- Email: [email protected]
The Findings of Surgical Lung Biopsy in HIV Patients in the Era of Highly Active Antiretroviral Therapy
GENERAL THORACIC
The Findings of Surgical Lung Biopsy in HIV Patients in the Era of Highly Active Antiretroviral Therapy Ashfaq Ahmad, MD, Shafeeq Khan, MD, and Ayman O. Soubani, MD Division of Pulmonary, Critical Care and Sleep Medicine, Wayne State University School of Medicine, Detroit, Michigan
Background. Surgical lung biopsy (SLB) is rarely necessary in the evaluation of pulmonary infiltrates in patients with human immunodeficiency virus (HIV). The aim of this study is to describe the findings of SLB in a recent cohort of HIV-infected patients with unexplained pulmonary infiltrates, and determine the impact of this procedure on management and outcome of these patients. Methods. We used a retrospective review of medical records for this study. Results. Twenty-six patients who underwent SLB where evaluated. The procedure was performed either by video-assisted thoracoscopy, in 42% of the cases, or by thoracotomy in 58%. Eleven patients (42%) had focal infiltrates, and 15 (58%) had diffuse infiltrates. There were 31 diagnoses in the 26 patients; 26 (84%) were specific and 5(16%) were nonspecific. The most common
specific diagnoses were infection (58%), malignancy (16%), and inflammatory conditions (10%). Pneumocystis carinii, alone or in combination with other conditions, was the most common diagnosis of all biopsies (34%). Surgical lung biopsy led to a change in management in 17 patients (65%), and was associated with complications in 4 patients (15%). The in-hospital mortality rate was 27%. Conclusions. Surgical lung biopsy is a relatively safe procedure, and provides a specific diagnosis in the majority of patients with HIV infection and unexplained pulmonary infiltrates. Pneumocystis carinii alone or in combination with other conditions is the most common finding.
F
evaluate the impact of the procedure on the management of these patients.
ibroptic bronchoscopy is the procedure of choice in evaluating pulmonary disease in patients infected with human immunodeficiency virus (HIV). The diagnostic yield of this procedure in detecting most of the pulmonary complications of HIV infection is very high, and in the case of Pneumocystis carinii it exceeds 90% [1, 2]. However, the role of fibroptic bronchoscopy is limited in diagnosing certain conditions such as Kaposi’s sarcoma, cytomegalovirus (CMV) pneumonitis, and lymphocytic interstitial pneumonitis [2]. Rarely do HIV patients require SLB to diagnose pulmonary infiltrates. There were limited studies in the last 2 decades to determine the value and findings of SLB in this patient population [3–7]. All these studies were conducted before the highly active antiretroviral therapy (HAART) era, and the SLB procedures were done by thoracotomy. There have been no recent studies that describe the findings of SLB in HIVinfected patients in the HAART era, especially in light of the recent advances in surgical techniques and postoperative care. The primary aim of this study is to describe the findings and safety of SLB in patients with HIV infection and unexplained pulmonary infiltrates, and to Accepted for publication Nov 17, 2004. Address reprint requests to Dr Soubani, Harper University Hospital, Division of Pulmonary, Critical Care and Sleep Medicine, 3990 John R-3 Hudson, Detroit, MI 48201; e-mail: [email protected].
© 2005 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2005;79:1862–5) © 2005 by The Society of Thoracic Surgeons
Patients and Methods We reviewed the medical records of all patients who had a SLB (either by thoracotomy or by video-assisted thoracoscopic [VATS] biopsy) for the diagnosis of pulmonary infiltrates at the Detroit Medical Center, which is an urban, tertiary, university-affiliated hospital, during the period between July 1997 and June 2003. Patients were included in the analysis if they had HIV infection and underwent a SLB for the diagnosis of pulmonary infiltrates. The data collected included demographic information, the time from HIV diagnosis to SLB, white cell count, CD4 count, viral load, HAART status, P carinii prophylaxis, computed tomography (CT) findings, prior diagnostic studies, histologic and microbiologic diagnosis by SLB, change in management, type of procedure and its complications, and final hospital outcome. Computed tomography findings were considered diffuse when there were bilateral pulmonary infiltrates involving all lobes of the lung; conversely, focal findings were defined as localized lesions such as masses, nodules, or focal infiltrates. Change in management was defined as addition or withdrawal of therapy including antimicrobials (antibac0003-4975/05/$30.00 doi:10.1016/j.athoracsur.2004.11.030
AHMAD ET AL SURGICAL LUNG BIOPSY IN HIV PATIENTS
Table 1. Clinical Characteristics of All HIV-Infected Patients Who Underwent SLB Characteristics Age, mean ⫾ SD, years (range) Male/female White/black/others The time from HIV diagnosis to SLB, mean ⫾ SD, years (range) CD4 count, mean ⫾ SD, cells/mm3 (range) Pneumocystis carinii prophylaxis HAART Mechanical ventilation Radiologic findings Focal Diffuse
All Patients (n ⫽ 26) 41.3 ⫾ 12.5 (26–74) 18/8 4/19/3 4.4 ⫾ 4.2 (0–18) 155.9 ⫾ 209.9 (1–680) 12 18 5 11 15
HAART ⫽ highly active antiretroviral therapy; HIV ⫽ human immunodeficiency virus; SLB ⫽ surgical lung biopsy.
terial, antiviral, antifungal or antimycobacterial), corticosteroids, or treatment for malignancy. The diagnosis was considered specific if the pathologic or microbiological findings were characteristic enough to establish a diagnosis and direct therapy, such as P carinii, CMV, mycobacteria, lung cancer, Kaposi’s sarcoma, hypersensitivity pneumonitis, lymphocytic interstitial pneumonitis, and bronchiolitis obliterans organizing pneumonia. The diagnosis was considered nonspecific when the findings were not characteristic enough to establish an etiology or direct therapy, such as diffuse alveolar damage, interstitial fibrosis, hyalinizing granuTable 2. Surgical Lung Biopsy Findings in All HIV-Infected Patientsa Diagnosis Specific Infectious Pneumocystis carinii Cytomegalovirus Atypical mycobacteria Mycobacterium tuberculosis Malignancy Lung cancer Kaposi’s sarcoma Inflammatory Bronchiolitis obliterans organizing pneumonia Lymphocytic interstitial pneumonitis Hypersensitivity pneumonitis Nonspecific Chronic organizing pneumonia Diffuse alveolar damage Interstitial fibrosis Hyalinizing granulomas a
Four patients had two or more diagnoses by SLB.
Number (n ⫽ 31) 26 (84%) 18 (58%) 9 5 3 1 5 (16%) 3 2 3 (10%) 1 1 1 5 (16%) 2 1 1 1
1863
lomas, or chronic organizing pneumonia with negative special stains and cultures. Surgical complications were defined as death directly related to the procedure, abnormal bleeding, return to operating room, prolonged chest tube insertion (⬎7 days), prolonged pneumomediastinum (⬎48 hours), wound infection, or failure of extubation after 48 hours of the procedure if the patient was not intubated before the procedure. The decision to proceed with SLB was made on a case-by-case basis by the treating physicians. The indications were slow response or lack of response to treatment of a condition diagnosed by bronchoscopy (8 patients), or a nondiagnostic bronchoscopy (11 patients), or the clinical judgment of the treating physicians that bronchoscopy would have a low diagnostic yield or be associated with high risk due to respiratory compromise or coagulopathy (7 patients). Thoracic surgeons performed surgical lung biopsies, either by VATS in 42% of the cases, or by thoracotomy in 58%. The surgeon individualized the choice of the surgical approach. There were no identified criteria for the type of SLB procedure. Multiple wedges of pulmonary tissue were taken using the stapling technique; intercostal tube drainage was routinely employed. Tissue samples were submitted for immediate frozen section as well as permanent sections with Gram, hematoxylin and eosin, and Gimsa stains. Aerobic and anaerobic, acid fast bacilli, fungal, and viral cultures were routinely performed. Data were analyzed using the Fisher exact test. All p values less than 0.05 were considered statistically significant.
Results Twenty-six patients aged 41.3 ⫾ 12.5 years (mean age ⫾ SD) underwent SLB and met the inclusion criteria. The baseline patients’ characteristics are shown in Table 1. A specific diagnosis was determined in 21 patients (81%), and a nonspecific diagnosis was made in 5 patients (19%). Four patients had two (3 patients) or three (1 patient) simultaneous specific pulmonary diagnoses on SLB, so there were 31 diagnoses in the 26 patients. Of the 31 pulmonary diagnoses, 26 (84%) were specific and 5 (16%) were nonspecific (Table 2). The 4 patients who had more than one diagnosis on SLB had P carinii and CMV (2 patients), P carinii, CMV and M avium complex (1 patient), and P carinii and hypersensitivity pneumonitis (1 patient). The patients were divided into three groups: the first group was patients who had positive results by bronchoscopy and were started on treatment. Their condition deteriorated, so a SLB was done (8 patients). All of these patients were diagnosed to have P carinii by bronchoscopy except 1 who had CMV. Surgical lung biopsy revealed P carinii in 3 patients, P carinii and an additional condition in 3 patients (CMV, hypersensitivity pneumonitis, CMV and M avium complex). One patient had CMV, and another patient had atypical mycobacteria
GENERAL THORACIC
Ann Thorac Surg 2005;79:1862–5
GENERAL THORACIC
1864
AHMAD ET AL SURGICAL LUNG BIOPSY IN HIV PATIENTS
(Mycobacterium xenopi). The SLB led to a change in management in 5 of these patients, and 6 patients died during their hospitalization. The second group was patients who had negative bronchoscopy (11 patients); the SLB findings were chronic organizing pneumonia with negative cultures (3 patients), P carinii (2 patients), and Mycobacterium tuberculosis, CMV, nonspecific interstitial fibrosis, bronchiolitis obliterans organizing pneumonia, lymphocytic interstitial pneumonitis, and Kaposi’s sarcoma (1 patient each). In this group, there was a change in management because of SLB in 8 patients, and only 1 patient died in the hospital. The third group was patients who did not undergo bronchoscopy before SLB (7 patients), and the diagnoses obtained by SLB were nonsmall cell carcinoma of the lung (3 patients—2 had adenocarcinoma and 1 had poorly differentiated nonsmall cell carcinoma), Kaposi’s sarcoma, both P carinii and CMV, M avium complex, and hyalinizing granulomas (each 1 patient). The procedure led to a change in management in 4 patients, and none of these patients died during their hospitalization. There were 5 patients on mechanical ventilation before SLB. All underwent bronchoscopy before SLB. In 3 patients, the bronchoscopy showed P carinii; however, the patients did not improve with therapy, and SLB findings were P carinii and hypersensitivity pneumonitis, P carinii, and CMV—1 patient each. The other 2 patients had a nondiagnostic bronchoscopy, and SLB results were lymphocytic interstitial pneumonitis, and diffuse alveolar damage. Three patients died during their hospitalizations owing to the underlying disease. Review of the CT scan findings revealed that 11 patients (42%) had a focal pattern and 15 (58%) showed diffuse findings. A specific diagnosis was found in 82% of the cases with a focal abnormality on chest CT scan, compared with 80% with a diffuse pattern (p ⫽ 0.38). Overall, the SLB led to a change in management in 17 patients (65%). Therapeutic changes were made in 15 of the 21 patients (71%) with specific diagnosis; on the other hand, 2 of 5 patients (40%) with nonspecific diagnoses had a change in therapy as a result of the SLB findings (p ⫽ 0.17). The overall in-hospital mortality for this cohort was 27% (7 patients). Those who died had P carinii (3 patients), CMV (2 patients), P carinii and CMV (1 patient), and CMV, P carinii, and M avium complex (1 patient). Complications related to the SLB procedure were reported in 4 patients (15%). None of the patients died because of the procedure. Two patients had wound infections, 1 patient had pneumomediastinum, and 1 developed excessive bleeding. Three of these complications happened in patients who underwent VATS and 1 developed in a patient who had thoracotomy.
Comment In the pre-HAART era, there were few studies describing open-lung biopsy findings in HIV patients with pulmonary infiltrates. Pass and colleagues [3], reported the findings of 23 open lung biopsies of 19 AIDS patients that
Ann Thorac Surg 2005;79:1862–5
were diagnostic in 70% of cases (most common findings were Kaposi’s sarcoma and P carinii). The procedure led to a change in therapy in 67% of the patients, and the hospital mortality was 32% (2 patients died owing to the procedure). In a report by Bonfils-Roberts and colleagues [4], the most common finding of open lung biopsy on 66 patients with AIDS was P carinii, either alone or with other pathogens (74%). The procedure led to a change in therapy in only 1 patient, and the mortality within 1 month of the procedure was 33% (3 during the operation). In the report by Trachiotis and colleagues [5], 60% of 25 patients with AIDS who underwent thoracotomy had a change in therapy, and the hospital mortality rate was 36%. The most common findings of the procedure were P carinii and Kaposi’s sarcoma. In 20% of the patients, the procedure was nondiagnostic. Finally, the report by LiCicero and colleagues [6], on 15 patients with HIV infection who underwent thoracotomy, indicated that the procedure led to a change in therapy in 73% of those patients and the hospital mortality rate was 20%. The most common findings of thoracotomy were Kaposi’s sarcoma and P carinii, and the most important predictor of mortality was mechanical ventilation. This study is similar to previous ones in the high yield of a specific diagnosis, and that P carinii, either alone or in combination with an additional diagnosis, is the most common finding by SLB. Furthermore, this study shows that in 26% of such patients, the etiology of the pulmonary infiltrates is due to underlying malignancy (such as Kaposi’s sarcoma or lung cancer), or is due to an inflammatory condition. On the other hand, in 16% of the patients the SLB yielded a nonspecific diagnosis. The etiology of the nonspecific findings is not clear but may be related to an infectious process that was either completely or partially treated, resulting in negative microbiologic examination, a complication of therapy, or an inflammatory response to recurrent infections. In nearly half of the patients in this study, the SLB was done by VATS. This approach was well tolerated by these patients, with a high diagnostic yield. The overall complications rate related to SLB was much lower than historical studies. There were no deaths related to the procedure. Therefore, in selected patients with HIV infection and unexplained pulmonary infiltrates, SLB by VATS is an acceptable procedure, with high diagnostic yield and low risk of complications. In 5 cases, SLB was done on patients who were mechanically ventilated. Two of these patients were discharged alive from the hospital. Previous reports suggested that SLB should be avoided in patients who are mechanically ventilated [6]. While the number of such patients is small, the study suggests that in selected HIV patients on mechanical ventilation, SLB is a diagnostic option, especially that the prognosis of these patients is improving [8]. The study shows that the in-hospital mortality rate among those patients who undergo SLB is lower than previously reported. Although this may represent a selection bias, or may be due to improved surgical tech-
niques, postoperative care, or more effective treatment of the conditions diagnosed by SLB. The decision to do SLB in the study group was made on a case-by-case basis by the treating physicians. The study could not identify patients who are likely to have a specific diagnosis or change in management because of the SLB, probably owing to the small number of patients and the retrospective nature of this study. The aim of this study was to determine the findings in HIV-infected patients with unexplained pulmonary infiltrates who underwent SLB. The retrospective nature of this analysis was limited in providing specific criteria for the best candidate for SLB, or the type of procedure (VATS versus thoracotomy). Furthermore, the study was not designed to provide data on all HIV-infected patients who presented with pulmonary infiltrates and did not undergo SLB, or to compare the findings and outcome between those who did and did not undergo SLB. Prospective, multicenter studies are needed to address these questions, and to identify predictors of outcome after SLB and provide guidelines on how to proceed in the evaluation of unexplained pulmonary infiltrates in this patient population. In summary, SLB is a useful diagnostic procedure in the evaluation of HIV-infected patients with unexplained pulmonary infiltrates. It has a high yield for specific diagnosis, and P carinii, either alone or in combination with other conditions, remains the most common finding.
AHMAD ET AL SURGICAL LUNG BIOPSY IN HIV PATIENTS
1865
The procedure is likely to lead to a change in management, and is relatively well tolerated in this patient population.
References 1. Golden JA, Hollander H, Stulbarg MS, Gamsu G. Bronchoalveolar lavage as the exclusive diagnostic modality for Pneumocystis carinii pneumonia. A prospective study among patients with acquired immunodeficiency syndrome. Chest 1986;90:18 –22. 2. Vander Els NJ, Stover DE. Approach to the patient with pulmonary disease. Clin Chest Med 1996;17:767– 85. 3. Pass HI, Potter D, Shelhammer J, et al. Indications for and diagnostic efficacy of open-lung biopsy in the patient with acquired immunodeficiency syndrome (AIDS). Ann Thorac Surg 1986;41:307–12. 4. Bonfils-Roberts EA, Nickodem A, Nealon TF Jr. Retrospective analysis of the efficacy of open lung biopsy in acquired immunodeficiency syndrome. Ann Thorac Surg 1990;49: 115–7. 5. Trachiotis GD, Hafner GH, Hix WR, Aaron BL. Role of open lung biopsy in diagnosing pulmonary complications of AIDS. Ann Thorac Surg 1992;54:898 –901. 6. LoCicero J III. Does every patient with enigmatic lung disease deserve a lung biopsy? The continuing dilemma. Chest 1994; 106:706 – 8. 7. Fitzgerald W, Bevelaqua FA, Garay SM, Aranda CP. The role of open lung biopsy in patients with the acquired immunodeficiency syndrome. Chest 1987;91:659 – 61. 8. Narasimhan M, Posner AJ, DePalo VA, et al. Intensive care in patients with HIV infection in the era of highly active antiretroviral therapy. Chest 2004;125:1800 – 4.
GENERAL THORACIC
Ann Thorac Surg 2005;79:1862–5
The Findings of Surgical Lung Biopsy in HIV Patients in the Era of Highly Active Antiretroviral Therapy Ashfaq Ahmad, MD, Shafeeq Khan, MD, and Ayman O. Soubani, MD Division of Pulmonary, Critical Care and Sleep Medicine, Wayne State University School of Medicine, Detroit, Michigan
Background. Surgical lung biopsy (SLB) is rarely necessary in the evaluation of pulmonary infiltrates in patients with human immunodeficiency virus (HIV). The aim of this study is to describe the findings of SLB in a recent cohort of HIV-infected patients with unexplained pulmonary infiltrates, and determine the impact of this procedure on management and outcome of these patients. Methods. We used a retrospective review of medical records for this study. Results. Twenty-six patients who underwent SLB where evaluated. The procedure was performed either by video-assisted thoracoscopy, in 42% of the cases, or by thoracotomy in 58%. Eleven patients (42%) had focal infiltrates, and 15 (58%) had diffuse infiltrates. There were 31 diagnoses in the 26 patients; 26 (84%) were specific and 5(16%) were nonspecific. The most common
specific diagnoses were infection (58%), malignancy (16%), and inflammatory conditions (10%). Pneumocystis carinii, alone or in combination with other conditions, was the most common diagnosis of all biopsies (34%). Surgical lung biopsy led to a change in management in 17 patients (65%), and was associated with complications in 4 patients (15%). The in-hospital mortality rate was 27%. Conclusions. Surgical lung biopsy is a relatively safe procedure, and provides a specific diagnosis in the majority of patients with HIV infection and unexplained pulmonary infiltrates. Pneumocystis carinii alone or in combination with other conditions is the most common finding.
F
evaluate the impact of the procedure on the management of these patients.
ibroptic bronchoscopy is the procedure of choice in evaluating pulmonary disease in patients infected with human immunodeficiency virus (HIV). The diagnostic yield of this procedure in detecting most of the pulmonary complications of HIV infection is very high, and in the case of Pneumocystis carinii it exceeds 90% [1, 2]. However, the role of fibroptic bronchoscopy is limited in diagnosing certain conditions such as Kaposi’s sarcoma, cytomegalovirus (CMV) pneumonitis, and lymphocytic interstitial pneumonitis [2]. Rarely do HIV patients require SLB to diagnose pulmonary infiltrates. There were limited studies in the last 2 decades to determine the value and findings of SLB in this patient population [3–7]. All these studies were conducted before the highly active antiretroviral therapy (HAART) era, and the SLB procedures were done by thoracotomy. There have been no recent studies that describe the findings of SLB in HIVinfected patients in the HAART era, especially in light of the recent advances in surgical techniques and postoperative care. The primary aim of this study is to describe the findings and safety of SLB in patients with HIV infection and unexplained pulmonary infiltrates, and to Accepted for publication Nov 17, 2004. Address reprint requests to Dr Soubani, Harper University Hospital, Division of Pulmonary, Critical Care and Sleep Medicine, 3990 John R-3 Hudson, Detroit, MI 48201; e-mail: [email protected].
© 2005 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2005;79:1862–5) © 2005 by The Society of Thoracic Surgeons
Patients and Methods We reviewed the medical records of all patients who had a SLB (either by thoracotomy or by video-assisted thoracoscopic [VATS] biopsy) for the diagnosis of pulmonary infiltrates at the Detroit Medical Center, which is an urban, tertiary, university-affiliated hospital, during the period between July 1997 and June 2003. Patients were included in the analysis if they had HIV infection and underwent a SLB for the diagnosis of pulmonary infiltrates. The data collected included demographic information, the time from HIV diagnosis to SLB, white cell count, CD4 count, viral load, HAART status, P carinii prophylaxis, computed tomography (CT) findings, prior diagnostic studies, histologic and microbiologic diagnosis by SLB, change in management, type of procedure and its complications, and final hospital outcome. Computed tomography findings were considered diffuse when there were bilateral pulmonary infiltrates involving all lobes of the lung; conversely, focal findings were defined as localized lesions such as masses, nodules, or focal infiltrates. Change in management was defined as addition or withdrawal of therapy including antimicrobials (antibac0003-4975/05/$30.00 doi:10.1016/j.athoracsur.2004.11.030
AHMAD ET AL SURGICAL LUNG BIOPSY IN HIV PATIENTS
Table 1. Clinical Characteristics of All HIV-Infected Patients Who Underwent SLB Characteristics Age, mean ⫾ SD, years (range) Male/female White/black/others The time from HIV diagnosis to SLB, mean ⫾ SD, years (range) CD4 count, mean ⫾ SD, cells/mm3 (range) Pneumocystis carinii prophylaxis HAART Mechanical ventilation Radiologic findings Focal Diffuse
All Patients (n ⫽ 26) 41.3 ⫾ 12.5 (26–74) 18/8 4/19/3 4.4 ⫾ 4.2 (0–18) 155.9 ⫾ 209.9 (1–680) 12 18 5 11 15
HAART ⫽ highly active antiretroviral therapy; HIV ⫽ human immunodeficiency virus; SLB ⫽ surgical lung biopsy.
terial, antiviral, antifungal or antimycobacterial), corticosteroids, or treatment for malignancy. The diagnosis was considered specific if the pathologic or microbiological findings were characteristic enough to establish a diagnosis and direct therapy, such as P carinii, CMV, mycobacteria, lung cancer, Kaposi’s sarcoma, hypersensitivity pneumonitis, lymphocytic interstitial pneumonitis, and bronchiolitis obliterans organizing pneumonia. The diagnosis was considered nonspecific when the findings were not characteristic enough to establish an etiology or direct therapy, such as diffuse alveolar damage, interstitial fibrosis, hyalinizing granuTable 2. Surgical Lung Biopsy Findings in All HIV-Infected Patientsa Diagnosis Specific Infectious Pneumocystis carinii Cytomegalovirus Atypical mycobacteria Mycobacterium tuberculosis Malignancy Lung cancer Kaposi’s sarcoma Inflammatory Bronchiolitis obliterans organizing pneumonia Lymphocytic interstitial pneumonitis Hypersensitivity pneumonitis Nonspecific Chronic organizing pneumonia Diffuse alveolar damage Interstitial fibrosis Hyalinizing granulomas a
Four patients had two or more diagnoses by SLB.
Number (n ⫽ 31) 26 (84%) 18 (58%) 9 5 3 1 5 (16%) 3 2 3 (10%) 1 1 1 5 (16%) 2 1 1 1
1863
lomas, or chronic organizing pneumonia with negative special stains and cultures. Surgical complications were defined as death directly related to the procedure, abnormal bleeding, return to operating room, prolonged chest tube insertion (⬎7 days), prolonged pneumomediastinum (⬎48 hours), wound infection, or failure of extubation after 48 hours of the procedure if the patient was not intubated before the procedure. The decision to proceed with SLB was made on a case-by-case basis by the treating physicians. The indications were slow response or lack of response to treatment of a condition diagnosed by bronchoscopy (8 patients), or a nondiagnostic bronchoscopy (11 patients), or the clinical judgment of the treating physicians that bronchoscopy would have a low diagnostic yield or be associated with high risk due to respiratory compromise or coagulopathy (7 patients). Thoracic surgeons performed surgical lung biopsies, either by VATS in 42% of the cases, or by thoracotomy in 58%. The surgeon individualized the choice of the surgical approach. There were no identified criteria for the type of SLB procedure. Multiple wedges of pulmonary tissue were taken using the stapling technique; intercostal tube drainage was routinely employed. Tissue samples were submitted for immediate frozen section as well as permanent sections with Gram, hematoxylin and eosin, and Gimsa stains. Aerobic and anaerobic, acid fast bacilli, fungal, and viral cultures were routinely performed. Data were analyzed using the Fisher exact test. All p values less than 0.05 were considered statistically significant.
Results Twenty-six patients aged 41.3 ⫾ 12.5 years (mean age ⫾ SD) underwent SLB and met the inclusion criteria. The baseline patients’ characteristics are shown in Table 1. A specific diagnosis was determined in 21 patients (81%), and a nonspecific diagnosis was made in 5 patients (19%). Four patients had two (3 patients) or three (1 patient) simultaneous specific pulmonary diagnoses on SLB, so there were 31 diagnoses in the 26 patients. Of the 31 pulmonary diagnoses, 26 (84%) were specific and 5 (16%) were nonspecific (Table 2). The 4 patients who had more than one diagnosis on SLB had P carinii and CMV (2 patients), P carinii, CMV and M avium complex (1 patient), and P carinii and hypersensitivity pneumonitis (1 patient). The patients were divided into three groups: the first group was patients who had positive results by bronchoscopy and were started on treatment. Their condition deteriorated, so a SLB was done (8 patients). All of these patients were diagnosed to have P carinii by bronchoscopy except 1 who had CMV. Surgical lung biopsy revealed P carinii in 3 patients, P carinii and an additional condition in 3 patients (CMV, hypersensitivity pneumonitis, CMV and M avium complex). One patient had CMV, and another patient had atypical mycobacteria
GENERAL THORACIC
Ann Thorac Surg 2005;79:1862–5
GENERAL THORACIC
1864
AHMAD ET AL SURGICAL LUNG BIOPSY IN HIV PATIENTS
(Mycobacterium xenopi). The SLB led to a change in management in 5 of these patients, and 6 patients died during their hospitalization. The second group was patients who had negative bronchoscopy (11 patients); the SLB findings were chronic organizing pneumonia with negative cultures (3 patients), P carinii (2 patients), and Mycobacterium tuberculosis, CMV, nonspecific interstitial fibrosis, bronchiolitis obliterans organizing pneumonia, lymphocytic interstitial pneumonitis, and Kaposi’s sarcoma (1 patient each). In this group, there was a change in management because of SLB in 8 patients, and only 1 patient died in the hospital. The third group was patients who did not undergo bronchoscopy before SLB (7 patients), and the diagnoses obtained by SLB were nonsmall cell carcinoma of the lung (3 patients—2 had adenocarcinoma and 1 had poorly differentiated nonsmall cell carcinoma), Kaposi’s sarcoma, both P carinii and CMV, M avium complex, and hyalinizing granulomas (each 1 patient). The procedure led to a change in management in 4 patients, and none of these patients died during their hospitalization. There were 5 patients on mechanical ventilation before SLB. All underwent bronchoscopy before SLB. In 3 patients, the bronchoscopy showed P carinii; however, the patients did not improve with therapy, and SLB findings were P carinii and hypersensitivity pneumonitis, P carinii, and CMV—1 patient each. The other 2 patients had a nondiagnostic bronchoscopy, and SLB results were lymphocytic interstitial pneumonitis, and diffuse alveolar damage. Three patients died during their hospitalizations owing to the underlying disease. Review of the CT scan findings revealed that 11 patients (42%) had a focal pattern and 15 (58%) showed diffuse findings. A specific diagnosis was found in 82% of the cases with a focal abnormality on chest CT scan, compared with 80% with a diffuse pattern (p ⫽ 0.38). Overall, the SLB led to a change in management in 17 patients (65%). Therapeutic changes were made in 15 of the 21 patients (71%) with specific diagnosis; on the other hand, 2 of 5 patients (40%) with nonspecific diagnoses had a change in therapy as a result of the SLB findings (p ⫽ 0.17). The overall in-hospital mortality for this cohort was 27% (7 patients). Those who died had P carinii (3 patients), CMV (2 patients), P carinii and CMV (1 patient), and CMV, P carinii, and M avium complex (1 patient). Complications related to the SLB procedure were reported in 4 patients (15%). None of the patients died because of the procedure. Two patients had wound infections, 1 patient had pneumomediastinum, and 1 developed excessive bleeding. Three of these complications happened in patients who underwent VATS and 1 developed in a patient who had thoracotomy.
Comment In the pre-HAART era, there were few studies describing open-lung biopsy findings in HIV patients with pulmonary infiltrates. Pass and colleagues [3], reported the findings of 23 open lung biopsies of 19 AIDS patients that
Ann Thorac Surg 2005;79:1862–5
were diagnostic in 70% of cases (most common findings were Kaposi’s sarcoma and P carinii). The procedure led to a change in therapy in 67% of the patients, and the hospital mortality was 32% (2 patients died owing to the procedure). In a report by Bonfils-Roberts and colleagues [4], the most common finding of open lung biopsy on 66 patients with AIDS was P carinii, either alone or with other pathogens (74%). The procedure led to a change in therapy in only 1 patient, and the mortality within 1 month of the procedure was 33% (3 during the operation). In the report by Trachiotis and colleagues [5], 60% of 25 patients with AIDS who underwent thoracotomy had a change in therapy, and the hospital mortality rate was 36%. The most common findings of the procedure were P carinii and Kaposi’s sarcoma. In 20% of the patients, the procedure was nondiagnostic. Finally, the report by LiCicero and colleagues [6], on 15 patients with HIV infection who underwent thoracotomy, indicated that the procedure led to a change in therapy in 73% of those patients and the hospital mortality rate was 20%. The most common findings of thoracotomy were Kaposi’s sarcoma and P carinii, and the most important predictor of mortality was mechanical ventilation. This study is similar to previous ones in the high yield of a specific diagnosis, and that P carinii, either alone or in combination with an additional diagnosis, is the most common finding by SLB. Furthermore, this study shows that in 26% of such patients, the etiology of the pulmonary infiltrates is due to underlying malignancy (such as Kaposi’s sarcoma or lung cancer), or is due to an inflammatory condition. On the other hand, in 16% of the patients the SLB yielded a nonspecific diagnosis. The etiology of the nonspecific findings is not clear but may be related to an infectious process that was either completely or partially treated, resulting in negative microbiologic examination, a complication of therapy, or an inflammatory response to recurrent infections. In nearly half of the patients in this study, the SLB was done by VATS. This approach was well tolerated by these patients, with a high diagnostic yield. The overall complications rate related to SLB was much lower than historical studies. There were no deaths related to the procedure. Therefore, in selected patients with HIV infection and unexplained pulmonary infiltrates, SLB by VATS is an acceptable procedure, with high diagnostic yield and low risk of complications. In 5 cases, SLB was done on patients who were mechanically ventilated. Two of these patients were discharged alive from the hospital. Previous reports suggested that SLB should be avoided in patients who are mechanically ventilated [6]. While the number of such patients is small, the study suggests that in selected HIV patients on mechanical ventilation, SLB is a diagnostic option, especially that the prognosis of these patients is improving [8]. The study shows that the in-hospital mortality rate among those patients who undergo SLB is lower than previously reported. Although this may represent a selection bias, or may be due to improved surgical tech-
niques, postoperative care, or more effective treatment of the conditions diagnosed by SLB. The decision to do SLB in the study group was made on a case-by-case basis by the treating physicians. The study could not identify patients who are likely to have a specific diagnosis or change in management because of the SLB, probably owing to the small number of patients and the retrospective nature of this study. The aim of this study was to determine the findings in HIV-infected patients with unexplained pulmonary infiltrates who underwent SLB. The retrospective nature of this analysis was limited in providing specific criteria for the best candidate for SLB, or the type of procedure (VATS versus thoracotomy). Furthermore, the study was not designed to provide data on all HIV-infected patients who presented with pulmonary infiltrates and did not undergo SLB, or to compare the findings and outcome between those who did and did not undergo SLB. Prospective, multicenter studies are needed to address these questions, and to identify predictors of outcome after SLB and provide guidelines on how to proceed in the evaluation of unexplained pulmonary infiltrates in this patient population. In summary, SLB is a useful diagnostic procedure in the evaluation of HIV-infected patients with unexplained pulmonary infiltrates. It has a high yield for specific diagnosis, and P carinii, either alone or in combination with other conditions, remains the most common finding.
AHMAD ET AL SURGICAL LUNG BIOPSY IN HIV PATIENTS
1865
The procedure is likely to lead to a change in management, and is relatively well tolerated in this patient population.
References 1. Golden JA, Hollander H, Stulbarg MS, Gamsu G. Bronchoalveolar lavage as the exclusive diagnostic modality for Pneumocystis carinii pneumonia. A prospective study among patients with acquired immunodeficiency syndrome. Chest 1986;90:18 –22. 2. Vander Els NJ, Stover DE. Approach to the patient with pulmonary disease. Clin Chest Med 1996;17:767– 85. 3. Pass HI, Potter D, Shelhammer J, et al. Indications for and diagnostic efficacy of open-lung biopsy in the patient with acquired immunodeficiency syndrome (AIDS). Ann Thorac Surg 1986;41:307–12. 4. Bonfils-Roberts EA, Nickodem A, Nealon TF Jr. Retrospective analysis of the efficacy of open lung biopsy in acquired immunodeficiency syndrome. Ann Thorac Surg 1990;49: 115–7. 5. Trachiotis GD, Hafner GH, Hix WR, Aaron BL. Role of open lung biopsy in diagnosing pulmonary complications of AIDS. Ann Thorac Surg 1992;54:898 –901. 6. LoCicero J III. Does every patient with enigmatic lung disease deserve a lung biopsy? The continuing dilemma. Chest 1994; 106:706 – 8. 7. Fitzgerald W, Bevelaqua FA, Garay SM, Aranda CP. The role of open lung biopsy in patients with the acquired immunodeficiency syndrome. Chest 1987;91:659 – 61. 8. Narasimhan M, Posner AJ, DePalo VA, et al. Intensive care in patients with HIV infection in the era of highly active antiretroviral therapy. Chest 2004;125:1800 – 4.
GENERAL THORACIC
Ann Thorac Surg 2005;79:1862–5