Solid Organ Transplants (SOT) provide increased life expectancy and quality of life for various comorbidities of organ failure, with a one-year survival rate of more than 80%1 More than half of Liver Transplant (LT) recipients develop some type of infection post-transplant,2 mostly of bacterial etiology. However, fungal infections are an important cause of morbidity and mortality after SOT. The highest incidence of Invasive Fungal Infections (IFI) among SOT occurred in liver recipients in a TRANSNET cohort, with 339 cases (31.8%). The emergence of new pathogens with different susceptibility profiles renders it essential to review antifungal prophylaxis protocols in transplant services. In the 1990s, the incidence of IFI varied from 6% to 47% in LT recipients, with high mortality rates3,4 With the advent of better surgical techniques, adjustments in immunosuppression, and preventive strategies such as antifungal prophylaxis, these rates have been reduced, reaching incidence rates as low as 4% in a multicenter American study1,5,6 The mortality rate from invasive candidiasis in LT is 30.9% according to TRANSNET data1

In studies of risk factors for IFI post-LT, several factors were found, with results sometimes differing among studies. The most commonly cited risk factors are prolonged surgical time, reoperation, retransplantation, need for renal replacement therapy, fulminant hepatitis, choledochojejunostomy, Candida spp. colonization, extensive transfusions of blood components, and higher MELD score3,4,6–12 Studies in different centers identify different risk factors, with no consensus in the literature. Currently, guided antifungal prophylaxis is recommended only for patients considered at high risk for IFI. Some studies have already shown the safety of guided prophylaxis compared to universal prophylaxis3,13,14

It is important to define which patients will benefit from prophylaxis, determining poorly established risk factors for IFI in the post-LT period to facilitate proper prophylaxis, promoting greater graft and patient survival.

The main objective was to identify the risk factors associated with IFI in the first 100-days after LT, as well as to identify the risk factors associated with death in the studied cohort in the first 100-days after LT.

This is a retrospective multicenter case-control study conducted in two liver transplant reference hospitals in the city of São Paulo: Hospital das Clínicas da FMUSP (HCFMUSP) and Hospital Israelita Albert Einstein ‒ Unidade Vila Santa Catarina (HIAE-VSC). We included patients undergoing LT from January 2002 to December 2017 presenting with an IFI diagnosed within the first 100-days after transplantation, according to the definitions published by the EORTC in 202015 Patients who died within the first 48-hours after liver transplantation and IFI with positive cultures collected intraoperatively or within the first 24-hours after transplantation were excluded.

From 2002 to 2017, a total of 2759 liver transplants were performed in both centers. We excluded 107 patients who died within 48-hours post-transplant and 259 retransplant records, resulting in 2393 eligible subjects for the study. Cultures identified 66 cases and 198 controls across both centers, as depicted in Fig. 1 outlining patient inclusion and exclusion criteria.

Fig. 1.

Flowchart for identifying cases and controls. HCFMUSP, Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo; HIAE-VSC, Hospital Israelita Albert Einstein Vila Santa Catarina.

Clinical and demographic data of study cases and controls are presented in Table 1. Most patients were male in both groups. Median age was 51-years for cases and 54-years for controls. Median MELD scores were 21 for cases and 20 for controls. The most common causes of cirrhosis were chronic hepatitis C (34.9% of IFI cases) and alcoholic cirrhosis (19.7%). Hepatocellular carcinoma was present in 33% of IFI cases, with 4.5% of transplants were due to fulminant hepatitis. Three percent involved simultaneous liver-kidney transplants. Significant differences between groups were noted in median length of hospital stay and ICU admission: 42 and 12.5-days for cases, compared to 18.5 and 5-days for controls, respectively.

Among the 66 IFI cases, 21 (31.8%) resulted in death due to IFI, according to medical record data. Of 264 evaluated patients, 146 received antifungal prophylaxis, predominantly fluconazole (67.8% of cases). No difference was observed between fluconazole and amphotericin B prophylaxis. Among 101 patients undergoing RRT, 7/101 (6.9%) developed IFI without antifungal prophylaxis; 25/101 (24.8%) patients developed IFI despite prophylaxis, with no difference in IFI rate noted among different antifungal agents.

Table 2 details fungal agents and infection sites. Candida spp. predominated (58-cases), mostly non-albicans (57.6%), followed by Candida albicans (30.3%). Filamentous fungi included Aspergillus spp. (10.6%) and one Mucor spp. infection. Intrabdominal was the primary infection site (54%), followed by bloodstream infections (27%).

The distribution of IFI cases from 2002 to 2017 showed no statistical difference across years (Fig. 2, p p = 0.346).

Fig. 2.

The annual distribution of cases and controls.

Bivariate analysis identified several risk factors for IFI occurrence (Table 3), including antifungal prophylaxis, deceased donor, reoperation, renal replacement therapy, MELD > 30, antibiotic use, retransplantation, hospital stay > 30-days, graft vascular complications, and primary graft non-function. Dialytic renal dysfunction and early reoperation within 100-days were independent risk factors for IFI occurrence. Antifungal prophylaxis and deceased donor status were protective against IFI.

Bivariate analysis for death within 100-days post-transplant identified significant risk factors: IFI development, MELD > 30, ≥ 5 red blood cell transfusions, renal dysfunction with or without dialysis, vascular complications, primary graft non-function, and early graft dysfunction. Multivariate analysis identified IFI, vascular complications, and renal dysfunction within the first 100-days post-transplant as independent risk factors for mortality (Table 4).

Identifying risk factors associated with the development of IFI is crucial for implementing appropriate screening and tailored use of antifungal prophylaxis, thereby reducing morbidity and mortality following LT. In our study across two major LT centers in São Paulo, risk factors for IFI occurrence within 100-days post-transplant were identified as renal replacement therapy and reoperation. Protective factors included the use of antifungal prophylaxis and deceased donor transplantation. Regarding mortality within the same period, IFI occurrence, graft vascular disorders, and renal dysfunction were identified as worse prognostic factors.

Renal replacement therapy as a risk fator for IFI occurrence is accordance to previously studies (colocar as referencias Gianella, Melenotte e Fortun). In our cohort, 101 subjects underwent renal replacement therapy, and 32 IFI cases were identified in this population, with only 7 not receiving prophylaxis. Despite prophylactic antifungal use, breakthrough infections occurred in 25-cases. Cases of IFI occurring despite fluconazole prophylaxis may have occurred by potential variability in serum levels during renal replacement therapy, affecting drug pharmacokinetics and pharmacodynamics18

Reoperation was also identified as an independent risk factor for IFI. In a literature review, reoperation and laparotomy are associated with up to a 10-fold increase in IFI risk19 In an Australian case-control study investigating early IFI risk factors, reoperation was also independently associated in multivariate analysis, along with biliary-enteric anastomosis, and intraoperative transfusion of ≥40-units. Intestinal manipulation during reoperations or retransplantation predisposes to translocation of Candida species into the peritoneal cavity or bloodstream13 In our center, reoperation is considered a minor criterion for initiating antifungal prophylaxis; however, this study underscores its importance as a significant risk factor, and resulted in a protocol modification, and we have included reoperation as a major criterion for antifungal prophylaxis.

Contrary to the literature, our study did not identify fulminant hepatitis and retransplantation as independent risk factors for IFI, despite these being classically associated factors4,6,14,20 Retransplantation can increase the risk of invasive aspergillosis up to 29-fold19 Both participating centers consider these indications for prophylaxis. Therefore, the use of prophylaxis in our patients diagnosed with fulminant hepatitis and retransplantation may have proven effective in preventing IFI development in our cohort.

Regarding antifungal prophylaxis, Neyra et al.'s cohort did not find a difference in IFI incidence between high-risk patients receiving or not receiving prophylaxis (fluconazole or micafungin). The number needed to treat to prevent one IFI case was 105, however, IFI cases in this sample were small21 Winston et al. conducted a randomized clinical trial comparing anidulafungin and fluconazole in high-risk IFI patients, finding no significant difference, though no cases of aspergillosis occurred in the anidulafungin group. “Breakthrough” cases in the fluconazole group included 5 out of 8 Candida spp. isolates resistant to the drug9 Kang et al., in a study across 5 centers in South Korea, found no inferiority in micafungin prophylaxis compared to fluconazole universal prophylaxis in LDLT22 However, Fortun et al., in a multicenter study, compared caspofungin and fluconazole, identifying superiority in preventing breakthroughs in the caspofungin group, with no difference in mortality10 These findings are reinforced by a meta-analysis by Gatti et al., demonstrating non-inferiority of echinocandins, but a higher tendency toward breakthroughs in echinocandin groups in renal replacement therapy patients when excluding the previously mentioned Fortun et al. study23

The importance of studies comparing echinocandins with fluconazole use arises in the context of emerging Candida spp. strains resistant to fluconazole, as well as their lack of interaction with calcineurin inhibitors and potential activity against aspergillosis. The exclusively intravenous presentation and high cost are limitations for routine echinocandin use in IFI prevention. The incidence of Aspergillus spp. and other filamentous fungi infections is relatively low in these studies24,25 Literature consensus supports Candida spp. prophylaxis in high-risk patients, though drug, dosage, and duration are debated. For instance, fluconazole use varies from 21 to 60-days, at doses ranging from 100 to 400 mg20 In this cohort, Candida spp. were the main identified agents, primarily affecting the intra-abdominal site, consistent with other literature series3 The incidence of Aspergillus spp. infection in our two evaluated centers was low, with most occurring during a period of construction work in the hospital. Based on our study results, maintaining the current service antifungal prophylaxis protocol appears safe.

In our study, deceased donor transplantation was also identified as a protective factor against IFI, suggesting that Living Donor Transplantation (LDLT) may be associated with increased fungal infection risk. In our cohort, 19 LDLT cases were performed, with 42.1% developing IFI. The increased risk of IFI in LDLT may be associated with surgical complexity. In this technique, the graft is smaller with narrower blood vessel and bile duct diameters, potentially increasing the risk of biliary complications such as fistulas and strictures3,22,26 Due to the small number of living-donor recipients in this sample, the analysis of this finding may be overestimated. Therefore, studies including a larger population of LDLT are needed to better evaluate this finding. These are under-recognized risk factors in the literature and warrant further study and consideration for prophylactic measures in this setting.

In this cohort, controls weren’t selected through matched randomization. Although no statistically significant differences were observed between the groups, this should be considered a limitation of the study. Additionally, the small number of living-donor transplants in this cohort may have influenced the observed findings.

The main limitations of this study include the small number of fungal infections; a larger cohort might have identified different risk factors. Fungal infections identified exclusively by anatomopathological examination were also not included. The method used to identify IFI cases relied on positive cultures meeting the 2020 EORTC criteria, this choice excluded other IFI cases diagnosed with indirect evidence that could meet probable or possible criteria, such as invasive aspergillosis diagnosed by galactomannan or cryptococcosis diagnosed by latex agglutination. Infections like pneumocystosis, not commonly identified via cultures, were also excluded.

Another limitation is its retrospective nature, retrieving data from 2002 onwards, a time when routine antifungal sensitivity testing was not performed, precluding assessment of whether fungal disease cases occurring in patients on prophylaxis were associated with resistance patterns.

We identified that the use of antifungal prophylaxis and deceased donor were protective factors against IFI development within the first 100-days post-transplant. No superiority was observed between the regimens used. The lack of superiority among prophylactic regimens is consistent with the predominance of Candida spp. as the main causative agents, rather than filamentous fungi. Breakthrough infections occurred mainly in patients under renal replacement therapy, possibly due to underdosing of antifungals.

We observed that Candida spp. were the primary agents of IFI. When aggregating all non-albicans Candida species, they were the most frequently identified; however, upon separate species evaluation, C. albicans was the most isolated species. The main site of invasive fungal disease was the intra-abdominal site. The data from our cohort are consistent with series found in the literature.

The renal replacement therapy and reoperation were considered risk factors. Regarding mortality, IFI was a risk factor within the first 100-days post-liver transplant, as well as vascular graft complications and renal dysfunction.

ALT, Alanine Aminotransferase; AST, Aspartate Aminotransferase; CMV, Cytomegalovirus; EORTC/MSGERC, European Organization for Research and Treatment of Cancer and the Mycoses Study Group Education and Researh Consortium; FMUSP, Faculdade de Medicina da Universidade de São Paulo; HBV, Viral Hepatitis B; HCFMUSP, Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo; HCV, Viral Hepatitis C; HIAE-VSC, Hospital Israelita Albert Einstein ‒ Unidade Vila Santa Catarina; HR, Hazard Ratio; IC, Confidence Interval; ICU, Intensive Care Unit; IFI, Invasive Fungal Infections; IQR, Interquartile Interval; LT, Liver Transplant; MELD, Model for End-Stage Liver Disease; OR Odds Ratio; PCR, Polymerase Chain Reaction; RR, RRT, Relative Risk Renal Replacement Therapy; LDLT, Living Donor Transplantation; SBP, Spontaneous Bacterial Peritonitis; SOT Solid Organ Transplants; UI, Units.

Data must be requested from the corresponding author.

All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Larissa N de A Gouveia, Bruno G P Romeo and Julia Rezende. Analysis was performed by Maristela P. Freire, Alice T W Song and Larissa N de A Gouveia. The first draft of the manuscript was written by Larissa N de A Gouveia, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

There was no funding for this work.