Letters to the editor
Bronchoalveolar lavage is a safe and informative procedure in haematologic patients with nonresolving pneumonia To the editor: Pulmonary infection is a major complication during the treatment of haematological diseases (1). The initial choice of antimicrobial therapy is empiric. Positive blood culture or examination of the sputum may reveal the aetiological microorganisms, as may bronchoalveolar lavage (BAL) (2). To elucidate the benefit of the BAL procedure in haematological patients with nonresolving pneumonias, we studied the results of sampled material from 24 consecutive BAL procedures. The patients were 17 males and 7 females, suffering from acute myeoloid leukaemia (AML) (N = 8), acute lymphoid leukaemia (ALL) (N = 3), nonHodgkin lymphoma (N = 7), Hodgkin disease (N = 3) and chronic lymphocytic leukaemia (CLL) (N = 3). All patients had fever as the initial symptom. Simultaneous chest X-ray showed infiltrates in 13 patients. Subsequent pulmonary infiltrates occurred at a median of 1 day later (range 0-22) in 11 patients. Initially, most of the patients were treated with an aminoglycoside in combination with penicillin or cefuroxime. In some patients the antibiotic treatment was changed to either erythromycin, trimethoprim/ sulfamethoxazole, amphotericin-B, piperacillin or ceftazidime due to lack of improvement in the clinical condition. Furthermore BAL was performed a median of 6.5 d (range 0-28) after the infiltrates were first demonstrated. Fourteen patients had blood granulocytes above 0.5 x 10y/l, and 10 patients below 0.5 x 109/l. The platelet counts were below 10 x 109/1 in 3 patients, 10 to 20 x 10"l in 5 patients, 31 to 50 x 109/1 in 2 patients and above 50 x 109/1in 13 patients. All bronchoscopies were performed under local anaesthesia using a flexible fiberoptic bronchoscope. The procedure was performed with the bronchoscope wedged into the most suitable segment, usually in the middle lobe or lingula or in the most abnormal lobe according to chest X-ray. Lukewarm (37°C) sterile water (50 ml) followed by isotonic saline (1 00- 150 ml) were lavaged in 50 ml aliquots and aspirated. Transbronchial lung biopsies and brushing were not performed, in order to avoid bleeding due to thrombocytopenia (3). The BAL specimens were cultured and examined for bacteria, parasites (Pneumocystis carinii), fungi and viruses (herpes 280
Table 1. Identified microorganisms in the BAL specimens Number of patients Pneumocystis carinii Candida albicans Legionella pneumophila Haernophilus influenzae Mycobacteria CMV and aspergillus fragilis Streptococcus pneumoniae Enterobacter cloacae and micrococcus species BAL without identification of microorganisms
4 2 1 1 1 1 1 1 12
simplex virus and cytomegalovirus). Sixteen of the 24 BAL fluids were cultured for CMV. The microbiological examinations revealed a possible aetiological microorganism for the nonresolving pneumonia in 12 of the 24 episodes (50 ). Campbell et al. (4) found a similar percentage ( 5 5 %) of conclusive BAL examination in a comparable group of patients. The BAL results are given in Table 1. The procedure resulted in therapeutic changes in 75% of the patients, including start of antimicrobial treatment, change of antimicrobial treatment, and termination of therapy. Six patients (25%) had no change of treatment after BAL. No severe complications due to the BAL procedure were observed. Of special interest is that bleeding in the thrombocytopenic patients was not observed. Thrombocytopenic patients received 4 platelet transfusion before and after BAL. None of the patients needed ventilation support after the procedure. Our conclusion is that BAL is an uncomplicated, informative and necessary procedure in haematological patients with nonresolving pneumonia. Future prospective studies in our department will reveal the specificity and sensitivity of this diagnostic procedure in haematological patients with early pulmonary infiltrates.
References 1. MCCABERE. Diagnosis of pulmonary infections in inimunocompromised patients. Med Clin North Am 1988: 72: 10671083. 2. FEINSILVER SH, FEINA M , NIEDERMAN MS, SCHULTZ DE,
Letters to the editor FAEGENBURG DH. Utility of fiberoptic bronchoscopy in nonresolving pneumonia. Chest 1990: 98: 1322-1326. 3. HEURLINN, B R A T T S T R 0 M L0NNQVIST B, WESTMAN L, LIDMANC, ANDERSONJ. Aetiology of pulmonary diseases in imrnunocoinpriinised patients. Eur Respir J 1991: 4: 10-18. 4. CAMPBELL JH, RAINAV, BANHAMSW, CUNNINGHAM D, SOUKOP M. Pulmonary infiltrates - diagnostic problems in lymphoma. Postgrad Med J 1989: 65: 881-884
c,
Correspondence: Trine Lindhart Pedersen', Michael Pedersen', John Myhre', Nils Milman2, Hans Johnsen I , 'Departmerit of Haematology, Herlev Hospital, DK-2730 Herlev and 'Department of Pulmonary Medicine, Gentofte Hospital, DK-2900 Hellerup, Denmark,
28 1
Bronchoalveolar lavage is a safe and informative procedure in haematologic patients with nonresolving pneumonia To the editor: Pulmonary infection is a major complication during the treatment of haematological diseases (1). The initial choice of antimicrobial therapy is empiric. Positive blood culture or examination of the sputum may reveal the aetiological microorganisms, as may bronchoalveolar lavage (BAL) (2). To elucidate the benefit of the BAL procedure in haematological patients with nonresolving pneumonias, we studied the results of sampled material from 24 consecutive BAL procedures. The patients were 17 males and 7 females, suffering from acute myeoloid leukaemia (AML) (N = 8), acute lymphoid leukaemia (ALL) (N = 3), nonHodgkin lymphoma (N = 7), Hodgkin disease (N = 3) and chronic lymphocytic leukaemia (CLL) (N = 3). All patients had fever as the initial symptom. Simultaneous chest X-ray showed infiltrates in 13 patients. Subsequent pulmonary infiltrates occurred at a median of 1 day later (range 0-22) in 11 patients. Initially, most of the patients were treated with an aminoglycoside in combination with penicillin or cefuroxime. In some patients the antibiotic treatment was changed to either erythromycin, trimethoprim/ sulfamethoxazole, amphotericin-B, piperacillin or ceftazidime due to lack of improvement in the clinical condition. Furthermore BAL was performed a median of 6.5 d (range 0-28) after the infiltrates were first demonstrated. Fourteen patients had blood granulocytes above 0.5 x 10y/l, and 10 patients below 0.5 x 109/l. The platelet counts were below 10 x 109/1 in 3 patients, 10 to 20 x 10"l in 5 patients, 31 to 50 x 109/1 in 2 patients and above 50 x 109/1in 13 patients. All bronchoscopies were performed under local anaesthesia using a flexible fiberoptic bronchoscope. The procedure was performed with the bronchoscope wedged into the most suitable segment, usually in the middle lobe or lingula or in the most abnormal lobe according to chest X-ray. Lukewarm (37°C) sterile water (50 ml) followed by isotonic saline (1 00- 150 ml) were lavaged in 50 ml aliquots and aspirated. Transbronchial lung biopsies and brushing were not performed, in order to avoid bleeding due to thrombocytopenia (3). The BAL specimens were cultured and examined for bacteria, parasites (Pneumocystis carinii), fungi and viruses (herpes 280
Table 1. Identified microorganisms in the BAL specimens Number of patients Pneumocystis carinii Candida albicans Legionella pneumophila Haernophilus influenzae Mycobacteria CMV and aspergillus fragilis Streptococcus pneumoniae Enterobacter cloacae and micrococcus species BAL without identification of microorganisms
4 2 1 1 1 1 1 1 12
simplex virus and cytomegalovirus). Sixteen of the 24 BAL fluids were cultured for CMV. The microbiological examinations revealed a possible aetiological microorganism for the nonresolving pneumonia in 12 of the 24 episodes (50 ). Campbell et al. (4) found a similar percentage ( 5 5 %) of conclusive BAL examination in a comparable group of patients. The BAL results are given in Table 1. The procedure resulted in therapeutic changes in 75% of the patients, including start of antimicrobial treatment, change of antimicrobial treatment, and termination of therapy. Six patients (25%) had no change of treatment after BAL. No severe complications due to the BAL procedure were observed. Of special interest is that bleeding in the thrombocytopenic patients was not observed. Thrombocytopenic patients received 4 platelet transfusion before and after BAL. None of the patients needed ventilation support after the procedure. Our conclusion is that BAL is an uncomplicated, informative and necessary procedure in haematological patients with nonresolving pneumonia. Future prospective studies in our department will reveal the specificity and sensitivity of this diagnostic procedure in haematological patients with early pulmonary infiltrates.
References 1. MCCABERE. Diagnosis of pulmonary infections in inimunocompromised patients. Med Clin North Am 1988: 72: 10671083. 2. FEINSILVER SH, FEINA M , NIEDERMAN MS, SCHULTZ DE,
Letters to the editor FAEGENBURG DH. Utility of fiberoptic bronchoscopy in nonresolving pneumonia. Chest 1990: 98: 1322-1326. 3. HEURLINN, B R A T T S T R 0 M L0NNQVIST B, WESTMAN L, LIDMANC, ANDERSONJ. Aetiology of pulmonary diseases in imrnunocoinpriinised patients. Eur Respir J 1991: 4: 10-18. 4. CAMPBELL JH, RAINAV, BANHAMSW, CUNNINGHAM D, SOUKOP M. Pulmonary infiltrates - diagnostic problems in lymphoma. Postgrad Med J 1989: 65: 881-884
c,
Correspondence: Trine Lindhart Pedersen', Michael Pedersen', John Myhre', Nils Milman2, Hans Johnsen I , 'Departmerit of Haematology, Herlev Hospital, DK-2730 Herlev and 'Department of Pulmonary Medicine, Gentofte Hospital, DK-2900 Hellerup, Denmark,
28 1
