- Email: [email protected]
The diagnosis of pneumonia in the immunocompromised child: Use of bronchoalveolar lavage
The Diagnosis
of Pneumonia in the Immunocompromised Use of Bronchoalveolar Lavage
By Andrea
L. Winthrop,
Thomas
Waddell,
Child:
and Riccardo A. Superina
Toronto, Ontario l Between January 1987 and December 1988,28 immunocompromised children (aged 15 months to 17 years) underwent bronchoalveolar lavage (8AL) for evaluation of pneumonia (chemotherapy for malignancy, 12; orthotopic liver transplantation, 9: other hematologic disease, 5). Bilateral diffuse pulmonary disease was present in 25 children. All were receiving broad spectrum antibiotics. In addition, five were receiving antiviral therapy and two were receiving antifungal therapy. Sixteen patients underwent rigid end 10 underwent flexible bronchoscopy. Two lavages of 10 to 20 mL of normal saline were obtained from involved subsegmental bronchi of both lungs in each patient. Second wash samples from each lung were sent for bacterial and viral cultures, silver staining for pneumocystis, and direct electronmicroscopy analysis for viral particles. Samples were considered satisfactory if they contained an ambundance of alveolar macrophages and only small numbers of upper respiratory tract epithelial cells. Alveolar macrophages were present in 21 (81%) of the 8AL samples. A specific infectious agent was identified in 15 of these patients (cytomegalovirus [CMV], 8: Pneumocystis carinii, 4; gram-positive cocci. 3; Candida albicans, 2). and therapy was modified in 12. In the five patients in whom BAL samples were contaminated with upper respiratory tract cells no infectious agents were isolated. Because of continued clinical deterioration, open-lung biopsies were performed in three patients in whom 8AL had identified CMV and in three patients in whom no organisms had been obtained. Lung biopsies did not identify any new infectious agents. although in the latter group specific histological diagnosis of a noninfectious process was made (hemorrhagic infarct, bronchiolitis obliterans. and lymphoma). Our experience suggests that BAL is an excellent screening procedure for the diagnosis of pneumonia in immunosuppressed children and that it significantly decreases the need for open-lung biopsies in order to identify opportunistic pathogens in this patient population. @ 1990 by W.B. Saunders Company. INDEX WORDS: lmmunocompromised children; bronchoalveolar lavage.
T
HE IMMUNOCOMPROMISED child is susceptible to a variety of opportunistic pulmonary infections associated with significant morbidity and From the Department of Surgery. The Hospital for Sick Children. University of Toronto, Toronto, Ontario, Canada. Presented at the 38th Annual Meeting of the Surgical Section of the American Academy of Pediatrics, Chicago, Illinois, October 21-23.1989. Address reprint requests to R.A. Superina. MD. Department of Surgery, The Hospital for Sick Children, 555 University Ave. Toronto, Ontario M5G 1X8. Canada. 8 1990 by W.B. Saunders Company. 0022-3468/90/2508-0010$03.00/0 878
mortality. In many of these children, the clinical presentation and radiological abnormalities may be atypical, and often indistinguishable from other noninfectious pulmonary pathology.‘,* Because of the risks associated with the inappropriate administration of broad-spectrum antimicrobial therapy, a definitive diagnosis is often desirable prior to the initiation of therapy. Open-lung biopsy has long been the most accurate diagnostic procedure available.2-5 However, it is associated with significant morbidity**“’ and, although it provides a high diagnostic yield that may result in a change in therapy,2V5-7it has not been convincingly shown to contribute to an increase in patient survival.3T6 Bronchoalveolar lavage (BAL) has recently been advocated as a useful noninvasive diagnostic procedure for evaluating acute pulmonary pathology in immunocompromised children and adults.‘V8-‘4The purpose of this retrospective review is to evaluate our experience with BAL at the Hospital for Sick Children, and to better define the role of BAL in the diagnosis and management of pneumonia in the immunocompromised child. MATERIALS
AND
METHODS
Between January 1987 and December 1988,26 immunocomprcmised children aged 15 months to 17 years underwent BAL for evaluation of pneumonia. Twelve children were receiving chemotherapy for malignancy, nine were receiving immunosuppression following orthotopic liver transplantation, and five had other hematologic or immune abnormalities (Table 1). Bilateral diffuse pulmonary disease was present in 25 patients. One child with aplastic anemia had pyrexia of unknown origin and a coin lesion on chest roentgenogram. Ten children required supplemental oxygen by mask, and two were ventilated preoperatively. At the time of BAL, all children were receiving broad spectrum antibiotics. In addition, five were receiving antiviral therapy and two were receiving antifungal therapy. Treatment was initiated 1 to 21 days (mean, 4 + 2) prior to the BAL. Bronchoscopy was performed under general anesthesia in 23 patients. Sixteen of these children underwent rigid bronchoscopy using a Storz (Rexdale, Ontario) ventilating bronchoscope. In seven, flexible fiberoptic bronchoscopy was performed through an indwelling endotracheal tube using the Olympus pediatric bronchoscope BF3C4 (Olympus Corp, New Hyde Park, NY). Flexible bronchoscopy was performed in three additional patients who received intravenous sedation and topical lidocaine and had transnasal passage of the Olympus model BF4B2 (Olympus Carp) bronchoscope. In each patient, two lavages of 10 to 20 mL of normal saline were obtained from involved subsegmental bronchi of both lungs. Second wash samples from each lung were evaluated by Gram stain and by bacterial, fungal, and viral cultures. Additional aliquots were anaJournal of Pediatric Swgm-y, Vol 25, No 8 (August),1990:
pp 878-880
879
BAL FOR THE DIAGNOSIS OF PNEUMONIA
Table 1. The Hospitel for Sick Children: January 1997 to December 1966 (n = 26)
No.of Patients
Diagnosis Acute lymphocytic leukemia
9
Orthotopic liver transplantation
9
Acute myalogenous leukemia
2
Aplestic anemia
2
Severe combined immune deficiency
2
Ataxia telangactasia
1
Rhabdomyosarcoma
1
lysed by methanamine silver stain for Pneumocystis carinii, and direct electronmicroscopy (EM) analysis for viral particles. Samples were considered satisfactory if they contained an abundanceof alveolar macrophages and only small numbers of upper respiratory tract epithelial cells.
RESULTS
Alveolar macrophages were present in 21 (8 1%) of the lavage samples. A specific infectious agent was identified in 15 of these patients, resulting in an alteration in therapy in 12 (Table 2). Eight of the latter 12 children (75%) survived. No infectious agent was identified in six patients, despite adequate sampling. Five of these six children had received greater than 5 days of antimicrobial therapy prior to BAL. In five children (19%), BAL samples were unsatisfactory for analysis because of upper respiratory tract contamination. No infectious agents were isolated from these samples. Two of these five patients had continued clinical improvement without further intervention. Of the six patients requiring mechanical ventilation after BAL, two had been ventilated preoperatively. The remaining four were extubated by 48 hours postprocedure. There were no perioperative deaths and no other identifiable perioperative morbidity. DISCUSSION
Evaluation of a new diagnostic procedure must take into consideration the associated procedural morbidity and mortality, the diagnostic yield, the influence of results on patient management, and impact on patient survival. In the evaluation of immunocompromised patients, BAL has been shown to be a noninvasive procedure associated with no mortality and minimal Table 2. Results of Bronchoalveolar Lavage (n = 21 Satisfactory Samples1
Diagnosis
No.
in Therapy
SUNiWl
morbidity in this high-risk patient population.‘*4*8 In our series, there were no perioperative deaths, and four of six patients requiring postoperative ventilation were extubated by 48 hours. In contrast, open-lung biopsy is invasive and can be associated with significant morbidity such as pneumothorax and the need for prolonged ventilation.2,4-7 BAL identified an infectious agent in 15 (58%) of the 26 children in our series. Furthermore, in the 21 children in whom adequate samples were obtained, the calculated diagnostic yield was 71%. These results compare favorably with previously reported series, in which specific infectious agents were identified in 40% to 71% of patients.‘*8,‘1-14 Williams et al’ have shown that bronchoscopy in immunocompromised patients has a sensitivity of 90% for the diagnosis of pulmonary infection. In their series, a negative BAL was associated with a very low probability that an infectious process would be found at open-lung biopsy. Our experience is comparable. Because of continued clinical deterioration, open-lung biopsies were performed in six of the 26 children (Table 3). Open-lung biopsies were performed in three patients in whom BAL had identified cytomegalovirus (CMV) and in three patients in whom no organisms had been obtained. Lung biopsies did not identify any new infectious agents, although in the latter group specific histological diagnosis of a noninfectious process was made. In the immunocompromised host with acute pulmonary disease, controversy exists with respect to the clinical benefit of a definitive etiologic diagnosis. Definitive diagnosis of a specific infectious agent by either noninvasive (BAL) or invasive (open-lung biopsy) methods has been shown to alter therapy in a significant number of these patients.2*5~7~‘0 In this series, therapy was altered in 12 of 15 (80%) children in whom a specific infectious agent was identified by BAL and in one of six children where no infectious agent was isolated (Table 2). Alterations in therapy included cessation of broad-spectrum antibiotics and/or initiation of specific antifungal, antibiotic, or antiviral therapy. In the past, alterations in patient management have resulted in little demonstrable affect on overall patient survival.3*6However, with the recent advent of specific effective antiviral therapy for CMV and other viruses, Table 3. Results of Open-Lung Biopsy (n = 6) Bronchoaiveolsr Lava@ (No.1
LungBiopsy(No.)
Cytomegalovirus
6
5
2 (33%)
Pneumocystic carinii
4
3
4 (100%)
Cytomegalovirus (3)
Gram-positive cocci
3
2
2 (66%)
No infectious agent (3)
Candid8 albicans
2
2
2 (100%)
Bronchiolitis obliterans (1)
No infectious agent
6
1
5 (83%)
Lymphoma
Cytomegalovirus (3) Hamorrhagic infarct (1)
(1)
880
WINTHROP, WADDELL, AND SUPERINA
survival may be improved when the definitive diagnosis of viral pneumonitis results in initiation of appropriate treatment. Foglia et al7 demonstrated a 55% survival in nine patients with viral pneumonitis diagnosed by lung biopsy. In this series, there was a 23% incidence of CMV pneumonitis. This was associated with a 67% mortality, despite antiviral therapy. However, the two children who were treated with DHPG (Ganciclovir; Syntex, Palo Alto, CA) survived. Because DHPG is a myelosuppressive agent, therapy was only initiated after confirmed diagnosis of CMV infection by BAL. Our experience suggests that BAL is an excellent
noninvasive screening procedure for the diagnosis of pneumonia in immunosuppressed children. It signihcantly decreases the need for open-lung biopsies in order to identify opportunistic pathogens in this patient population. In the majority of patients in whom adequate samples are obtained, BAL yields a specific diagnosis of pulmonary infection that may result in an alteration in therapy and possibly improved survival in these critically ill children. We feel that open-lung biopsy should be reserved for those patients in whom BAL is nondiagnostic and in whom there is continued clinical deterioration.
REFERENCES 1. Williams D, Yungbluth M, Adams G, et al: The role of fiberoptic bronchoscopy in the evaluation of immunocompromised hosts with diffuse pulmonary infiltrates. Am Rev Resp Dis 131:880885,1985 2. Prober CG, Whyte H, Smith CR: Open-lung biopsy in immunocompromised children with pulmonary infiltrates. Am J Dis Child 138:60-63,1984 3. Shorter NA, Ross AJ III, August C, et al: The usefulness of open-lung biopsy in the pediatric bone marrow transplant population. J Pediatr Surg 23533-537, 1988 4. Ballantine TVN, Grosfeld JL, Knapek RM, et al: Interstitial pneumonitis in the immunologically suppressed child. An urgent surgical condition. J Pediatr Surg 12:501-508, 1977 5. Early GL, Williams TE, Kilman JW: Open-lung biopsy: Its effect on therapy in the pediatric patient. Chest 87:467-469, 1985 6. Doolin EJ, Luck SR, Sherman JO, et ai: Emergency lung biopsy: Friend or foe of the immunosuppressive child? J Pediatr Surg 21:485-487, 1986 7. Foglia RP, Shilyansky J, Fonkalsrud EW: Emergency lung biopsy in immunocompromised pediatric patients. Ann Surg 210:9092,1989
8. Stover DE, White DA, Roman0 PA, et al: Diagnosis of pulmonary disease in acquired immune deficiency syndrome (AIDS). Am Rev Resp Dis 130:659-662,1984 9. Leigh MW, Henshaw NG, Wood RE: Diagnosis of Pneumocystis corinii pneumonia in pediatric patients using bronchoscopic bronchoalveolar lavage. Ped Infect Dis J 4:408-410,198s 10. Frank01 LR, Smith DW, Lewiston NJ: Bronchoalveolar lavage for diagnosis of pneumonia in the immunocompromised child. Pediatrics 81:785-788, 1988 11. Pattishall EN, Noyes BE, Orenstein DM: Use of bronchoalvealar lavage in immunocompromised children with pneumonia. Pediatr Pulmonol5:1-5, 1988 12. Allen KA, Markin RS, Rennard SI, et al: Bronchoalveolar lavage in liver transplant patients. Acta Cytol33:539-543, 1989 13. Bye MR, Bernstein L, Shah K, et al: Diagnostic bronchoalveolar lavage in children with AIDS. Pediatr Pulmonol 3:425-428, 1987 14. de Blec J, McKelvie P, Le Bourgeois M, et al: Value of bronchoalveolar lavage in the management of severe acute pneumonia and interstitial pneumonitis in the immunocompromised child. Thorax 42:759-765, 1987
Discussion M.Reynolds (Chicago, IL): Most of these patients had to be anesthetized for bronchoscopy and some also for open-lung biopsy. Do you think it is reasonable to ask the oncologist which of the two procedures he/she would like in order to save the child one anesthetic? A. Winthrop (response): We actually were interested in doing both procedures simultaneously. In some of the patients, where the oncologist felt that the child had a noninfectious process, we did them that way
initially. I think that it’s probably not appropriate to ask the oncologists which technique they preferred. The gold standard in the literature is still open-lung biopsy and they would be very hesitant to accept BAL. We usually ask them what is the most likely cause of the child’s pulmonary infiltrate. If it is an organism that is shown to be found with high yield on BAL, then we usually tell them that’s what we’re initially going to do and don’t give them a choice.
of Pneumonia in the Immunocompromised Use of Bronchoalveolar Lavage
By Andrea
L. Winthrop,
Thomas
Waddell,
Child:
and Riccardo A. Superina
Toronto, Ontario l Between January 1987 and December 1988,28 immunocompromised children (aged 15 months to 17 years) underwent bronchoalveolar lavage (8AL) for evaluation of pneumonia (chemotherapy for malignancy, 12; orthotopic liver transplantation, 9: other hematologic disease, 5). Bilateral diffuse pulmonary disease was present in 25 children. All were receiving broad spectrum antibiotics. In addition, five were receiving antiviral therapy and two were receiving antifungal therapy. Sixteen patients underwent rigid end 10 underwent flexible bronchoscopy. Two lavages of 10 to 20 mL of normal saline were obtained from involved subsegmental bronchi of both lungs in each patient. Second wash samples from each lung were sent for bacterial and viral cultures, silver staining for pneumocystis, and direct electronmicroscopy analysis for viral particles. Samples were considered satisfactory if they contained an ambundance of alveolar macrophages and only small numbers of upper respiratory tract epithelial cells. Alveolar macrophages were present in 21 (81%) of the 8AL samples. A specific infectious agent was identified in 15 of these patients (cytomegalovirus [CMV], 8: Pneumocystis carinii, 4; gram-positive cocci. 3; Candida albicans, 2). and therapy was modified in 12. In the five patients in whom BAL samples were contaminated with upper respiratory tract cells no infectious agents were isolated. Because of continued clinical deterioration, open-lung biopsies were performed in three patients in whom 8AL had identified CMV and in three patients in whom no organisms had been obtained. Lung biopsies did not identify any new infectious agents. although in the latter group specific histological diagnosis of a noninfectious process was made (hemorrhagic infarct, bronchiolitis obliterans. and lymphoma). Our experience suggests that BAL is an excellent screening procedure for the diagnosis of pneumonia in immunosuppressed children and that it significantly decreases the need for open-lung biopsies in order to identify opportunistic pathogens in this patient population. @ 1990 by W.B. Saunders Company. INDEX WORDS: lmmunocompromised children; bronchoalveolar lavage.
T
HE IMMUNOCOMPROMISED child is susceptible to a variety of opportunistic pulmonary infections associated with significant morbidity and From the Department of Surgery. The Hospital for Sick Children. University of Toronto, Toronto, Ontario, Canada. Presented at the 38th Annual Meeting of the Surgical Section of the American Academy of Pediatrics, Chicago, Illinois, October 21-23.1989. Address reprint requests to R.A. Superina. MD. Department of Surgery, The Hospital for Sick Children, 555 University Ave. Toronto, Ontario M5G 1X8. Canada. 8 1990 by W.B. Saunders Company. 0022-3468/90/2508-0010$03.00/0 878
mortality. In many of these children, the clinical presentation and radiological abnormalities may be atypical, and often indistinguishable from other noninfectious pulmonary pathology.‘,* Because of the risks associated with the inappropriate administration of broad-spectrum antimicrobial therapy, a definitive diagnosis is often desirable prior to the initiation of therapy. Open-lung biopsy has long been the most accurate diagnostic procedure available.2-5 However, it is associated with significant morbidity**“’ and, although it provides a high diagnostic yield that may result in a change in therapy,2V5-7it has not been convincingly shown to contribute to an increase in patient survival.3T6 Bronchoalveolar lavage (BAL) has recently been advocated as a useful noninvasive diagnostic procedure for evaluating acute pulmonary pathology in immunocompromised children and adults.‘V8-‘4The purpose of this retrospective review is to evaluate our experience with BAL at the Hospital for Sick Children, and to better define the role of BAL in the diagnosis and management of pneumonia in the immunocompromised child. MATERIALS
AND
METHODS
Between January 1987 and December 1988,26 immunocomprcmised children aged 15 months to 17 years underwent BAL for evaluation of pneumonia. Twelve children were receiving chemotherapy for malignancy, nine were receiving immunosuppression following orthotopic liver transplantation, and five had other hematologic or immune abnormalities (Table 1). Bilateral diffuse pulmonary disease was present in 25 patients. One child with aplastic anemia had pyrexia of unknown origin and a coin lesion on chest roentgenogram. Ten children required supplemental oxygen by mask, and two were ventilated preoperatively. At the time of BAL, all children were receiving broad spectrum antibiotics. In addition, five were receiving antiviral therapy and two were receiving antifungal therapy. Treatment was initiated 1 to 21 days (mean, 4 + 2) prior to the BAL. Bronchoscopy was performed under general anesthesia in 23 patients. Sixteen of these children underwent rigid bronchoscopy using a Storz (Rexdale, Ontario) ventilating bronchoscope. In seven, flexible fiberoptic bronchoscopy was performed through an indwelling endotracheal tube using the Olympus pediatric bronchoscope BF3C4 (Olympus Corp, New Hyde Park, NY). Flexible bronchoscopy was performed in three additional patients who received intravenous sedation and topical lidocaine and had transnasal passage of the Olympus model BF4B2 (Olympus Carp) bronchoscope. In each patient, two lavages of 10 to 20 mL of normal saline were obtained from involved subsegmental bronchi of both lungs. Second wash samples from each lung were evaluated by Gram stain and by bacterial, fungal, and viral cultures. Additional aliquots were anaJournal of Pediatric Swgm-y, Vol 25, No 8 (August),1990:
pp 878-880
879
BAL FOR THE DIAGNOSIS OF PNEUMONIA
Table 1. The Hospitel for Sick Children: January 1997 to December 1966 (n = 26)
No.of Patients
Diagnosis Acute lymphocytic leukemia
9
Orthotopic liver transplantation
9
Acute myalogenous leukemia
2
Aplestic anemia
2
Severe combined immune deficiency
2
Ataxia telangactasia
1
Rhabdomyosarcoma
1
lysed by methanamine silver stain for Pneumocystis carinii, and direct electronmicroscopy (EM) analysis for viral particles. Samples were considered satisfactory if they contained an abundanceof alveolar macrophages and only small numbers of upper respiratory tract epithelial cells.
RESULTS
Alveolar macrophages were present in 21 (8 1%) of the lavage samples. A specific infectious agent was identified in 15 of these patients, resulting in an alteration in therapy in 12 (Table 2). Eight of the latter 12 children (75%) survived. No infectious agent was identified in six patients, despite adequate sampling. Five of these six children had received greater than 5 days of antimicrobial therapy prior to BAL. In five children (19%), BAL samples were unsatisfactory for analysis because of upper respiratory tract contamination. No infectious agents were isolated from these samples. Two of these five patients had continued clinical improvement without further intervention. Of the six patients requiring mechanical ventilation after BAL, two had been ventilated preoperatively. The remaining four were extubated by 48 hours postprocedure. There were no perioperative deaths and no other identifiable perioperative morbidity. DISCUSSION
Evaluation of a new diagnostic procedure must take into consideration the associated procedural morbidity and mortality, the diagnostic yield, the influence of results on patient management, and impact on patient survival. In the evaluation of immunocompromised patients, BAL has been shown to be a noninvasive procedure associated with no mortality and minimal Table 2. Results of Bronchoalveolar Lavage (n = 21 Satisfactory Samples1
Diagnosis
No.
in Therapy
SUNiWl
morbidity in this high-risk patient population.‘*4*8 In our series, there were no perioperative deaths, and four of six patients requiring postoperative ventilation were extubated by 48 hours. In contrast, open-lung biopsy is invasive and can be associated with significant morbidity such as pneumothorax and the need for prolonged ventilation.2,4-7 BAL identified an infectious agent in 15 (58%) of the 26 children in our series. Furthermore, in the 21 children in whom adequate samples were obtained, the calculated diagnostic yield was 71%. These results compare favorably with previously reported series, in which specific infectious agents were identified in 40% to 71% of patients.‘*8,‘1-14 Williams et al’ have shown that bronchoscopy in immunocompromised patients has a sensitivity of 90% for the diagnosis of pulmonary infection. In their series, a negative BAL was associated with a very low probability that an infectious process would be found at open-lung biopsy. Our experience is comparable. Because of continued clinical deterioration, open-lung biopsies were performed in six of the 26 children (Table 3). Open-lung biopsies were performed in three patients in whom BAL had identified cytomegalovirus (CMV) and in three patients in whom no organisms had been obtained. Lung biopsies did not identify any new infectious agents, although in the latter group specific histological diagnosis of a noninfectious process was made. In the immunocompromised host with acute pulmonary disease, controversy exists with respect to the clinical benefit of a definitive etiologic diagnosis. Definitive diagnosis of a specific infectious agent by either noninvasive (BAL) or invasive (open-lung biopsy) methods has been shown to alter therapy in a significant number of these patients.2*5~7~‘0 In this series, therapy was altered in 12 of 15 (80%) children in whom a specific infectious agent was identified by BAL and in one of six children where no infectious agent was isolated (Table 2). Alterations in therapy included cessation of broad-spectrum antibiotics and/or initiation of specific antifungal, antibiotic, or antiviral therapy. In the past, alterations in patient management have resulted in little demonstrable affect on overall patient survival.3*6However, with the recent advent of specific effective antiviral therapy for CMV and other viruses, Table 3. Results of Open-Lung Biopsy (n = 6) Bronchoaiveolsr Lava@ (No.1
LungBiopsy(No.)
Cytomegalovirus
6
5
2 (33%)
Pneumocystic carinii
4
3
4 (100%)
Cytomegalovirus (3)
Gram-positive cocci
3
2
2 (66%)
No infectious agent (3)
Candid8 albicans
2
2
2 (100%)
Bronchiolitis obliterans (1)
No infectious agent
6
1
5 (83%)
Lymphoma
Cytomegalovirus (3) Hamorrhagic infarct (1)
(1)
880
WINTHROP, WADDELL, AND SUPERINA
survival may be improved when the definitive diagnosis of viral pneumonitis results in initiation of appropriate treatment. Foglia et al7 demonstrated a 55% survival in nine patients with viral pneumonitis diagnosed by lung biopsy. In this series, there was a 23% incidence of CMV pneumonitis. This was associated with a 67% mortality, despite antiviral therapy. However, the two children who were treated with DHPG (Ganciclovir; Syntex, Palo Alto, CA) survived. Because DHPG is a myelosuppressive agent, therapy was only initiated after confirmed diagnosis of CMV infection by BAL. Our experience suggests that BAL is an excellent
noninvasive screening procedure for the diagnosis of pneumonia in immunosuppressed children. It signihcantly decreases the need for open-lung biopsies in order to identify opportunistic pathogens in this patient population. In the majority of patients in whom adequate samples are obtained, BAL yields a specific diagnosis of pulmonary infection that may result in an alteration in therapy and possibly improved survival in these critically ill children. We feel that open-lung biopsy should be reserved for those patients in whom BAL is nondiagnostic and in whom there is continued clinical deterioration.
REFERENCES 1. Williams D, Yungbluth M, Adams G, et al: The role of fiberoptic bronchoscopy in the evaluation of immunocompromised hosts with diffuse pulmonary infiltrates. Am Rev Resp Dis 131:880885,1985 2. Prober CG, Whyte H, Smith CR: Open-lung biopsy in immunocompromised children with pulmonary infiltrates. Am J Dis Child 138:60-63,1984 3. Shorter NA, Ross AJ III, August C, et al: The usefulness of open-lung biopsy in the pediatric bone marrow transplant population. J Pediatr Surg 23533-537, 1988 4. Ballantine TVN, Grosfeld JL, Knapek RM, et al: Interstitial pneumonitis in the immunologically suppressed child. An urgent surgical condition. J Pediatr Surg 12:501-508, 1977 5. Early GL, Williams TE, Kilman JW: Open-lung biopsy: Its effect on therapy in the pediatric patient. Chest 87:467-469, 1985 6. Doolin EJ, Luck SR, Sherman JO, et ai: Emergency lung biopsy: Friend or foe of the immunosuppressive child? J Pediatr Surg 21:485-487, 1986 7. Foglia RP, Shilyansky J, Fonkalsrud EW: Emergency lung biopsy in immunocompromised pediatric patients. Ann Surg 210:9092,1989
8. Stover DE, White DA, Roman0 PA, et al: Diagnosis of pulmonary disease in acquired immune deficiency syndrome (AIDS). Am Rev Resp Dis 130:659-662,1984 9. Leigh MW, Henshaw NG, Wood RE: Diagnosis of Pneumocystis corinii pneumonia in pediatric patients using bronchoscopic bronchoalveolar lavage. Ped Infect Dis J 4:408-410,198s 10. Frank01 LR, Smith DW, Lewiston NJ: Bronchoalveolar lavage for diagnosis of pneumonia in the immunocompromised child. Pediatrics 81:785-788, 1988 11. Pattishall EN, Noyes BE, Orenstein DM: Use of bronchoalvealar lavage in immunocompromised children with pneumonia. Pediatr Pulmonol5:1-5, 1988 12. Allen KA, Markin RS, Rennard SI, et al: Bronchoalveolar lavage in liver transplant patients. Acta Cytol33:539-543, 1989 13. Bye MR, Bernstein L, Shah K, et al: Diagnostic bronchoalveolar lavage in children with AIDS. Pediatr Pulmonol 3:425-428, 1987 14. de Blec J, McKelvie P, Le Bourgeois M, et al: Value of bronchoalveolar lavage in the management of severe acute pneumonia and interstitial pneumonitis in the immunocompromised child. Thorax 42:759-765, 1987
Discussion M.Reynolds (Chicago, IL): Most of these patients had to be anesthetized for bronchoscopy and some also for open-lung biopsy. Do you think it is reasonable to ask the oncologist which of the two procedures he/she would like in order to save the child one anesthetic? A. Winthrop (response): We actually were interested in doing both procedures simultaneously. In some of the patients, where the oncologist felt that the child had a noninfectious process, we did them that way
initially. I think that it’s probably not appropriate to ask the oncologists which technique they preferred. The gold standard in the literature is still open-lung biopsy and they would be very hesitant to accept BAL. We usually ask them what is the most likely cause of the child’s pulmonary infiltrate. If it is an organism that is shown to be found with high yield on BAL, then we usually tell them that’s what we’re initially going to do and don’t give them a choice.