- Email: [email protected]
Evaluation of Pulmonary Infections in Solid Organ Transplant Patients: 12 Years of Experience
Evaluation of Pulmonary Infections in Solid Organ Transplant Patients: 12 Years of Experience lu, E. Küpeli, S¸.S. Bozbas¸, Z.E. Özen, E.S. Akkurt, C. Aydog an, G. Ulubay, S¸. Akçay, F.Ö. Eyübog and M. Haberal ABSTRACT Background. Recipients of solid organ transplants (SOTs) are at higher risk to develop pulmonary infections (PIs) owing to their immunocompromised state. Flexible bronchoscopy (FB) is frequently performed to diagnose nature of these infections. The aim of 12-year review was to evaluate the demographic characteristics of SOT recipients with PIs and to study diagnostic utility of FB in this group of patients. Methods. Medical records of patients who underwent SOT as well as FB between 2000 and 2012 were retrospectively reviewed. Patients’ demographics, type of transplantation, primary diagnoses, thoracic computed tomography (CT) findings, total blood count and chemistry, indication for FB, FB results, specimen culture results, and suspected and final diagnoses were all recorded. If the bronchoscopy findings altered medical management and produced improvement in PI, the procedure was considered diagnostic. Results. Ninety of 998 liver, heart, or kidney transplant recipients underwent FB (73 renal, 16 liver, and 1 heart; mean age, 42.3 12.1 years) during the study period. CT findings were as follows: Consolidation (49.4%), lymphadenopathy (3.4%), nodular infiltrates (5.6%), and cavitary lesion (1.1%). FB was unremarkable in 29, but showed increased secretions in 33 patients (36.7%), chronic mucosal changes in 9 (10%), edema in 7 (7.8%), mucosal plaque in 7 (7.8%), friable mucosa in 3 (3.3%), and endobronchial lesion in 2 (2.2%). A total of 29 bronchial washings (BW; 32.6%) and 10 bronchoalveolar lavages (BAL; 11.2%) were performed. PI was diagnosed in 82% of the patients (n ¼ 73). In 32 patients (36%), micro-organism growth was observed on either BW or BAL. Mycobacterium tuberculosis was detected in 6 (6.7%), Staphylococcus aureus in 4 (4.4%), Moraxella catharralis in 4 (4.4%), Candida albicans in 6 (6.7%), Klebsiella pneumonia in 2 (2.2%), Escherichia coli in 2 (2.2%), Streptococcus pneumoni in 2 (2.2%), Stenotrofomonas maltofilia in 1 (1.1%), Aspergillus fumigatus in 4 (4.5%), and Pseudomonas aeruginosa in 1 (1.1%). Final diagnosis was established by FB (n ¼ 33) with a diagnostic yield of 36%. No significant finding was observed between the type of the transplant and the culture results (P > .05). Conclusion. Suspected PI is the most common indication for FB in SOT recipients. It may identify the causative organism in >30% of patients. Tuberculosis was found to be the most frequent agent, which is not surprising from such an endemic area. Bacteria were more common than fungal or viral micro-organisms. FB should be considered in SOT recipients presenting with lung infiltrates and suspected to have PI.
From the Departments of Pulmonary Medicine (F.Ö.E., E.K., S¸.S.B., Z.E.Ö. E.S.A., G.U., S¸.A.), General Surgery (C.A., M.H.), Baskent University School of Medicine, Ankara, Turkey.
lu, MD, Address correspondence to Füsun Öner Eyübog Department of Pulmonary Diseases, Baskent University Hospital, 5 Sokak No: 48 06490-Bahçelievler, Ankara, Turkey. E-mail: [email protected]
0041-1345/13/$esee front matter http://dx.doi.org/10.1016/j.transproceed.2013.09.024
ª 2013 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710
3458
Transplantation Proceedings, 45, 3458e3461 (2013)
PULMONARY INFECTIONS IN SOT
A
LTHOUGH THE INCIDENCE of pulmonary infections (PIs) after transplantation has declined with effective prophylactic strategies and refinements in immunosuppressive regimens, it account for significant mortality and morbidity among solid organ transplant (SOT) recipients. Most PIs present with lung infiltrates. Approximately two thirds of lung infiltrates are of infectious origin among recipients of SOT. In addition to the symptomatology and radiologic findings, microbiological findings on airway or pleural specimen help establish the diagnosis. Flexible bronchoscopy (FB) is frequently performed to evaluate nature of the PIs in this group. Performing bronchoscopy is a major component in the management of SOT recipients with lung infiltrates as it has been reported to have high diagnostic yield.1,2 The aim of this 12-year review was to evaluate the characteristics of SOT patients with PI and study the diagnostic utility of FB among these patients. MATERIALS AND METHODS The medical records of SOT recipients who underwent FB between 2000 and 2012 at our institution were reviewed retrospectively. Information on patients’ demographics, type of transplantation, primary diagnoses, and results of thorax computed tomography (CT), total blood count and chemistry, immunosuppressant therapy, indications for FB, FB results, antibiotic use, and suspected and final diagnoses were recorded. Patients with positive serology for HIV were excluded. For study purposes, PI was described as presence of new pulmonary infiltrates with either fever or leukocytosis or both. Bronchoscopy were performed in the presence of localized, multiple, bilateral, or diffuse pulmonary infiltrates or in the absence of a clinical or radiologic response to antimicrobial therapy. The institutional ethics committee approved this prospective study.
3459 Table 1. Systemic Diseases and Comorbidities in Solid Organ Transplant Recipients (n [ 49) Variable
n
%
Hypertension Diabetes mellitus Hepatitis C Diabetes mellitus and hepatitis C Diabetes mellitus and hypertension Malignancy Coronary artery disease Cardiomyopathy Systemic lupus erythematosus Malignancy and hypertension Hepatitis C and hypertension Cardiac cirrhosis and renal failure
14 10 7 5 4 3 1 1 1 1 1 1
15.6 11.1 7.8 5.6 4.4 3.3 1.1 1.1 1.1 1.1 1.1 1.1
Candida albicans, Klebsiella pneumonia, Escherichia coli, Streptococcus pneumonia, Stenotrofomonas maltofilia, Aspergillus fumigatus, and Pseudomonas aeruginosa, and led to alteration in the medical management. FB was considered diagnostic if a change in the medical management produced improvement in lung infiltrates and clinical status. The FB was introduced to the endobronchial tree without using any suction.
Statistical Techniques All continuous data were expressed as mean values standard deviation. Data were analyzed with SPSS software (Statistical Package for the Social Sciences, version 15.0, SSPS Inc., Chicago, IL). The c2 test and t test were used to compare the parameters of the drug levels and the culture results. Multivariate analysis was performed. All parameters were expressed as mean values standard deviation. P < .05 was considered significant.
Technical Information
RESULTS
All FB were performed in an inpatient bronchoscopy suite or in the intensive care unit with appropriate cardiopulmonary monitoring. The choice of sedating medications was at the discretion of the bronchoscopist. Local anesthesia was provided with 2% lidocaine instilled through the working channel of the bronchoscope. The bronchoscope was introduced into the lower airways without using suction to prevent contamination of the channel with the upper respiratory tract flora. The choice of sampling procedure during FB was also at the bronchoscopist’s discretion. The standard sampling techniques used during FB were bronchoalveolar lavage (BAL) and bronchial washings (BW). BAL was performed according to the American Thoracic Society Guidelines and analyzed according to our standard protocol as described previously.3,4 BAL and BW specimens and tissue biopsies were obtained if needed. Grame and Ziehl-Neelsenestained respiratory samples were cultured for bacterial pathogens, fungi, and mycobacterial agents. Fungal pneumonia was diagnosed in the presence of fungal hyphae identified by cytopathologic evaluation through the use of bronchoscopic lavage samples. Pneumonia owing to cytomegalovirus and other viruses were considered if they were isolated by cell culture from BAL fluid, elevated titers of antibody were discovered in serum or BAL, and inclusion bodies were present on cytopathologic evaluation. Information obtained from BAL, BW, and biopsies were considered positive when any of the following organisms were identified: Mycobacterium tuberculosis, Staphylococcus aureus, Moraxella catharralis,
According to our patient records, 90 of 998 SOTs (9%) underwent FB with suspected PI during the study period. Of these 90 patients, 16 had liver, 73 had kidney, and 1 had heart transplantation. The mean age was 42.3 12.1 years with a minimum value of 18 and a maximum of 64. Fiftythree of the transplanted organs (58.9%) were from living donors and 37 (41.1%) were cadaveric. Among all transplants, 49 patients (54.4%) had different comorbid systemic diseases (Table 1). Smoking history was present in 54 patients (60%). The most predominant thoracic CT finding was consolidation (49.4%), followed by pleural effusion (12%), nodular infiltration (5.6%), lymphadenopathy (3.4%), and cavitary lesions (1.1%). All patients were on empiric antibiotics at the time of FB. Protected specimen brush (PSB) was not used because of the unavailability of the instrument. Transbronchial biopsy (TBBx) was not performed owing to lack of fluoroscopy equipment and multiple comorbidities among the majority of SOT recipients. FB results were unremarkable in 32.2% (n ¼ 29). Other findings were increased secretions in 33 (36.7%33), chronic mucosal changes in 9 (10%), mucosal edema in 7 (7.8%), mucosal plaques in 7 (7.8%7), mucosal necrosis in 3 (3.3%),
3460
and endobronchial lesion in 2 (2.2%). Either BAL or BW was obtained in all patients. BW revealed positive cultures in 22 patients (24.4%), and BAL in 10 (11.2%), for an overall total of 32 patients (36%). In 7 patients, both BAL and BW cultures grew the same organisms (concordance). In the remainder of the patients, BAL and BW cultures were either negative or grew normal flora. Isolated microorganisms were as follows with different percentages: M tuberculosis 6.7% (n ¼ 6), S aureus 4.4% (n ¼ 4), M catharralis 4.4% (n ¼ 4), C albicans 6.7% (n ¼ 6), K pneumonia 2.2% (2), E coli 2.2% (n ¼ 2), S pneumonia 2.2% (n ¼ 2), S maltofilia 1.1% (1), A fumigatus 4.5% (n ¼ 4), and P aeruginosa 1.1% (n ¼ 1). This information led to successful alteration in medical management among these patients. In other words, new diagnosis was established by FB in 32 patients with a therapeutic impact of 36%. We observed no significant difference between the type of the transplant and the culture results (P > .05). PI was diagnosed in 82% of patients (n ¼ 73). Fifty-nine patients were renal transplant recipients. No significant difference was observed between the PI development and the age, gender, laboratory findings, or smoking status of the patient. PI development had no association with the type of the transplantation (either live or cadaver) and the rejection status (P > .05). No significant relationship was observed between the type of the transplantation and the patients with PI (P ¼ .52). Patients with PI who were treated either in hospital or in the intensive care unit when compared with the ones who did not have PI showed no difference (P ¼ .015). No association was observed between the PI development and comorbidities (P ¼ .08). Mortality was higher among patients with PI than the others (24.7% [n ¼ 22] vs 2.2% [n ¼ 2]). Mortality was also higher among patients treated in the intensive care unit (P ¼ .04). There were no differences between survival and PI development (P ¼ .15). DISCUSSION
PIs in SOTs constitute a medical emergency. Immediate, appropriate, empiric or specific antimicrobial treatment is mandatory. Because the spectrum of potential pathogens is far more diverse than in immunocompetent hosts, a systematic approach to the management of these patients is required. This approach should include a comprehensive diagnostic evaluation including history, physical examination, chest radiography, sputum studies, and blood cultures. A CT scan of the lung is usually indicated in patients in whom typical clinical or specific radiographic findings are missing. Bronchoscopy is indicated in patients with pulmonary infiltrates, unusual clinical and radiographic presentations, and those with treatment failure to empiric antibiotics.1 Among 998 transplant recipients, 90 underwent bronchoscopy to rule out and isolate pneumonia in our cohort. All were treated for pneumonia, even with negative culture growth. In the literature, respiratory tract infections were reported as 8.9% of all infectious episodes among this
EYÜBOGLU, KÜPELI, BOZBAS¸ ET AL
patient group; the incidence of pneumonia varied between 8% and 18%. We believe that the option for the prophylactic use of antibiotics may justify the differences observed in different studies regarding the incidence of respiratory tract infections. In our study, a positive history of smoking and acute rejection episodes in patients increased pneumonia risk. In addition, the presence of comorbidities has additional effects on the development of pneumonia. These factors have a potential role to delay the treatment period. Among 17 patients with rejection, 12 (13%) received pulse steroid therapy. The presence of cough, sputum, and fever was not significant among patients associated with PI in general or specific infections. It has been stated that if the bronchoscopy is performed before commencing empiric antibiotics, it provides greater diagnostic yield.5 Yet, this practice is not possible consistently owing to a variety of factors. Second, the bronchoscope should be introduced in the lower airways without using the suction to prevent contamination from the upper airways. BAL is the most frequently obtained specimen for the evaluation of lung infiltrates in immunocompromised patients. A comprehensive testing including stains and cultures for bacteria, mycobacterium, fungi, higher bacteria, and viruses should be carried out according to the endemicity of the organisms.5 If available, obtaining quantitative cultures on the BAL fluid or PSB specimen is also beneficial.5 Performing TBBx, if no contraindication exists, can add 10% to the diagnostic yield of the procedure.1 Tissue culture on the TBBx specimen if of limited value and is rarely performed.1 Occasionally, transbronchial needle aspiration may be helpful if mediastinal involvement is present.2 Therapeutic impact of bronchoscopy was significant; we were able to isolate agents of pneumonia in our patient group. PIs occur at predictable intervals in SOTs. The first month immediately after transplantation is influenced predominantly by nosocomial bacterial infections; over the following 6 months, opportunistic infections are high. After 6 months, community-acquired micro-organisms are responsible for PIs.1 Cases of M tuberculosis are mainly reported in areas of high endemicity such as Turkey. Nontuberculous mycobacterial pneumonias have been reported generally after the first year. Empiric antibiotic therapy should be targeted to the underlying suspected pathogens depending on several factors, including underlying immunosuppression and concurrent prophylactic therapy of SOT patients.6 It is well-established that empiric antibiotics do not cover causative organisms in >30% of SOTs. In a study by Ibrahim et al,7 42% of patients’ empiric antibiotherapy failed to cover concrete etiology. In our retrospective study, we determined that only 34% of our SOT recipient with PI empiric antibiotic treatment covered the causative organisms; among the remainder, further investigation similar to other studies was required. New diagnosis was established by FB in 32 (36%) of our SOT patients with a therapeutic impact. This was despite ongoing empiric antibiotic therapy. It is also reported that
PULMONARY INFECTIONS IN SOT
endobronchial findings can be helpful in supporting the suspected diagnosis of PI, especially in cases of mycobacterial or fungal infection. In one third of our patients, endobronchial findings were nonspecific. Nevertheless, these findings suggested that increased purulent secretions predominantly support PI. In our patient group, isolated microorganisms by FB were M tuberculosis (6.7%), S aureus (4.4%), M catharralis (4.4%), C albicans (6.7%), K pneumonia (2.2%), E coli (2.2%), S pneumonia (2.2%), S maltofilia (1.1%), A fumigatus (4.5%), and P aeruginosa (1.1%). The incidence of tuberculosis in transplant populations is reported to be 20 to 75 times greater than that of the general population based on endemicity. These results were despite of ongoing empiric antibiotic therapies. In areas with low endemicity, the prevalence of tuberculosis among SOT recipients is 0.5% to 6.4% in contrast with 15.2% in endemic countries.8 All of our M tuberculosis patients underwent SOT 1 to 6 years before the diagnosis. In addition, these patients were referred to us from cities with higher tuberculosis endemicity in our country. The remainder of PIs was detected at the predictable interval. We observed no difference between the type of the SOT and the culture results or type of PI. The value of FB in our SOT recipients was in line with previous studies with a therapeutic impact of 36%.9 Overall mortality in SOTs with PI was found to be between 21% and 35%; however, differences in mortality between nosocomial and community-acquired infection were extreme at 58% and 8%, respectively.9,10 We speculate that, given an empiric antibiotic therapy without waiting, culture results and using effective prophylactic antibiotic regimens might have decreased our mortality rate in patients with PI. In conclusion, suspected PI is the most common indication for FB in SOT recipients. Although endobronchial findings are nonspecific, FB may identify causative organism in >30% of patients despite ongoing empiric treatment. It can be assumed that yield could have been higher if FB was performed before ordering therapy. Bacteria were the most
3461
common pathogens versus fungal or viral micro-organisms, but tuberculosis should always kept in mind in areas with high endemicity. Our study has several weaknesses in addition to being retrospective in nature. The majority of our patients were mainly renal and liver transplant recipients, limiting its observations and applications. In addition, we were not able to perform PSB nor TBBx. Quantitative cultures were also not performed. We do, however, feel that our findings are applicable to the majority of the practices because these limitations are not unique to our institution. REFERENCES 1. Küpeli E, Eyüboglu FÖ, Haberal M. Pulmonary infections in transplant recipients. Curr Opin Pulm Med. 2012;18:202e212. 2. Baughman RP, Lower EE. Diagnosis of pneumonia in immunocompromised patient. In: Agusti C, Torres A, eds. Pulmonary Infection in the Immunocompromised Patient. New York: Wiley; 2009:53e80. 3. American Thoracic Society. Clinical role of bronchoalveolar lavage in adults with pulmonary disease. Am Rev Respir Dis. 1990;142:481e486. 4. Lee FYW, Mehta AC. Basic techniques in flexible bronchoscopy. In: Wang KP, Mehta AC, eds. Flexible Bronchoscopy. Cambridge, UK: Blackwell Science; 1995:95e118. 5. Jain P, Sandur S, Meli Y, et al. Role of flexible bronchoscopy in immunocompromised patients with lung infiltrates. Chest. 2004;125:712e722. 6. Kottloff RM, Ahya VN, Crawford SW. Pulmonary complications of solid organ transplantation and hematopoietic stem cell transplantation. Am J Respir Crit Care Med. 2004;170:367e416. 7. Ibrahim EH, Ward S, Sherman G, et al. Experience with a clinical guideline for the treatment of ventilator-associated pneumonia. Crit Care Med. 2001;29:1109e1115. 8. Arslan H, Ergin F, Öner Eyüboglu F, et al. Tuberculosis in renal transplant recipients. Transplant Proc. 2003;35:2670e2681. 9. Bonatti H, Pruet TL, Brandacher G, et al. Pneumonia in solid organ recipients: spectrum of pathogens in 217 episodes. Transplant Proc. 2009;41:371e374. 10. Cervera C, Agusti C, Angeles Marcos M, et al. Microbiologic features and outcome of pneumonia in transplanted patients. Diagn Microbiol Infect Dis. 2006;55:47e54.
From the Departments of Pulmonary Medicine (F.Ö.E., E.K., S¸.S.B., Z.E.Ö. E.S.A., G.U., S¸.A.), General Surgery (C.A., M.H.), Baskent University School of Medicine, Ankara, Turkey.
lu, MD, Address correspondence to Füsun Öner Eyübog Department of Pulmonary Diseases, Baskent University Hospital, 5 Sokak No: 48 06490-Bahçelievler, Ankara, Turkey. E-mail: [email protected]
0041-1345/13/$esee front matter http://dx.doi.org/10.1016/j.transproceed.2013.09.024
ª 2013 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710
3458
Transplantation Proceedings, 45, 3458e3461 (2013)
PULMONARY INFECTIONS IN SOT
A
LTHOUGH THE INCIDENCE of pulmonary infections (PIs) after transplantation has declined with effective prophylactic strategies and refinements in immunosuppressive regimens, it account for significant mortality and morbidity among solid organ transplant (SOT) recipients. Most PIs present with lung infiltrates. Approximately two thirds of lung infiltrates are of infectious origin among recipients of SOT. In addition to the symptomatology and radiologic findings, microbiological findings on airway or pleural specimen help establish the diagnosis. Flexible bronchoscopy (FB) is frequently performed to evaluate nature of the PIs in this group. Performing bronchoscopy is a major component in the management of SOT recipients with lung infiltrates as it has been reported to have high diagnostic yield.1,2 The aim of this 12-year review was to evaluate the characteristics of SOT patients with PI and study the diagnostic utility of FB among these patients. MATERIALS AND METHODS The medical records of SOT recipients who underwent FB between 2000 and 2012 at our institution were reviewed retrospectively. Information on patients’ demographics, type of transplantation, primary diagnoses, and results of thorax computed tomography (CT), total blood count and chemistry, immunosuppressant therapy, indications for FB, FB results, antibiotic use, and suspected and final diagnoses were recorded. Patients with positive serology for HIV were excluded. For study purposes, PI was described as presence of new pulmonary infiltrates with either fever or leukocytosis or both. Bronchoscopy were performed in the presence of localized, multiple, bilateral, or diffuse pulmonary infiltrates or in the absence of a clinical or radiologic response to antimicrobial therapy. The institutional ethics committee approved this prospective study.
3459 Table 1. Systemic Diseases and Comorbidities in Solid Organ Transplant Recipients (n [ 49) Variable
n
%
Hypertension Diabetes mellitus Hepatitis C Diabetes mellitus and hepatitis C Diabetes mellitus and hypertension Malignancy Coronary artery disease Cardiomyopathy Systemic lupus erythematosus Malignancy and hypertension Hepatitis C and hypertension Cardiac cirrhosis and renal failure
14 10 7 5 4 3 1 1 1 1 1 1
15.6 11.1 7.8 5.6 4.4 3.3 1.1 1.1 1.1 1.1 1.1 1.1
Candida albicans, Klebsiella pneumonia, Escherichia coli, Streptococcus pneumonia, Stenotrofomonas maltofilia, Aspergillus fumigatus, and Pseudomonas aeruginosa, and led to alteration in the medical management. FB was considered diagnostic if a change in the medical management produced improvement in lung infiltrates and clinical status. The FB was introduced to the endobronchial tree without using any suction.
Statistical Techniques All continuous data were expressed as mean values standard deviation. Data were analyzed with SPSS software (Statistical Package for the Social Sciences, version 15.0, SSPS Inc., Chicago, IL). The c2 test and t test were used to compare the parameters of the drug levels and the culture results. Multivariate analysis was performed. All parameters were expressed as mean values standard deviation. P < .05 was considered significant.
Technical Information
RESULTS
All FB were performed in an inpatient bronchoscopy suite or in the intensive care unit with appropriate cardiopulmonary monitoring. The choice of sedating medications was at the discretion of the bronchoscopist. Local anesthesia was provided with 2% lidocaine instilled through the working channel of the bronchoscope. The bronchoscope was introduced into the lower airways without using suction to prevent contamination of the channel with the upper respiratory tract flora. The choice of sampling procedure during FB was also at the bronchoscopist’s discretion. The standard sampling techniques used during FB were bronchoalveolar lavage (BAL) and bronchial washings (BW). BAL was performed according to the American Thoracic Society Guidelines and analyzed according to our standard protocol as described previously.3,4 BAL and BW specimens and tissue biopsies were obtained if needed. Grame and Ziehl-Neelsenestained respiratory samples were cultured for bacterial pathogens, fungi, and mycobacterial agents. Fungal pneumonia was diagnosed in the presence of fungal hyphae identified by cytopathologic evaluation through the use of bronchoscopic lavage samples. Pneumonia owing to cytomegalovirus and other viruses were considered if they were isolated by cell culture from BAL fluid, elevated titers of antibody were discovered in serum or BAL, and inclusion bodies were present on cytopathologic evaluation. Information obtained from BAL, BW, and biopsies were considered positive when any of the following organisms were identified: Mycobacterium tuberculosis, Staphylococcus aureus, Moraxella catharralis,
According to our patient records, 90 of 998 SOTs (9%) underwent FB with suspected PI during the study period. Of these 90 patients, 16 had liver, 73 had kidney, and 1 had heart transplantation. The mean age was 42.3 12.1 years with a minimum value of 18 and a maximum of 64. Fiftythree of the transplanted organs (58.9%) were from living donors and 37 (41.1%) were cadaveric. Among all transplants, 49 patients (54.4%) had different comorbid systemic diseases (Table 1). Smoking history was present in 54 patients (60%). The most predominant thoracic CT finding was consolidation (49.4%), followed by pleural effusion (12%), nodular infiltration (5.6%), lymphadenopathy (3.4%), and cavitary lesions (1.1%). All patients were on empiric antibiotics at the time of FB. Protected specimen brush (PSB) was not used because of the unavailability of the instrument. Transbronchial biopsy (TBBx) was not performed owing to lack of fluoroscopy equipment and multiple comorbidities among the majority of SOT recipients. FB results were unremarkable in 32.2% (n ¼ 29). Other findings were increased secretions in 33 (36.7%33), chronic mucosal changes in 9 (10%), mucosal edema in 7 (7.8%), mucosal plaques in 7 (7.8%7), mucosal necrosis in 3 (3.3%),
3460
and endobronchial lesion in 2 (2.2%). Either BAL or BW was obtained in all patients. BW revealed positive cultures in 22 patients (24.4%), and BAL in 10 (11.2%), for an overall total of 32 patients (36%). In 7 patients, both BAL and BW cultures grew the same organisms (concordance). In the remainder of the patients, BAL and BW cultures were either negative or grew normal flora. Isolated microorganisms were as follows with different percentages: M tuberculosis 6.7% (n ¼ 6), S aureus 4.4% (n ¼ 4), M catharralis 4.4% (n ¼ 4), C albicans 6.7% (n ¼ 6), K pneumonia 2.2% (2), E coli 2.2% (n ¼ 2), S pneumonia 2.2% (n ¼ 2), S maltofilia 1.1% (1), A fumigatus 4.5% (n ¼ 4), and P aeruginosa 1.1% (n ¼ 1). This information led to successful alteration in medical management among these patients. In other words, new diagnosis was established by FB in 32 patients with a therapeutic impact of 36%. We observed no significant difference between the type of the transplant and the culture results (P > .05). PI was diagnosed in 82% of patients (n ¼ 73). Fifty-nine patients were renal transplant recipients. No significant difference was observed between the PI development and the age, gender, laboratory findings, or smoking status of the patient. PI development had no association with the type of the transplantation (either live or cadaver) and the rejection status (P > .05). No significant relationship was observed between the type of the transplantation and the patients with PI (P ¼ .52). Patients with PI who were treated either in hospital or in the intensive care unit when compared with the ones who did not have PI showed no difference (P ¼ .015). No association was observed between the PI development and comorbidities (P ¼ .08). Mortality was higher among patients with PI than the others (24.7% [n ¼ 22] vs 2.2% [n ¼ 2]). Mortality was also higher among patients treated in the intensive care unit (P ¼ .04). There were no differences between survival and PI development (P ¼ .15). DISCUSSION
PIs in SOTs constitute a medical emergency. Immediate, appropriate, empiric or specific antimicrobial treatment is mandatory. Because the spectrum of potential pathogens is far more diverse than in immunocompetent hosts, a systematic approach to the management of these patients is required. This approach should include a comprehensive diagnostic evaluation including history, physical examination, chest radiography, sputum studies, and blood cultures. A CT scan of the lung is usually indicated in patients in whom typical clinical or specific radiographic findings are missing. Bronchoscopy is indicated in patients with pulmonary infiltrates, unusual clinical and radiographic presentations, and those with treatment failure to empiric antibiotics.1 Among 998 transplant recipients, 90 underwent bronchoscopy to rule out and isolate pneumonia in our cohort. All were treated for pneumonia, even with negative culture growth. In the literature, respiratory tract infections were reported as 8.9% of all infectious episodes among this
EYÜBOGLU, KÜPELI, BOZBAS¸ ET AL
patient group; the incidence of pneumonia varied between 8% and 18%. We believe that the option for the prophylactic use of antibiotics may justify the differences observed in different studies regarding the incidence of respiratory tract infections. In our study, a positive history of smoking and acute rejection episodes in patients increased pneumonia risk. In addition, the presence of comorbidities has additional effects on the development of pneumonia. These factors have a potential role to delay the treatment period. Among 17 patients with rejection, 12 (13%) received pulse steroid therapy. The presence of cough, sputum, and fever was not significant among patients associated with PI in general or specific infections. It has been stated that if the bronchoscopy is performed before commencing empiric antibiotics, it provides greater diagnostic yield.5 Yet, this practice is not possible consistently owing to a variety of factors. Second, the bronchoscope should be introduced in the lower airways without using the suction to prevent contamination from the upper airways. BAL is the most frequently obtained specimen for the evaluation of lung infiltrates in immunocompromised patients. A comprehensive testing including stains and cultures for bacteria, mycobacterium, fungi, higher bacteria, and viruses should be carried out according to the endemicity of the organisms.5 If available, obtaining quantitative cultures on the BAL fluid or PSB specimen is also beneficial.5 Performing TBBx, if no contraindication exists, can add 10% to the diagnostic yield of the procedure.1 Tissue culture on the TBBx specimen if of limited value and is rarely performed.1 Occasionally, transbronchial needle aspiration may be helpful if mediastinal involvement is present.2 Therapeutic impact of bronchoscopy was significant; we were able to isolate agents of pneumonia in our patient group. PIs occur at predictable intervals in SOTs. The first month immediately after transplantation is influenced predominantly by nosocomial bacterial infections; over the following 6 months, opportunistic infections are high. After 6 months, community-acquired micro-organisms are responsible for PIs.1 Cases of M tuberculosis are mainly reported in areas of high endemicity such as Turkey. Nontuberculous mycobacterial pneumonias have been reported generally after the first year. Empiric antibiotic therapy should be targeted to the underlying suspected pathogens depending on several factors, including underlying immunosuppression and concurrent prophylactic therapy of SOT patients.6 It is well-established that empiric antibiotics do not cover causative organisms in >30% of SOTs. In a study by Ibrahim et al,7 42% of patients’ empiric antibiotherapy failed to cover concrete etiology. In our retrospective study, we determined that only 34% of our SOT recipient with PI empiric antibiotic treatment covered the causative organisms; among the remainder, further investigation similar to other studies was required. New diagnosis was established by FB in 32 (36%) of our SOT patients with a therapeutic impact. This was despite ongoing empiric antibiotic therapy. It is also reported that
PULMONARY INFECTIONS IN SOT
endobronchial findings can be helpful in supporting the suspected diagnosis of PI, especially in cases of mycobacterial or fungal infection. In one third of our patients, endobronchial findings were nonspecific. Nevertheless, these findings suggested that increased purulent secretions predominantly support PI. In our patient group, isolated microorganisms by FB were M tuberculosis (6.7%), S aureus (4.4%), M catharralis (4.4%), C albicans (6.7%), K pneumonia (2.2%), E coli (2.2%), S pneumonia (2.2%), S maltofilia (1.1%), A fumigatus (4.5%), and P aeruginosa (1.1%). The incidence of tuberculosis in transplant populations is reported to be 20 to 75 times greater than that of the general population based on endemicity. These results were despite of ongoing empiric antibiotic therapies. In areas with low endemicity, the prevalence of tuberculosis among SOT recipients is 0.5% to 6.4% in contrast with 15.2% in endemic countries.8 All of our M tuberculosis patients underwent SOT 1 to 6 years before the diagnosis. In addition, these patients were referred to us from cities with higher tuberculosis endemicity in our country. The remainder of PIs was detected at the predictable interval. We observed no difference between the type of the SOT and the culture results or type of PI. The value of FB in our SOT recipients was in line with previous studies with a therapeutic impact of 36%.9 Overall mortality in SOTs with PI was found to be between 21% and 35%; however, differences in mortality between nosocomial and community-acquired infection were extreme at 58% and 8%, respectively.9,10 We speculate that, given an empiric antibiotic therapy without waiting, culture results and using effective prophylactic antibiotic regimens might have decreased our mortality rate in patients with PI. In conclusion, suspected PI is the most common indication for FB in SOT recipients. Although endobronchial findings are nonspecific, FB may identify causative organism in >30% of patients despite ongoing empiric treatment. It can be assumed that yield could have been higher if FB was performed before ordering therapy. Bacteria were the most
3461
common pathogens versus fungal or viral micro-organisms, but tuberculosis should always kept in mind in areas with high endemicity. Our study has several weaknesses in addition to being retrospective in nature. The majority of our patients were mainly renal and liver transplant recipients, limiting its observations and applications. In addition, we were not able to perform PSB nor TBBx. Quantitative cultures were also not performed. We do, however, feel that our findings are applicable to the majority of the practices because these limitations are not unique to our institution. REFERENCES 1. Küpeli E, Eyüboglu FÖ, Haberal M. Pulmonary infections in transplant recipients. Curr Opin Pulm Med. 2012;18:202e212. 2. Baughman RP, Lower EE. Diagnosis of pneumonia in immunocompromised patient. In: Agusti C, Torres A, eds. Pulmonary Infection in the Immunocompromised Patient. New York: Wiley; 2009:53e80. 3. American Thoracic Society. Clinical role of bronchoalveolar lavage in adults with pulmonary disease. Am Rev Respir Dis. 1990;142:481e486. 4. Lee FYW, Mehta AC. Basic techniques in flexible bronchoscopy. In: Wang KP, Mehta AC, eds. Flexible Bronchoscopy. Cambridge, UK: Blackwell Science; 1995:95e118. 5. Jain P, Sandur S, Meli Y, et al. Role of flexible bronchoscopy in immunocompromised patients with lung infiltrates. Chest. 2004;125:712e722. 6. Kottloff RM, Ahya VN, Crawford SW. Pulmonary complications of solid organ transplantation and hematopoietic stem cell transplantation. Am J Respir Crit Care Med. 2004;170:367e416. 7. Ibrahim EH, Ward S, Sherman G, et al. Experience with a clinical guideline for the treatment of ventilator-associated pneumonia. Crit Care Med. 2001;29:1109e1115. 8. Arslan H, Ergin F, Öner Eyüboglu F, et al. Tuberculosis in renal transplant recipients. Transplant Proc. 2003;35:2670e2681. 9. Bonatti H, Pruet TL, Brandacher G, et al. Pneumonia in solid organ recipients: spectrum of pathogens in 217 episodes. Transplant Proc. 2009;41:371e374. 10. Cervera C, Agusti C, Angeles Marcos M, et al. Microbiologic features and outcome of pneumonia in transplanted patients. Diagn Microbiol Infect Dis. 2006;55:47e54.