for C. glabrata. Interpretation by the FICI showed synergy for 90% of the strains, while interpretation by response surface analysis exhibited indifference for 90% of the strains. Nevertheless, despite the formal indifference of the C. glabrata results by response surface analysis, the mean of the SUM-SYN-ANT metric of all strains was quite high (40.42). The discrepancy could indeed be related to the stringent threshold (56.0) used in the present work compared to previous studies (Schwarz et al., 2020a; Schwarz et al., 2022).

In conclusion, we found in vitro synergy of the combination of isavuconazole with colistin against C. glabrata, C. krusei, C. tropicalis, and C. kefyr. Against C. albicans and C. parapsilosis, the combination only exhibited indifference. Antagonism was never seen regardless of interpretation model used. These results warrant further animal experiments.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://www.ncbi.nlm.nih.gov/genbank/, OM859334 to OM859357 https://www.ncbi.nlm.nih.gov/genbank/, OL351325 to OL351356.

PS wrote the first draft of the manuscript. PS and IN carried out the experiments. PS, IN, A-LB, and ED performed the analysis of the results. ED, PS, GB, and FS contributed to the revisions. All authors have read and agreed to publish the final version of the manuscript.

Open Access funding provided by the Open Acess Publication Fund of Philipps-Universität Marburg with support of the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation). This study was supported by internal funding.

PS has received research grants from Basilea Pharmaceutica, Gilead, and Pfizer; travel grants from Gilead and Pfizer; and speaking fees from Pfizer. During the past 5 years, ED has received research grants from MSD and Gilead; travel grants from Gilead, MSD, Pfizer, and Astellas; and speaker’s fees from Gilead, MSD, and Astellas.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.