Objectives: Pulmonary fungal infections remain the most important cause of morbidity and mortality in liver transplant recipients. Fast and accurate causative diagnoses are essential for a good outcome. Bronchoscopy with bronchoalveolar lavage frequently is performed to diagnose pulmonary infections in immunocompromised patients. The aim of this study was to evaluate the diagnostic use of bronchoalveolar lavage in liver transplant recipients with pulmonary infections.
Materials and Methods: We retrospectively analyzed the data of 408 patients who underwent liver transplant from January 1990 to December 2012. Patients who underwent bronchoalveolar lavage after transplant were included in this study.
Results: There were 18 of 408 liver transplant recipients (4.41%) who underwent bronchoalveolar lavage after transplant. The mean age was 49.5 ± 18 years. In 5 patients (27.8%), fungal microorganisms were observed in the cytology of bronchoalveolar lavage specimens, including Aspergillus fumigatus in 3 patients and Candida albicans in 2 patients. Death occurred in 4 of 5 patients (80%) with fungal infections. No association was observed between the presence of fungal infection and clinical and radiographic findings of the patients.
Conclusions: Bronchoscopy with bronchoalveolar lavage is a useful, noninvasive diagnostic tool for the rapid diagnosis of infections in solid-organ transplant recipients.
Key words : Aspergillus fumigatus, Candida albicans, Cytology, Pulmonary
Solid-organ transplant recipients are at a high risk for infectious complications, and the most common infections include pulmonary fungal infections.1,2 These infections are associated with major morbidity and mortality. Therefore, rapid diagnosis of fungal infections in immunocompromised patients is important.3 Bronchoscopy with bronchoalveolar lavage (BAL) is an important tool for the diagnosis of pulmonary infections. This study sought to analyze the effectiveness of BAL in establishing the diagnosis of pulmonary infections in liver transplant recipients.
Materials and Methods
This study was approved by the Institutional Review Board at Başkent University. Patients with BAL were selected retrospectively from 408 liver allograft recipients who underwent liver transplant from January 1990 to December 2012 at Başkent University. Clinical findings of these patients were obtained from patient charts including age, sex, immunosuppressive regimen, clinical symptoms, thoracic computed tomography results, bronchoscopic findings, culture results of BAL fluid, complete blood count, age at transplant, and time between transplant and BAL. The baseline (preoperative to postoperative) immunosuppression protocol included tacrolimus or cyclosporine with low-dose steroids. For tacrolimus, the posttransplant target whole blood trough level was 10 to 12 ng/mL during the first 2 weeks, and the dose subsequently was tapered. Sirolimus was administered only in selected patients. Biopsy-proven rejection was treated with steroid bolus therapy.
The data analysis was performed with statistical software (SPSS for Windows, Version 16.0, SPSS Inc., Chicago, IL, USA). Average values were reported as mean ± SD and analyzed with Kruskal-Wallis test and Mann-Whitney U test. Values of P ≤ .05 were considered statistically significant.
During follow-up, only 18 of 408 liver transplant recipients (4.41%) underwent BAL after transplant. Of these 18 patients, 16 patients were (88.9%) male and 2 patients were (11.1%) female. The mean patient age was 49.5 ± 18 years (age range, 10-72 y) at the time of bronchoscopy. The underlying liver diseases were hepatocellular carcinoma secondary to viral hepatitis (8 patients), cryptogenic cirrhosis (3 patients), Wilson disease (2 patients), alcoholic cirrhosis (2 patients), autoimmune hepatitis (2 patients), and Alagille syndrome (1 patients). The graft source was a living-related donor in 10 patients (55.6%) and a deceased donor in 8 patients (44.4%). For immunosuppressive therapy, 6 patients (33.3%) received cyclosporine (mean dose, 162 ng/mL), 9 patients (50%) received tacrolimus (mean dose, 10.78 ng/mL), and 3 patients (16.7%) received sirolimus (mean dose, 10.33 ng/mL). In the 18 patients, acute rejection was observed only in 4 patients (22.2%). There were 8 patients (44.4%) who had diabetes mellitus before liver transplant.
The clinical symptoms were fever in 14 patients (77.8%) and cough in 4 patients (22.2%). In radiographic examination, lung infiltrates were observed in 16 patients (88.9%) and lung nodules were detected in the other 2 patients (11.1%). Bronchoscopic findings included increased secretions (5 patients), necrotic plaque (4 patients), and edematous mucosa (1 patient), and bronchoscopy otherwise was unremarkable (8 patients). In 5 patients (27.8%), fungal microorganisms were observed in BAL samples; 3 fungi were Aspergillus fumigatus and 2 were Candida albicans. In the cytology, Aspergillus was demonstrated with the presence of septate, 45°-branching hyphae, and Candida albicans was observed with pseudohyphae and multiple budding yeasts (Figure 1). The average interval between transplant and infection was 1.33 ± 0.5 months for Aspergillus infections and 1 month for Candida infections. All patients with fungal infection were men aged > 50 years. In the 5 patients, 3 patients were receiving cyclosporine and 2 patients were receiving tacrolimus-based immunosuppressive therapy. None of these patients had an acute rejection episode. The mean white blood cell count was 9.3 ×109/L in patients who had fungal infection and 6.6 ×109/L in patients who did not have fungal infection. Although the white blood cell count was higher in patients with fungal infections, the difference was not statistically significant (P > .05). The clinical and radiographic findings of the patients who had fungal infections in BAL cytology are summarized in the table (Table 1).
In the 18 patients who had BAL, the other 13 patients had nonspecific findings on cytologic examination. Synchronous transbronchial biopsy was taken only in 5 patients; the histologic diagnosis in 4 patients was nonspecific and in 1 patient, histologically invasive aspergillosis was reported which also was observed in BAL cytology (Figure 1). Microbiologic culture was performed in all 18 patients, and only 2 patients (11.1%) had positive culture, including 1 culture that showed Aspergillus fumigatus and the other culture that showed Candida albicans; both had already been detected in cytology. There were 3 patients who had positive cytology for fungi that were not detected in the cultures. In 3 patients who had positive BAL cytology, galactomannan level was investigated and was normal.
All fungal infections were treated with antifungal agents; 3 were treated with amphotericin B and the other 2 were treated with fluconazole and caspofungin. The antifungal treatment was started during the first 5 days in all patients, but shortly after BAL, 4 of the 5 patients who had invasive fungal infection had died. The mortality rate with aspergillosis was 100%, and these patients died at mean 10.66 days (range, 3-24 d) after diagnosis. For Candida, the mortality rate was 50%; 1 patient died after 86 days and the other patient still is alive after 72 months.
In our study, no significant differences were detected in donor type (living-related or deceased-donor transplant), cause of primary liver disease, clinical or radiographic findings, immunosuppressive treatment, or presence of acute rejection between patients who had or did not have fungal infection.
Fungal infections are the most important cause of morbidity and mortality in liver transplant patients.1,2 The most important reason for the high mortality is the difficulty in early diagnosis because of the low index of suspicion and nonspecific clinical and radiographic findings. Therefore, rapid and accurate diagnosis is essential for a good outcome. As a useful tool, flexible bronchoscopy with BAL is simple, safe, fast, and reliable. This procedure has been used extensively as a diagnostic procedure in transplant recipients with suspected pulmonary infections, especially opportunistic infections.3,4
Candida and Aspergillus species are the most important causes of invasive fungal infections in solid-organ transplant recipients, and are associated with almost 100% mortality.1,5 In most studies, the infection started during the first 6 months after transplant.1,2,5,6 In our study, Aspergillus was the most common cause of fungal infection in liver recipients, consistent with other studies,7,8 and the average interval between transplant and infection was 1.33 ± 0.5 months with Aspergillus and 1 month with Candida. Therefore, the time between transplant and fungal infection is shorter in our study than previous reports in the literature. The mortality rate of invasive aspergillosis is 50% to 100%, and the mortality from Candida infections is less than the mortality from Aspergillus infections.2,5,9 In our study, the mortality rate in aspergillosis was 100%, and the patients died at mean 10.66 days after diagnosis. The mortality rate of our patients who had Candida was 50%; 1 patient died at 86 days after diagnosis and the other patient still is alive after 72 months.
The most important risk factors for fungal infections in liver transplant recipients include high-dose immunosuppressive regimens, prolonged duration of transplant surgery, broad spectrum antibiotics, and acute rejection episodes.1,5,6,10 In our study, there was no significant association between the presence of fungal infection and the clinical or radiographic findings or immunosuppressive regimen that the patients received.
The BAL is the preferred method to investigate pulmonary infections in immunosuppressed patients. The procedure has low risk and can be safely performed in most patients, including those with hypoxia. However, the value of the cytologic examination of BAL specimens is controversial because of low sensitivity. In the study of Rañó and coworkers,8 BAL was performed in 135 cases and a specific diagnosis was obtained in 68 cases, with diagnostic yield 51%. The diagnosis of infection was made by culture or cytologic evaluation of BAL specimens. In these cases, 12% cases were fungal infections and the most common organism was Aspergillus fumigatus. As a result, the authors emphasized the importance of the use of noninvasive bronchoscopic procedures, and noted that BAL provided the highest diagnostic yield in diagnosing pulmonary infiltrates in immunocompromised patients.8
Another large study about pulmonary infections diagnosed by BAL was published by Joos and associates11; 1066 immunocompromised patients underwent BAL, including 173 solid-organ transplant recipients, and the diagnostic yield of BAL in diagnosing aspergillosis was 4%. The most common infections in this population (n = 1066) were bacterial and viral.11 Similar results were reported in the study of Reichenberger and coworkers, who evaluated the diagnostic yield of BAL in renal transplant recipients; microorganisms were isolated from 69% BAL cases, but most were bacterial and viral, and Aspergillus was detected only in 1 case.3
Hohenadel and associates12 investigated the role of BAL in 95 immunocompromised patients who had hematologic disorders with pneumonia. Pathologically relevant isolates were observed in 65% cases; BAL provided the definitive diagnosis in only 29 cases (31%), and 11% of these infections were fungal.12 In a study by Al-Za’abi and coworkers in immunosuppressed patients (96% lung transplant recipients), the most common identified organism was Aspergillus species, and 17% Aspergillus cases were detected by BAL cytologic examination.13 However, in a study by Pugliese and associates who investigated fungal infections in all solid-organ recipients (37.7% liver recipients), Candida albicans infections were the most common in their population, and the most frequently contaminated biological sample was from BAL (25.3%).14
Although the number of the patients who underwent BAL in our study was limited (18 of 408 patients), the diagnostic yield of BAL in detecting fungal infections was high (27.8%). The clinical and radiographic findings of the patients in the current study were nonspecific; therefore, BAL was useful in detecting fungal infections in our liver transplant recipients.
In conclusion, the mortality of fungal infections in liver transplant recipients is high. Early diagnosis and aggressive antifungal therapy is essential for successful treatment. Therefore, BAL is valuable as a noninvasive method for detecting fungal micro-organisms in liver transplant recipients.
Volume : 13
Issue : 1
Pages : 331 - 334
DOI : 10.6002/ect.mesot2014.P179
From the Departments of 1Pathology and 2Transplantation
Surgery, Baskent University Faculty of Medicine, Ankara, Turkey
Acknowledgements: The authors declare that they have no sources of funding for this study, and they have no conflicts of interest to declare.
Corresponding author: Merih Tepeoglu, MD, Baskent University, Department of Pathology,
79 Sokak, No. 7/4, Bahcelievler, Ankara 06490, Turkey
Phone: +90 312 212 6591
Fax: +90 312 212 7572
Figure 1. Fungal Microorganisms in Bronchoalveolar Lavage Cytology
Table 1. Clinical and Radiographic Findings in Patients Who Had Fungal Infections Diagnosed in Brochoalveolar Lavage Specimens