Predictors for Prolonged Hospital Stay Solely to Complete Intravenous Antifungal Treatment in Patients with Candidemia: Results from the ECMM Candida III Multinational European Observational Cohort Study

Background To date, azoles represent the only viable option for oral treatment of invasive Candida infections, while rates of azole resistance among non-albicans Candida spp. continue to increase. The objective of this sub-analysis of the European multicenter observational cohort study Candida III was to describe demographical and clinical characteristics of the cohort requiring prolonged hospitalization solely to complete intravenous (iv) antifungal treatment (AF Tx). Methods Each participating hospital (number of eligible hospitals per country determined by population size) included the first ~ 10 blood culture proven adult candidemia cases occurring consecutively after July 1st, 2018, and treating physicians answered the question on whether hospital stay was prolonged only for completion of intravenous antifungal therapy. Descriptive analyses as well as binary logistic regression was used to assess for predictors of prolonged hospitalization solely to complete iv AF Tx. Findings Hospital stay was prolonged solely for the completion of iv AF Tx in 16% (100/621) of candidemia cases by a median of 16 days (IQR 8 – 28). In the multivariable model, initial echinocandin treatment was a positive predictor for prolonged hospitalization to complete iv AF Tx (aOR 2.87, 95% CI 1.55 – 5.32, p < 0.001), while (i) neutropenia, (ii) intensive care unit admission, (iii) catheter related candidemia, (iv) total parenteral nutrition, and (v) C. parapsilosis as causative pathogen were found to be negative predictors (aOR 0.22 – 0.45; p < 0.03). Interpretation Hospital stays were prolonged due to need of iv AF Tx in 16% of patients with candidemia. Those patients were more likely to receive echinocandins as initial treatment and were less severely ill and less likely infected with C. parapsilosis.


Introduction
Invasive candidiasis (IC) including candidemia remains the most frequent invasive fungal infection in European and worldwide hospitals [1,2].Mortality rates are high with around 37% expected to have a fatal outcome at day 30 [3,4].Early adequate antifungal treatment is efficacious and one of the most important predictors of survival [5].The European Society for Clinical Microbiology and Infectious Diseases (ESCMID) [6] and the Infectious Diseases Society of America (IDSA) guidelines [7] recommend echinocandins as first line treatment, which can be administered intravenously only.While stepdown to oral fluconazole treatment can be considered after 5 days or even earlier [8,9], fluconazole has a narrower spectrum of activity paired with the continuous emergence of azole resistance among nonalbicans Candida spp., including C. glabrata [10], is less well tolerated than echinocandins causing hepatotoxicity, and also lacks activity against biofilms [11].Therefore, stepdown to an oral agent is not always feasible and prolonged intravenous (iv) antifungal treatment (AF Tx) may be necessary.This may prevent otherwise timely discharge of patients, meaning prolonged hospitalization to complete iv AF Tx or daily outpatient iv treatment, if available.Results are increased costs, risk for nosocomial infections, and avoidable waste of resources.New antifungals, including rezafungin (echinocandin with prolonged half-life allowing once-a-week administration) and ibrexafungerp (oral triterpenoid glucan synthase inhibitor) are currently in late stage clinical development and may help tackle this problem [12,13].To utilize these new drugs most efficaciously, studies are needed to identify patient populations that would benefit most from these new antifungal treatments.However, demographic and clinical factors that are associated with prolonged hospitalization solely to complete iv AF Tx have to date not been sufficiently described.
In order to delineate characteristics, as well as identify predictors for prolonged hospitalization solely to complete AF Tx, we performed a sub-analysis of the ECMM Candida III dataset [4].

Study Design and Participating Centers
The ECMM designed and conducted the CANDIDA III study-its third pan European multicenter observational cohort study over the past 25 years [4,14,15]-to collect data on epidemiology, risk factors, treatment, and outcomes of culture proven candidemia across Europe, as well as to assess the impact of guideline adherence on survival [4].This represents a sub-analysis of CANDIDA III [4], where each participating hospital included the first * 10 blood culture proven adult candidemia cases occurring consecutively after July 1 st , 2018.In short, participating centres/treating clinicians entered data on patient demographics, risk factors and characteristics, duration of hospitalization, diagnostic procedures, causative Candida species, treatment characteristics including antifungal treatment, whether hospital stay was prolonged only for completion of parenteral antifungal treatment, and outcomes, into the ECMM Candida Registry-CandiReg -FungiScopeÒ (NCT 01731353) [16,17], on www.clinicalsurveys.net(EFS Fall 2018 Questback, Cologne, Germany) [4].Candidemia was defined as the isolation of Candida species from blood culture, and the subanalysis was performed within the pool of those 621 where Candida spp. was reported.The main objective of this subanalysis was to describe demographic and clinical characteristics of the cohort requiring prolonged hospitalization to complete iv AF Tx.

Statistical Analysis and Ethics
For patients with prolonged hospitalization solely to complete iv AF Tx descriptive statistical analysis was performed for clinical characteristics and demographic variables, as well as for the distribution of Candida species.Case level data is available from corresponding authors by request.Data were summarized employing frequencies, percentages, mean and standard deviation (SD) or median and interquartile range (IQR) as appropriate.Binary logistic regression was used to assess for predictors of prolonged hospitalization solely to complete iv AF Tx in 30-day candidemia survivors.The reason for focusing only on 30-day survivors was to avoid bias that could have been introduced by including patients who had a fatal outcome before they could even possibly meet the event definitions of prolonged hospitalization solely to complete iv AF Tx.First, univariable analysis of clinical predictors was performed.All variables significant at p B 0.15 in univariable analyses were considered as possible predictor variables for multivariable analysis.
This study was performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments.For the database, retrospective data entry, and data analysis a central ethical approval was obtained at the University of Cologne, Germany (EK 17-485) that indicates that, generally, neither informed consent nor institutional review board (IRB) approval of each participating hospital would be required.Each participating hospital was required to obtain local IRB confirmation or approval as deemed necessary by the respective local regulations/authorities.Two-sided p \ 0.05 was taken as cut-off for statistical significance.All statistical analyses were performed using IBM SPSS Statistics 25 (SPSS Inc., Chicago, IL, USA) and R (version 4.3.1;www.r-project.org).

Results
A total of 621 unique patients with culture proven candidemia were included in the parent study.Of those, 16.1% (100/621), reported from 33 institutions in 16 countries had prolonged hospitalization specifically for the completion of iv AF Tx according to the treating physician.Hospitalization was prolonged solely to complete iv AF Tx by a median of 16 days (IQR 8 -28; mean 22 days, SD 19 days; data available from 81 patients); the total duration of hospitalization in that group was median 21 days (IQR 15-45, mean 35 days).The median duration of hospitalization in those without prolonged hospitalization for iv AF Tx (n = 521) was 14 days (IQR 5-30 days; mean 22 days, SD 42).
In about half the cases (55%) echocardiography was performed, showing cardiac involvement in 13% of those examined.Eye exam was reported in 42% of cases showing ocular involvement in 12%.The vast majority (85%) received treatment consultation by an infectious diseases and/or microbiology expert.Allcause mortality at day 30 was 10%, and in 6/10 cases with a fatal outcome by day 30 investigators classified the death as unrelated to candidemia.
Echinocandins were the first line antifungal drug in 73% of cases.Another 5% were started on fluconazole but were then switched to an echinocandin, while 3% were either started on or switched to liposomal amphotericin B (Table 1).
In addition, 4.3% (27/621) of patients included in the parent study from 16 institutions (including three institutions from France and two from Belgium, Germany, Netherlands, Spain and United Kingdom each) received outpatient intravenous antifungal therapy (all did not have prolonged hospital stay for completion of iv AF Tx).Among those, C. albicans and C. glabrata were the most frequent pathogens (each 38%; 10/26), followed by C. tropicalis, C. auris and C. parapsilosis (each 8%; 2/26).30-day mortality was 11% (3/27) and overall mortality 22% (6/27).

Discussion
We performed a sub-analysis of a multicenter observational study of candidemia, involving 64 hospitals from 20 countries across Europe, and found that 16.1% of patients with candidemia had prolonged hospitalization solely to complete iv AF Tx.These were less severely ill, had higher survival rates [4] and were less likely infected with C. parapsilosis.Outpatient iv AF Tx was performed in an additional 4.3% of the study population, outlining that in 20.4% no (sufficient) oral treatment options were available.
Our study highlights that there was a significant group of patients where step-down to oral fluconazole [8] was not performed.In the multivariable model, indicators of severe disease, namely neutropenia, ICU admission and TPN, but also CRBSI and C. parapsilosis as a causative pathogen remained significant negative predictors for prolonged hospitalization solely to complete iv AF Tx, likely because some of these factors were per se associated with longer duration of hospitalization.In contrast, initial echinocandin treatment and candidemia caused by C. glabrata were positive predictors for prolonged hospitalization solely to complete iv AF Tx, with the latter not reaching statistical significance in the multivariate model.However, C. glabrata was even more overrepresented among the smaller group of 27 patients who received outpatient iv AF Tx, of whom 38% had candidemia caused by C. glabrata, underpinning the impact of this organism that often shows resistance to azoles.The fact that C. parapsilosis, a pathogen associated with lower mortality and often causing CRBSI [10], was a negative predictor for prolonged hospitalization solely to complete iv AF Tx, indicates that this pathogen is still often (deemed) susceptible to oral azoles in Europe.Of note, azole resistance is also increasing among C. parapsilosis with several outbreaks reported [18,19].Therefore, lack of reliable oral treatment options may also depict a future problem for infections caused by C. parapsilosis.
In our cohort 30-day all-cause mortality was 10% (11% in those with outpatient iv AF Tx) compared to 44% in the remaining Candida III cohort without prolonged hospitalization solely to complete iv AF Tx or outpatient iv AF Tx (including patients who did not survive long enough to receive prolonged therapy),   outlining the better prognosis and less severe disease course [4].Importantly, initial echinocandin therapy was not only a positive predictor for prolonged hospitalization solely to complete iv AF Tx, but was also associated with better outcomes and survival in the overall study cohort [4].More and better options for antifungal treatment outside the hospital will emerge in the near future, with a promising antifungal pipeline [20].Rezafungin, an echinocandin with improved penetration into the peritoneal fluid and prolonged half-life, allowing once weekly infusion, has shown noninferiority to caspofungin for treatment of IC in a landmark phase III trial [21].Rezafungin will also be a viable option for patients that are incapable of taking oral medication.Ibrexafungerp [22], representing a novel antifungal class with an echinocandin like mechanism of action and excellent oral bioavailability [12], is FDA approved for vulvovaginal candidiasis and has shown promising results for IC in an ongoing phase IIb study [23].Both drugs may facilitate earlier hospital discharge of those patients in whom stepping down to fluconazole is not an option and may overcome the limitations of currently available echinocandins which require daily intravenous administration.In addition, several other antifungal drugs for Candida infections are in clinical development [20], including oteseconazole which was FDA approved for vulvovaginal candidiasis.
Our study comes along with some limitations.Not all requested data were available for all patients, and the presented data reflect a real-life scenario with no predefined fungal diagnostic strategies or treatment protocols.While the geographical distribution of our samples is reflective of Europe including its laboratory capacities [24], it is still likely that settings with better access to diagnostics [25,26] and antifungals are overrepresented.Also exact reasons for prolonged hospitalization to complete iv AF Tx were not assessed, and while some patients may have simply not been able to take oral medication, the fact that the population with prolonged hospitalization to complete iv AF Tx tended to be younger, was less severely ill and mostly on echinocandins argues against this being a major proportion of the study population.Besides resistance to fluconazole (31% of patients had isolates in the intermediate or resistant range), lack of biofilm activity an concerns about tolerability and hepatotoxicity may have been other reasons prolonged hospitalization to complete iv AF Tx.Fourthly, we did not use a standardized definition for the primary outcome of interest ''prolonged hospitalization solely to complete iv AF Tx'', but relied on the treating physicians judgment.This may be considered primarily a strength, as the multiple variables factoring into the decision whether to discharge a patient are difficult to catch in a standardized definition.However, this may also be considered a weakness, as our definition is impacted by regional and national variabilities across Europe in terms of hospital length of stays, discharge criteria and opportunities for outpatient intravenous treatment, thereby limiting applicability for some hospital settings.
In conclusion, we found that across Europe * 1 in 7 patients (16%) with candidemia had prolonged hospitalization specifically to complete iv AF Tx underscoring that no effective oral or long acting iv alternatives are available.C. parapsilosis was not associated with prolonged hospitalization in contrast to C. glabrata, which predicted prolonged hospitalization solely to complete iv AF Tx, and which was also overrepresented among the cohort receiving outpatient iv AF Tx.Limitations could be overcome by new antifungals with oral formulations available, broad antifungal activity, lower rates of resistance and longer half-life.Properties, which may allow for earlier discharge and outpatient therapy reducing costs, risk for nosocomial infections and other inpatient complications.
from Merck and Pfizer and speaker honoraria from Pfizer, Gilead, Astellas, Basilea, Merck, Angelini and Shionogi HH has received speaker honoraria from Gilead Sciences, Sanofi and Novartis as well as honorarium for antifungal workshops through Daman P/S, all outside the submitted work All other authors declare no conflict of interest for this study.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material.If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
IC = invasive candidiasis, SOT = solid organ transplantation, BMI = body mass index, ICU = intensive care unit, ECMO = extracorporeal membrane oxygenation, CRBSI = catheter related bloodstream infection, ID = infectious diseases, LAmB = liposomal amphotericin B, IQR = interquartile range *Echocardiography performed in 55 patients; Ophthalmoscopy performed in 42 patients with prolonged hospital stay solely to complete iv AF Tx. °Based on local susceptibility classifications.

Table 1
Patient characteristics and distribution of Candida species in patients with prolonged hospital stay solely to complete intravenous (iv) antifungal treatment (AF Tx) and in those without (excluding patients who received ambulatory parenteral treatment)