- Email: [email protected]
Utility of bronchoalveolar lavage in the assessment of diffuse pulmonary infiltrates in nonAIDS immunocompromized patients
Resph'atory Medicine (1990) 84, 313-316
Utility of bronchoalveolar lavage in the assessment of diffuse pulmonary infiltrates in nonAIDS immunocompromized patients RAPHAEL BREUER, IZIDORES. L0SS0S, JOEL S. LAFAIRAND DAN ENGELHARD*
Pulmonary Unit and Department of Clinical Microbiology*, Hadassah University Hospital, P.O. Box 12000, Jerusalem 91120, Israel
The efficacy of bronchoscopy with bronchoalveolar lavage for diagnosing 32 episodes of diffuse pulmonary infiltrates was studied in 30 nonAIDS immunocompromized patients. Bronchoalveolar lavage had an overall diagnostic yield of 84% (27 of 32 episodes). Bronchoalveolar lavage was noncontributive in five episodes of pneumonitis: drug induced in one, nonspecific in three and pneumonitis of unestablished etiology in one. Overall, the procedure is safe and did not miss diagnoses for which conventional treatment was available. Our data support the use of bronchoscopy with bronchoalveolar lavage as the primary diagnostic procedure in immunocompromized patients with diffuse pulmonary infiltrates.
Introduction
Methods
Pulmonary diseases are a major cause of morbidity and mortality in immunocompromized patients (1,2). Diffuse pulmonary infiltrates in these patients may be due to various infectious agents, neoplastic disorders, drugs, radiation toxicity, hemorrhage, pulmonary edema and nonspecific pneumonitis. The diversity of etiologies and the nonspecific nature of the clinical and radiologic findings frequently require invasive procedures to establish a diagnosis. The availability of effective treatment for many infectious agents, evidence of decreased mortality with early institution of appropriate treatment (3) and prevention of potentially serious side effects of broad spectrum empiric therapy emphasize the importance of rapid accurate diagnosis. Bronchoalveolar lavage (BAL) performed by fibreoptic bronchoscopy has been found to be a useful diagnostic technique in patients with the acquired immune deficiency syndrome (AIDS) (4). The purpose of the present study was to assess the value of BAL in nonAIDS immunocompromized patients with diffuse bilateral pulmonary infiltrates of unknown etiology. Our data show that BAL is a safe procedure with a high diagnostic yield.
PATIENTS
Received 22 November 1989 and accepted 4 April 1990. Address for reprints and correspondence: Dr R. Breuer, Pulmonary Unit, Hadassah University Hospital, P.O. Box 1200, Jerusalem 91120, Israel, Tel 972-2-447548, Fax 972-2-427-427.
0954-6111/90]040313+ 04 $03.00/0
During a 2.5 year period a total of 32 bronchoscopic procedures with BAL were performed in 30 nonAIDS irnmunocompromized patients for the evaluation of diffuse bilateral pulmonary infiltrates. Patients were considered immunocompromized if they had the diagnoses of lymphoma (9), leukemia (3), cancer (2) (one with breast carcinoma and one with nasopharyngeal carcinoma), or following bone marrow transplantation (13) and kidney transplantation (3). Mean age was 35.1 years (range 1-71). Sixteen were male and 14 female. Routine prebronchoscopy studies included sputum culture when available and at least two blood cultures, all of which were negative. BRONCHOSCOPYTECHNIQUE Bronchoscopy was done if patients could achieve an arterial Pa02 of at least 70 mmHg and in mechanically ventilated patients only if positive end expiratory pressure did not exceed 10 cm H20. The procedure was done with fibreoptic bronchoscopes; Olympus P10 (4.Smm diameter) in adults, and 3C 10 (3.5 mm diameter) in children. Following a 6 h fast, patients were premedicated with a mild sedative and atropin. Lidocaine hydrochloride was applied topically to the pharynx and upper airways before bronchoscopic examination and used as needed in the bronchial tree duing the procedure. The bronchoscope was passed transnasally into the trachea. In mechanically 9 1990Bailli~reTindall
314
R. Breuer et al.
ventilated patients, the bronchoscope was passed directly into the trachea through an endotracheal tube. Initially, the entire tracheobronchial tree was inspected. After inspection, the bronchoscope was wedged in the bronchial segment which showed the worst radiographic appearance. When pulmonary infiltrates were homogenous, BAL was performed in both the right middle lobe and the left lingula. A volume of 100-120ml o f normal saline was lavaged in 20-30 ml aliquots, aspirated separately by vacuum suctioning, and pooled. If the fluid returned was less than 60% of the total volume instilled, the lavage was discontinued and performed in another location. Transbronchial or open lung biopsies were performed when BAL was noncontributive or when the clinical course mandated additional diagnostic studies. Decisions regarding which additional procedures would be performed in individual patients were based on the suspected diagnosis and clinical condition at the time of bronchoscopy. SPECIMENSSTUDIES A sample o f pooled lavage fluid was centrifugated for 5 min at 3000 rpm and smears were prepared from the deposit. These were stained by the Papanicolaou, silver methanamine, Ziehl Neelson and potassium hydroxide (KOH) methods. Following centrifugation, samples were cultured for fungi, mycobacteria and cytomegalovirus (CMV). Samples of pooled lavage fluid were directly stained with Gram Weigert method and cultured for bacteria including legionella. Transbronchial and open lung biopsy specimens were embedded in paraffin, cut and stained with eosin hematoxilin, silver methanamine, Gram Weigert and Ziehl Neelson stains. DIAGNOSTIC CRITERIA
Pneumocystis carinii pneumonia was diagnosed by the demonstration of cysts in BAL fluid or lung tissue. CMV was diagnosed when the typical cytopathic changes were seen in pulmonary macrophages of BAL fluid or lung tissue or when the virus was cultured from BAL of patients in whom interstitial pneumonitis was present in the biopsy specimen. A diagnosis of fungal pneumonia was made when in addition to BAL there was tissue histologic demonstration of pulmonary invasive disease. Bacterial pneumonia was diagnosed by Gram Werget stain and isolation of a single organism in culture of BAL fluid. Diagnosis of nonspecific pneumonitis were made when lung tissue showed interstitial fibrosis, mononuclear cell infiltrate or both and all stains and cultures for organisms were negative. Drug toxicity was diagnosed when the patient had a
clinical history compatible with drug toxicity and pathologic examination revealed the presence of nonspecific pneumonitis with bronchial and alveolar lining atypia. Cure from pneumonitis occurred if there were improvements in both arterial blood gases and chest roentgenogram, and death, if it occurred, was not related to pneumonitis. Statistical analysis of variables was performed by the Chi-square test.
Results
Thirty-two flexible fibreoptic bronchoscopies with BAL were carried out in 30 immunocompromized patients. In no patient were complications attributed to the BAL procedure. Specific etiologic diagnoses were obtained by 27 (84%) BAL procedures (Table 1). A single organism was found in 22 and a mixed infection with two organisms were found in five. Invasive pulmonary aspergillosis was suspected, in one patient, by BAL findings of characteristic hyphae in KOH stain and culture. This diagnosis was confirmed by compatible findings in open lung biopsy specimen. BAL was non-contributive in five episodes of pneumonitis. Four of these were diagnosed by transbronchial or open lung biopsies: drug induced pneumonitis in one and nonspecific pneumonitis in three. No diagnosis was established in one episode of pneumonitis in a patient who refused other diagnostic procedures and died despite broad spectrum antibiotic coverage against bacteria, fungi and Pneumocystis carinii. Transbronchial biopsy was performed in eight patients; five with contributive and three with noncontributive BAL. The overall yield of transbronchial biopsy was 25%. In none of the patients with contributive BAL was transbronchial biopsy positive. In two of three patients with n9/ar BAL, the transbronchial biopsy, was diagnostic: drug induced pneumonitis in one and nonspecific pneumonitis in another. Open lung biopsy was performed in seven patients. It confirmed the BAL diagnosis in five patients. In two patients, with noncontributive BAL, the open biopsy revealed nonspecific pneumonitis. Mortality in bone marrow transplant patients was higher (69%) compared to the other patients (16%) (P<0.01). Cure rate in patients with contributive BAL was 18/27 (67%) and 2/5 (40%) in patients with non-contributive BAL. Six of nine patients who had contributive BAL and died underwent open lung or autopsy studies. None revealed additional specific etiology for the pneumonitis.
Bronchoalveolar lavage in n o n A I D S hnmunocompromized patients
315
Table 1 Specific diagnoses by BAL in 32 episodes of diffuse bilateral pneumtmitis in nonAIDS immunocompromized patients Outcome Specificdiagnosis
Episodes diagnosed byBAL
'
Cure
Death
9 9 3 1
8 3 2 0
I 6 I 1
3 1 1 27/32 (84%)
3
0
1
0
I 18
0 9
Single infection: Pneumocystis carinii Cytomegalovirus Bacteria ~ Aspergillosis Combh~ed infections: Pneumocystis carinii and cytomegalovirus Pneumocystis earinii and bacteria b Cytomegalovirus and bacteria~ Total
-
'One case each of Hemophilus hzfluenzae, Staphylococcus aureus and Klebsiellapneumoniae bPseudomonas aeruginosa "Klebsiellapneumoniae.
Discussion
Diffuse bilateral pneumonitis is a frequent cause of mobidity and mortality in immunocompromized patients (1,2). Early therapy necessitates the application of invasive diagnostic techniques. Open lung biopsy offers sufficient lung tissue for thorough histologic examination and has a diagnostic yield of about 70% (5). It requires general anesthesia and may be the cause of morbidity in up to 30% (6) and mortality in up to 4% (7). Cutting needle biopsy is associated with an unacceptably high complication rate and has a diagnostic yield of 46-82% (8,9). Fine transthoracic needle aspiration has a low diagnostic sensitivity (10) and was complicated by pneumothorax in a number of patients (ll). Fibreoptic bronchoseopy was not the procedure of choice until the introduction of BAL. Studies in AIDS patients with pulmonary infiltrates showed BAL to be a safe procedure with a diagnostic yield as high as 86% (12). The present study demonstrates the use of BAL in nonAIDS immunocompromized patients, showing a diagnostic yield of 84%. Our results are higher than that reported by others (13,14). The difference may be attributed to the high prevalence of infectious etiologies for which BAL has a high diagnostic yield (13,15,16). BAL is less effective in diagnosis of pneumonitis caused by drug reaction, radiation injury or malignant infiltration of the lungs (13). Indeed, in our patients BAL was not diagnostic in four cases with drug or nonspecific interstitial pneumonitis, where the diagnosis was established by examination of lung tissue obtained by transbronchial or open lung biopsies.
The most common findings were opportunistic infections due to Pneumocystis carinii and CMV. The former was the most common cause ofpneumonitis in patients with hematologic malignancies and the latter in patients undergoing bone marrow transplantation. These findings are in agreement with previously reported studies (3,7,14,17). Overall, in our group of patients no infectious diagnosis for which conventional treatment is available were missed by BAL. Our study group, however, did not include a large group of neutropenic patients who are more prone to invasive fungi and bacteria which may be more difficult to diagnose by BAL (18,19). It seems that a negative BAL in non-neutropenic immunocompromized patients is likely to exclude an infectious etiology. BAL should be the first step in the diagnostic work-up of diffuse pulmonary infiltrates. Open lung biopsy should be persued as the second step ifBAL is noncontributive, in a clinical set up with high suspicion of fungi or drug related pneumonitis or if the clinical course is not compatible with the diagnosis derived by the BAL studies. In conclusion, BAL is a safe procedure with a high diagnostic yield for infectious etiologies in nonAIDS immunocompromized patients with diffuse bilateral pneumonitis. Bronchoscopy with BAL should be the primary diagnostic approach in these patients.
References
l. Bode FR., Pare .IA, Fraser RG. Pulmonary disease in the compromised host. Medicine 1974; 53: 255-293.
316
R . B r e u e r et al.
2. Pennington JF, Fedman NT. Pulmonary infiltrates and fever in patients with hematologic malignancy: assessment of transbronchial biopsy. A m J Med 1977; 62: 581-587. 3. Aisner J, Schimpff SC, Wiemik PH. Treatment of invasive aspergillosis. Relation to early diagnosis and treatment to response. Ann h~tern Med 1977; 86: 539-543. 4. Stover DE, White DA, Romano PA, Gellene RA. Diagnosis of pulmonary disease in acquired immune deficiency syndrome (AIDS). Am Rev Respir Dis 1984; 130: 659-662. 5. Mathay RA, Moritz ED. Invasive procedures for diagnosing pulmonary infection: a critical review. Clin Chest Med 1981; 2: 2-18. 6. Cheson BD, Samlowski WE, Tang TT, Spruance SL. Value of open-lung biopsy in 87 immunocompromised patients with pulmonary infiltrates. Cancer 1985; 55: 453-459. 7. Jaffe JP, Maki DG. Lung biopsy in immunocompromised patients: one institution's experience and approach to management of pulmonary disease in the compromised host. Cancer 1981; 48:1144-1153. 8. Cunningham JH, Zavala DC, Corry RJ, Keim LW. Trephine air drill, bronchial brush and fiberoptic transbronchial lung biopsies in immunosuppressed patients. Am RevRespirDis 1977; 115: 213-220. 9. Greenman RL, Goodall PT, King D. Lung biopsy in immunocompromisedhost. Am J M e d 1975;59: 488- 496. 10. Springmeyer SC, Hackman RC, Holle R, et al. Use of bronchoalveolar lavage to diagnose acute diffuse pneumonia in the immunocompromised host. J hlfect Dis 1986; 154: 604-610.
11. Palmer DL, Davidson M, Lusk R. Needle aspiration of the lung in complex pneumonias. Chest 1980;78: 16-21. 12. Broaddus C, Dake MD, Stulbarg MS et al. Bronchoalveolar lavage and transbronchial biopsy for the diagnosis of pulmonary infections in acquired immunodificiency syndrome. Ann Intern Med 1985; 102: 747-752. 13. Stover DE, Zaman MB, Hajdu SI, Lange M, Gold J, Armstrong D. Bronchoalveolar lavage in the diagnosis of diffuse pulmonary infiltrates in the immunosuppressed host. Ann lntern Med 1984; 101: I-7. 14. Cardonnier C, Bernaudin JF, Fleury J, et al. Diagnostic yieldof bronchoalveolar Iavagein pneumonitis occurring after allogeneic bone marrow transplantation. Am Rev Respir Dis 1985; 132:1118-1123. 15. Martin II WJ, Smith TF, Sanderson DR, Brutinel WM, Cockerill III FR, Douglas WW. Role ofbronchoalveolar lavage in the assessment of opportunistic pulmonary infections: utility and complications. Mayo Clin Proe 1987; 62: 549-557. 16. Young JA, Hopkin JM, Cuthbertson WP. Pulmonary infiltrates in immunocompromised patients: diagnosis by cytological examination of bronchoalvealar lavage fluid. J Clin Patho11984; 37: 390-397. 17. Crawford SW, Hackman RC, Clark JG. Open lungbiopsy diagnosis of diffuse pulmonary infiltrates after marrow transplantation. Chest 1988; 94: 949-953. 18. Kahn FW, Jones JM, England DM. The role of bronchoalveolar lavage in the diagnosis of invasive pulmonary aspergillosis. Am J Clin Patho11986; 86:518-522 19. Thorpe JE, Baughman RP, Frame PT, Wesseler TA, Staneck JL. Bronchoalveolar Iavagefor diagnosingacute bacterial pneumonia. J h f e c Dis 1987; 155:855 861.
Utility of bronchoalveolar lavage in the assessment of diffuse pulmonary infiltrates in nonAIDS immunocompromized patients RAPHAEL BREUER, IZIDORES. L0SS0S, JOEL S. LAFAIRAND DAN ENGELHARD*
Pulmonary Unit and Department of Clinical Microbiology*, Hadassah University Hospital, P.O. Box 12000, Jerusalem 91120, Israel
The efficacy of bronchoscopy with bronchoalveolar lavage for diagnosing 32 episodes of diffuse pulmonary infiltrates was studied in 30 nonAIDS immunocompromized patients. Bronchoalveolar lavage had an overall diagnostic yield of 84% (27 of 32 episodes). Bronchoalveolar lavage was noncontributive in five episodes of pneumonitis: drug induced in one, nonspecific in three and pneumonitis of unestablished etiology in one. Overall, the procedure is safe and did not miss diagnoses for which conventional treatment was available. Our data support the use of bronchoscopy with bronchoalveolar lavage as the primary diagnostic procedure in immunocompromized patients with diffuse pulmonary infiltrates.
Introduction
Methods
Pulmonary diseases are a major cause of morbidity and mortality in immunocompromized patients (1,2). Diffuse pulmonary infiltrates in these patients may be due to various infectious agents, neoplastic disorders, drugs, radiation toxicity, hemorrhage, pulmonary edema and nonspecific pneumonitis. The diversity of etiologies and the nonspecific nature of the clinical and radiologic findings frequently require invasive procedures to establish a diagnosis. The availability of effective treatment for many infectious agents, evidence of decreased mortality with early institution of appropriate treatment (3) and prevention of potentially serious side effects of broad spectrum empiric therapy emphasize the importance of rapid accurate diagnosis. Bronchoalveolar lavage (BAL) performed by fibreoptic bronchoscopy has been found to be a useful diagnostic technique in patients with the acquired immune deficiency syndrome (AIDS) (4). The purpose of the present study was to assess the value of BAL in nonAIDS immunocompromized patients with diffuse bilateral pulmonary infiltrates of unknown etiology. Our data show that BAL is a safe procedure with a high diagnostic yield.
PATIENTS
Received 22 November 1989 and accepted 4 April 1990. Address for reprints and correspondence: Dr R. Breuer, Pulmonary Unit, Hadassah University Hospital, P.O. Box 1200, Jerusalem 91120, Israel, Tel 972-2-447548, Fax 972-2-427-427.
0954-6111/90]040313+ 04 $03.00/0
During a 2.5 year period a total of 32 bronchoscopic procedures with BAL were performed in 30 nonAIDS irnmunocompromized patients for the evaluation of diffuse bilateral pulmonary infiltrates. Patients were considered immunocompromized if they had the diagnoses of lymphoma (9), leukemia (3), cancer (2) (one with breast carcinoma and one with nasopharyngeal carcinoma), or following bone marrow transplantation (13) and kidney transplantation (3). Mean age was 35.1 years (range 1-71). Sixteen were male and 14 female. Routine prebronchoscopy studies included sputum culture when available and at least two blood cultures, all of which were negative. BRONCHOSCOPYTECHNIQUE Bronchoscopy was done if patients could achieve an arterial Pa02 of at least 70 mmHg and in mechanically ventilated patients only if positive end expiratory pressure did not exceed 10 cm H20. The procedure was done with fibreoptic bronchoscopes; Olympus P10 (4.Smm diameter) in adults, and 3C 10 (3.5 mm diameter) in children. Following a 6 h fast, patients were premedicated with a mild sedative and atropin. Lidocaine hydrochloride was applied topically to the pharynx and upper airways before bronchoscopic examination and used as needed in the bronchial tree duing the procedure. The bronchoscope was passed transnasally into the trachea. In mechanically 9 1990Bailli~reTindall
314
R. Breuer et al.
ventilated patients, the bronchoscope was passed directly into the trachea through an endotracheal tube. Initially, the entire tracheobronchial tree was inspected. After inspection, the bronchoscope was wedged in the bronchial segment which showed the worst radiographic appearance. When pulmonary infiltrates were homogenous, BAL was performed in both the right middle lobe and the left lingula. A volume of 100-120ml o f normal saline was lavaged in 20-30 ml aliquots, aspirated separately by vacuum suctioning, and pooled. If the fluid returned was less than 60% of the total volume instilled, the lavage was discontinued and performed in another location. Transbronchial or open lung biopsies were performed when BAL was noncontributive or when the clinical course mandated additional diagnostic studies. Decisions regarding which additional procedures would be performed in individual patients were based on the suspected diagnosis and clinical condition at the time of bronchoscopy. SPECIMENSSTUDIES A sample o f pooled lavage fluid was centrifugated for 5 min at 3000 rpm and smears were prepared from the deposit. These were stained by the Papanicolaou, silver methanamine, Ziehl Neelson and potassium hydroxide (KOH) methods. Following centrifugation, samples were cultured for fungi, mycobacteria and cytomegalovirus (CMV). Samples of pooled lavage fluid were directly stained with Gram Weigert method and cultured for bacteria including legionella. Transbronchial and open lung biopsy specimens were embedded in paraffin, cut and stained with eosin hematoxilin, silver methanamine, Gram Weigert and Ziehl Neelson stains. DIAGNOSTIC CRITERIA
Pneumocystis carinii pneumonia was diagnosed by the demonstration of cysts in BAL fluid or lung tissue. CMV was diagnosed when the typical cytopathic changes were seen in pulmonary macrophages of BAL fluid or lung tissue or when the virus was cultured from BAL of patients in whom interstitial pneumonitis was present in the biopsy specimen. A diagnosis of fungal pneumonia was made when in addition to BAL there was tissue histologic demonstration of pulmonary invasive disease. Bacterial pneumonia was diagnosed by Gram Werget stain and isolation of a single organism in culture of BAL fluid. Diagnosis of nonspecific pneumonitis were made when lung tissue showed interstitial fibrosis, mononuclear cell infiltrate or both and all stains and cultures for organisms were negative. Drug toxicity was diagnosed when the patient had a
clinical history compatible with drug toxicity and pathologic examination revealed the presence of nonspecific pneumonitis with bronchial and alveolar lining atypia. Cure from pneumonitis occurred if there were improvements in both arterial blood gases and chest roentgenogram, and death, if it occurred, was not related to pneumonitis. Statistical analysis of variables was performed by the Chi-square test.
Results
Thirty-two flexible fibreoptic bronchoscopies with BAL were carried out in 30 immunocompromized patients. In no patient were complications attributed to the BAL procedure. Specific etiologic diagnoses were obtained by 27 (84%) BAL procedures (Table 1). A single organism was found in 22 and a mixed infection with two organisms were found in five. Invasive pulmonary aspergillosis was suspected, in one patient, by BAL findings of characteristic hyphae in KOH stain and culture. This diagnosis was confirmed by compatible findings in open lung biopsy specimen. BAL was non-contributive in five episodes of pneumonitis. Four of these were diagnosed by transbronchial or open lung biopsies: drug induced pneumonitis in one and nonspecific pneumonitis in three. No diagnosis was established in one episode of pneumonitis in a patient who refused other diagnostic procedures and died despite broad spectrum antibiotic coverage against bacteria, fungi and Pneumocystis carinii. Transbronchial biopsy was performed in eight patients; five with contributive and three with noncontributive BAL. The overall yield of transbronchial biopsy was 25%. In none of the patients with contributive BAL was transbronchial biopsy positive. In two of three patients with n9/ar BAL, the transbronchial biopsy, was diagnostic: drug induced pneumonitis in one and nonspecific pneumonitis in another. Open lung biopsy was performed in seven patients. It confirmed the BAL diagnosis in five patients. In two patients, with noncontributive BAL, the open biopsy revealed nonspecific pneumonitis. Mortality in bone marrow transplant patients was higher (69%) compared to the other patients (16%) (P<0.01). Cure rate in patients with contributive BAL was 18/27 (67%) and 2/5 (40%) in patients with non-contributive BAL. Six of nine patients who had contributive BAL and died underwent open lung or autopsy studies. None revealed additional specific etiology for the pneumonitis.
Bronchoalveolar lavage in n o n A I D S hnmunocompromized patients
315
Table 1 Specific diagnoses by BAL in 32 episodes of diffuse bilateral pneumtmitis in nonAIDS immunocompromized patients Outcome Specificdiagnosis
Episodes diagnosed byBAL
'
Cure
Death
9 9 3 1
8 3 2 0
I 6 I 1
3 1 1 27/32 (84%)
3
0
1
0
I 18
0 9
Single infection: Pneumocystis carinii Cytomegalovirus Bacteria ~ Aspergillosis Combh~ed infections: Pneumocystis carinii and cytomegalovirus Pneumocystis earinii and bacteria b Cytomegalovirus and bacteria~ Total
-
'One case each of Hemophilus hzfluenzae, Staphylococcus aureus and Klebsiellapneumoniae bPseudomonas aeruginosa "Klebsiellapneumoniae.
Discussion
Diffuse bilateral pneumonitis is a frequent cause of mobidity and mortality in immunocompromized patients (1,2). Early therapy necessitates the application of invasive diagnostic techniques. Open lung biopsy offers sufficient lung tissue for thorough histologic examination and has a diagnostic yield of about 70% (5). It requires general anesthesia and may be the cause of morbidity in up to 30% (6) and mortality in up to 4% (7). Cutting needle biopsy is associated with an unacceptably high complication rate and has a diagnostic yield of 46-82% (8,9). Fine transthoracic needle aspiration has a low diagnostic sensitivity (10) and was complicated by pneumothorax in a number of patients (ll). Fibreoptic bronchoseopy was not the procedure of choice until the introduction of BAL. Studies in AIDS patients with pulmonary infiltrates showed BAL to be a safe procedure with a diagnostic yield as high as 86% (12). The present study demonstrates the use of BAL in nonAIDS immunocompromized patients, showing a diagnostic yield of 84%. Our results are higher than that reported by others (13,14). The difference may be attributed to the high prevalence of infectious etiologies for which BAL has a high diagnostic yield (13,15,16). BAL is less effective in diagnosis of pneumonitis caused by drug reaction, radiation injury or malignant infiltration of the lungs (13). Indeed, in our patients BAL was not diagnostic in four cases with drug or nonspecific interstitial pneumonitis, where the diagnosis was established by examination of lung tissue obtained by transbronchial or open lung biopsies.
The most common findings were opportunistic infections due to Pneumocystis carinii and CMV. The former was the most common cause ofpneumonitis in patients with hematologic malignancies and the latter in patients undergoing bone marrow transplantation. These findings are in agreement with previously reported studies (3,7,14,17). Overall, in our group of patients no infectious diagnosis for which conventional treatment is available were missed by BAL. Our study group, however, did not include a large group of neutropenic patients who are more prone to invasive fungi and bacteria which may be more difficult to diagnose by BAL (18,19). It seems that a negative BAL in non-neutropenic immunocompromized patients is likely to exclude an infectious etiology. BAL should be the first step in the diagnostic work-up of diffuse pulmonary infiltrates. Open lung biopsy should be persued as the second step ifBAL is noncontributive, in a clinical set up with high suspicion of fungi or drug related pneumonitis or if the clinical course is not compatible with the diagnosis derived by the BAL studies. In conclusion, BAL is a safe procedure with a high diagnostic yield for infectious etiologies in nonAIDS immunocompromized patients with diffuse bilateral pneumonitis. Bronchoscopy with BAL should be the primary diagnostic approach in these patients.
References
l. Bode FR., Pare .IA, Fraser RG. Pulmonary disease in the compromised host. Medicine 1974; 53: 255-293.
316
R . B r e u e r et al.
2. Pennington JF, Fedman NT. Pulmonary infiltrates and fever in patients with hematologic malignancy: assessment of transbronchial biopsy. A m J Med 1977; 62: 581-587. 3. Aisner J, Schimpff SC, Wiemik PH. Treatment of invasive aspergillosis. Relation to early diagnosis and treatment to response. Ann h~tern Med 1977; 86: 539-543. 4. Stover DE, White DA, Romano PA, Gellene RA. Diagnosis of pulmonary disease in acquired immune deficiency syndrome (AIDS). Am Rev Respir Dis 1984; 130: 659-662. 5. Mathay RA, Moritz ED. Invasive procedures for diagnosing pulmonary infection: a critical review. Clin Chest Med 1981; 2: 2-18. 6. Cheson BD, Samlowski WE, Tang TT, Spruance SL. Value of open-lung biopsy in 87 immunocompromised patients with pulmonary infiltrates. Cancer 1985; 55: 453-459. 7. Jaffe JP, Maki DG. Lung biopsy in immunocompromised patients: one institution's experience and approach to management of pulmonary disease in the compromised host. Cancer 1981; 48:1144-1153. 8. Cunningham JH, Zavala DC, Corry RJ, Keim LW. Trephine air drill, bronchial brush and fiberoptic transbronchial lung biopsies in immunosuppressed patients. Am RevRespirDis 1977; 115: 213-220. 9. Greenman RL, Goodall PT, King D. Lung biopsy in immunocompromisedhost. Am J M e d 1975;59: 488- 496. 10. Springmeyer SC, Hackman RC, Holle R, et al. Use of bronchoalveolar lavage to diagnose acute diffuse pneumonia in the immunocompromised host. J hlfect Dis 1986; 154: 604-610.
11. Palmer DL, Davidson M, Lusk R. Needle aspiration of the lung in complex pneumonias. Chest 1980;78: 16-21. 12. Broaddus C, Dake MD, Stulbarg MS et al. Bronchoalveolar lavage and transbronchial biopsy for the diagnosis of pulmonary infections in acquired immunodificiency syndrome. Ann Intern Med 1985; 102: 747-752. 13. Stover DE, Zaman MB, Hajdu SI, Lange M, Gold J, Armstrong D. Bronchoalveolar lavage in the diagnosis of diffuse pulmonary infiltrates in the immunosuppressed host. Ann lntern Med 1984; 101: I-7. 14. Cardonnier C, Bernaudin JF, Fleury J, et al. Diagnostic yieldof bronchoalveolar Iavagein pneumonitis occurring after allogeneic bone marrow transplantation. Am Rev Respir Dis 1985; 132:1118-1123. 15. Martin II WJ, Smith TF, Sanderson DR, Brutinel WM, Cockerill III FR, Douglas WW. Role ofbronchoalveolar lavage in the assessment of opportunistic pulmonary infections: utility and complications. Mayo Clin Proe 1987; 62: 549-557. 16. Young JA, Hopkin JM, Cuthbertson WP. Pulmonary infiltrates in immunocompromised patients: diagnosis by cytological examination of bronchoalvealar lavage fluid. J Clin Patho11984; 37: 390-397. 17. Crawford SW, Hackman RC, Clark JG. Open lungbiopsy diagnosis of diffuse pulmonary infiltrates after marrow transplantation. Chest 1988; 94: 949-953. 18. Kahn FW, Jones JM, England DM. The role of bronchoalveolar lavage in the diagnosis of invasive pulmonary aspergillosis. Am J Clin Patho11986; 86:518-522 19. Thorpe JE, Baughman RP, Frame PT, Wesseler TA, Staneck JL. Bronchoalveolar Iavagefor diagnosingacute bacterial pneumonia. J h f e c Dis 1987; 155:855 861.