The Utility of Surgical Lung Biopsy in Immunocompromised Children

Transcription

1 The Utility of Surgical Lung Biopsy in Immunocompromised Children Jessica A. Naiditch, MD, Katherine A. Barsness, MD, and David H. Rothstein, MD Objective To determine the utility of lung biopsy in immunocompromised pediatric patients with suspected infectious lung disease and to evaluate the risks and benefits of biopsy in the era of minimally invasive thoracic surgery. Study design We reviewed charts for 50 immunocompromised patients who underwent surgical lung biopsy between January 2000 and July 2011 at a free-standing, tertiary care, urban children s hospital. The primary outcome variable was benefit from biopsy, defined as change in therapy based on biopsy results. The secondary outcome variable was survival to discharge. The c 2 analysis was used for categorical variables and Student t test for continuous variables. Results Biopsy provided a definitive histopathologic or microbiologic diagnosis in 25 patients (50%), the most common diagnosis being fungal infection (22%). Diagnostic and nondiagnostic biopsy results yielded benefit in 25 surviving patients (50%) for whom the biopsy results were used to tailor treatment. Taking more than one biopsy specimen did not improve diagnostic yield. Six patients (12%) had a major morbidity including reinsertion of chest tube after initial chest tube removal (3), prolonged air leak (1), and a new requirement for mechanical ventilation postoperatively (2). Two patients died postoperatively, but the mortalities were not clearly related to surgery. Underlying diagnoses included hematologic malignancy (64%), primary immunodeficiency (12%), organ transplant recipient (12%), and solid malignancy (10%). Twelve patients (24%) had undergone stem cell transplantation. Conclusion Lung biopsy in immunocompromised pediatric patients alters therapy in 50% of cases, but predictably carries identifiable morbidities. This study is limited by its retrospective nature. (J Pediatr 2013;162:133-6). Pulmonary 1-3 infections are a leading cause of therapy-related morbidity and mortality in pediatric oncology patients. The management of immunocompromised patients with persistent pulmonary infiltrates is often challenging. Diagnostic procedures should only be performed if results have the potential to change treatment. Bronchoalveolar lavage (BAL), although relatively noninvasive, fails to provide a diagnosis in up to two-thirds of patients. 4,5 Surgical lung biopsy has been recommended as a rapid and accurate method to obtain diagnosis in cases of nondiagnostic BALs or when BAL is deemed unlikely to yield results. Even though open lung biopsy has previously been shown to allow definitive diagnosis in 50%-80% of pediatric patients, complication rates range from 2%-52% depending on patient comorbidities. 4,6 Over the past 20 years, surgical techniques for lung biopsy have advanced and the spectrum of pulmonary infectious disease in immunocompromised pediatric patients has changed. The incidence of Pneumocystis carinii pneumonia, for example once a common diagnosis at lung biopsy has decreased dramatically with the introduction of antimicrobial prophylaxis. In addition, video assisted thoracoscopic surgery (VATS) has largely replaced open lung biopsy as a preferred technique. Although VATS biopsy for diffuse interstitial lung disease has been studied, diagnostic VATS lung biopsy in immunocompromised children for infectious concerns has not previously been evaluated for diagnostic yield and outcomes. 7 We hypothesized that complications from surgical lung biopsy have decreased in the VATS era in comparison with the era of open lung biopsy. Additionally, we hypothesized that the benefit derived from surgical lung biopsy may have diminished due to the decreased incidence of Pneumocystis carinii pneumonia. Methods We conducted a retrospective chart review of all patients undergoing surgical lung biopsy between January 2000 and July 2011 at a free-standing, tertiary care, urban children s hospital. This was conducted with Institutional Review Board approval (# ). Patients who underwent a diagnostic lung biopsy for a mass primarily concerning malignancy were excluded (n = 23). Immunocompetent patients who underwent a diagnostic surgical lung biopsy due to concern for infection were also excluded (n = 2). Fifty immunocompromised patients, comprised of children with either primary immunodeficiency syndromes or immunodeficiency due to immunosuppressive treatments, remained and were included in the study. Patient variables included age, sex, preoperative diagnosis, symptoms, and imaging. Preoperative and postoperative treatments variables included antibiotics, BAL PPP VATS Bronchoalveolar lavage Post-pull pneumothorax Video assisted thoracoscopic surgery From the Division of Pediatric Surgery, Ann and Robert H. Lurie Children s Hospital of Chicago, Northwestern University Feinberg School of Medicine, Chicago, IL The authors declare no conflicts of interest /$ - see front matter. Copyright ª 2013 Mosby Inc. All rights reserved

2 THE JOURNAL OF PEDIATRICS Vol. 162, No. 1 antiviral agents, antifungals, and steroids. Pathologic and microbiology culture results for lung biopsies were abstracted. Operative factors including approach (VATS vs open), operative time, duration of postoperative thoracostomy tube placement, and postoperative morbidities and mortalities were recorded. Patients were determined to have benefited from surgical biopsy if (1) there was a change in therapy based on biopsy results; and (2) the patients survived their hospitalization. Patient factors and outcomes of interest were compared using c 2 test for categorical data and Student t test for continuous data with significance determined by P <.05. Results We identified 50 immunocompromised patients who underwent a diagnostic lung biopsy in order to identify an infectious agent. Average patient age was years and 20 (40%) were female. The primary underlying diagnoses included hematologic malignancy (n = 32, 64%), primary immunodeficiency (n = 6, 12%), organ transplant recipient (n = 6, 12%), solid malignancy undergoing chemotherapy (n = 5, 10%), and chronic steroid use (n = 1, 2%). Twelve patients were stem cell transplant recipients (24%). Preoperative symptoms prompting radiographic imaging varied greatly, with the most common being fever in 29 (58%), cough in 21 (42%), tachypnea in 5 (10%), and chest pain in 5 (10%). Two patients (4%) were in respiratory failure preoperatively. Seven patients were asymptomatic and presumed to be undergoing routine radiographic screening. Preoperative chest computed tomography revealed nodular infiltrates in 40 patients (80%) and interstitial disease in 10 patients (20%). Prior to surgery, all patients underwent nasopharyngeal swabs for viruses and 9 patients (18%) underwent BAL; none of these examinations was diagnostic. The decision to proceed with surgical lung biopsy was made on a case-by-case basis. The main indication for biopsy in all patients included in this study was suspicion for an infectious etiology of persistent pulmonary infiltrates with either a nondiagnostic BAL or the clinical judgment that BAL would have a low diagnostic yield. Noninfectious causes for the infiltrates may have been in the differential diagnosis, but the primary suspicion and drive for biopsy was undiagnosed infection. Biopsy sites were chosen by the surgeon and were generally based on the location of prominent imaging findings. A thoracostomy tube was left in place postoperatively in all patients. Surgical lung biopsy provided a definitive histopathologic or microbiologic diagnosis in 25 patients (50%). Fungal infection was most commonly diagnosed with surgical lung biopsy (n = 11, 22%). Other diagnoses included bacterial pneumonia (n = 4, 8%), posttransplant lymphoproliferative disease (n = 3, 6%), tuberculosis (n = 1, 2%), viral pneumonia (n = 1, 2%), aspiration pneumonia (n = 1, 2%), radiation induced change (n = 1, 2%), graft vs host disease (n = 1, 2%), Hodgkin s relapse (n = 1, 2%), and xanthomatous disease (n = 1, 2%). Lung biopsy was nondiagnostic in 25 patients (50%). We sought to determine predictors for a diagnostic biopsy (Table I). Patients with a diagnostic biopsy were less likely to have a fever (P =.022). Taking more than 1 biopsy specimen did not predict a diagnostic biopsy (P =.74). Diagnostic and nondiagnostic biopsy results yielded benefit in 25 surviving patients (50%) for whom the biopsy results were used to tailor treatment (Table II). Treatment changes included change in antimicrobial therapy (n = 19, 38%) and change in immunosuppression (n = 6, 12%). Among 25 patients in whom biopsies did not result in a change in therapy, 23 survived to discharge. Surgical outcomes are detailed in Table III. The majority of patients underwent a VATS lung biopsy (n = 46, 92%). The choice for an open biopsy in 4 patients was attributable to surgeon preference. Postoperative major morbidity occurred in 6 patients (12%): pneumothorax requiring chest tube reinsertion (n = 3), prolonged air leak requiring 16 chest tube days (n = 1), and postoperative mechanical ventilation (n = 2). Overall, 24 patients (48%) had pneumothoraces after chest tube removal; 21 resolved without intervention. Two patients who were not intubated preoperatively but experienced preoperative respiratory distress were kept Table I. Predictors of diagnostic lung biopsy in immunocompromised children Diagnostic biopsy (n = 25) Nondiagnostic biopsy (n = 25) P Female 8 (32%) 12 (48%).39 Age (y; mean SD) Primary diagnosis.32 Hematologic cancer 15 (60%) 17 (68%) Primary immunodeficiency 2 (8%) 4 (16%) Organ transplant recipient 5 (20%) 1 (4%) Solid cancer 2 (8%) 3 (12%) Chronic steroid use 1 (4%) 0 Stem cell transplant recipient 8 (32%) 4 (16%).32 Preoperative CT findings Interstitial infiltrates 3 (12%) 7 (28%) Nodular infiltrates 22 (88%) 18 (72%).28 Preoperative symptoms Fever 10 (40%) 19 (76%).02 Cough 10 (40%) 11 (44%).77 Tachypnea 5 (20%) 0.06 Chest pain 3 (12%) 2 (8%).64 Fatigue 1 (4%) 1 (4%).47 Weight loss 1 (4%) 0.31 Night sweats 1 (4%) 0.31 Septic shock 0 (4%) 1 (4%).31 Asymptomatic 5 (4%) 2 (8%).42 Operative technique.18 Open 4 (16%) 0 VATS 21 (84%) 25 (100%) Number of biopsies taken (72%) 20 (80.0%) $2 7 (28%) 5 (20.0%) Preoperative absolute neutrophil count (10 3 /ml; mean SD) Preoperative treatment Antibiotics preoperatively 16 (64%) 15 (60%).77 Antifungals preoperatively 14 (56%) 17 (68%).56 CT, computed tomography. 134 Naiditch, Barsness, and Rothstein

3 January 2013 ORIGINAL ARTICLES Table II. Evaluating benefit to immunocompromised children after undergoing diagnostic lung biopsy All lung biopsies (n = 50) Biopsy yielded benefit Survival and treatment changed based on results 25 (50%) Change in antimicrobials 19 (38%) Change in immunosuppression 6 (12%) Biopsy did not yield benefit Survival and no change in treatment based on results 23 (46%) Death Treatment change based on biopsy results 1 (2%) No treatment change based on biopsy results 1 (2%) intubated and on mechanical ventilation postoperatively (4%). Two patients died of sepsis during their hospitalization; their deaths were not clearly related to surgery. As pneumothorax was common after removal of the thoracostomy tube, we attempted to identify factors associated with post-pull pneumothoraces (PPP) by comparing patients with and without PPP by sex, age, diagnosis, operative technique, chest tube duration, and operative time (Table IV; available at Patients with PPP were older ( vs ; P =.01). They also tended to have more than 1 biopsy specimen taken, although this difference did not reach statistical significance (38% in those with PPP vs 12% in those without PPP; P =.07). Discussion Diagnostic VATS lung biopsy in immunocompromised children has not previously been studied for diagnostic yield and outcomes. Furthermore, with the advent of effective antibiotic treatment for Pneumocystis carinii pneumonia, the potential diagnostic yield for a surgical lung biopsy has become more limited. In the current study, we demonstrate that in an era of minimally invasive surgery, surgical lung biopsy in the immunocompromised pediatric patient yields a histopathologic or culture diagnosis and results in benefit in up to 50% of patients. We also found that patients with preoperative fever were less likely to have a diagnostic biopsy, which may indicate a more robust immune response in this sub-group of patients. Taking more than 1 biopsy did not improve diagnostic yield. Table III. Surgical outcomes following lung biopsy All lung biopsies (n = 50) Operative time (min; avg SD) Duration of chest tube placement (d; avg SD) Morbidity Any complication 26 (52%) PPP 24 (48%) Resolved without intervention 21 (42%) Required reinsertion of chest tube 3 (6%) Prolonged air leak 1 (2%) Postoperative mechanical ventilation 2 (4%) Mortality 2 (4%) Open lung biopsy has previously been shown to yield a definitive diagnosis in 50-80% of pediatric patients In addition, open biopsy led to a change in empiric therapy 76%-90% of the time and was more likely to lead to a change compared with BAL. 4,9,11-14 A comparative retrospective study found that open lung biopsy has a diagnostic yield of 50%, compared with 38% for computed tomography guided biopsy, and 32% for BAL. 4 Our data suggests that the diagnostic yield has not suffered with the change in standard surgical approach and change in potential infectious etiologies. Even in the absence of a specific diagnosis, the information provided by open lung biopsy has been shown to make it possible to limit the differential diagnoses and to select more appropriate drug treatment. 11 Open lung biopsy in immunocompromised children has been shown to be a safe procedure, with a minor complication rate of 36% and major complication rate of 7%. 4 In the series presented here for which most patients underwent VATS lung biopsy, postoperative morbidities were common, with 48% of patients developing a pneumothorax after removal of the thoracostomy tube, or a PPP. The majority of these small pneumothoraces did not become clinically significant and likely represented either a closed parenchymal leak, air entrained during chest tube removal, or residual carbon dioxide following thoracoscopy. Although only three patients required replacement of the thoracotomy tube, these frequent PPP likely resulted in further radiograph examinations and possibly an increase in length of hospital stay. Patients with PPP were generally older. Due to the retrospective design of this study, it was not feasible to reliably abstract data regarding the technique used for thoracostomy tube removal or what level of trainee performed the removal. It should also be taken into consideration that patients with more than one biopsy taken had a trend toward a higher rate of PPP. One may consider therefore only taking one biopsy specimen, as taking more may predispose to a higher complication rate without resulting in an improved diagnostic yield. Although not a pediatric study, Zihlif et al retrospectively reviewed 62 immunocompromised adult patients who underwent surgical lung biopsy in the current era of Pneumocystis cariniil prophylaxis, utilizing VATS in half of their patients. 15 They found that biopsy led to a specific diagnosis in 60%, a change in therapy in 40%, and was associated with major complications in 11% of patients; these findings are similar to those described here in children. However, the mortality rate in their series was 40%. This higher mortality rate may reflect a higher acuity of the patients undergoing surgical lung biopsy; it is not possible, however, to ascertain this from the data provided in their study. We acknowledge there are limitations inherent to our study. The retrospective design predisposes to incomplete data, as the consistency with which data were recorded at the time of each patient s hospitalization likely varies. That said, the majority of the data analyzed in this study were objective data points recorded in a standard fashion, such as radiology reports, pathology reports, and medication administration records, minimizing the possible variability inherent in the data collection process. The Utility of Surgical Lung Biopsy in Immunocompromised Children 135

4 THE JOURNAL OF PEDIATRICS Vol. 162, No. 1 In conclusion, persistent pulmonary infiltrates in immunocompromised pediatric patients continue to present a diagnostic dilemma. Surgical lung biopsy may be undertaken with the understanding that this invasive procedure continues to lead to diagnosis in only half of patients with an equal proportion of patients experiencing complications. In the era of minimally invasive surgery and a different spectrum of infectious etiologies that may lead to persistent pulmonary infiltrates in immunocompromised children, outcomes and benefit derived from surgical lung biopsy have not changed. These findings may help inform discussions regarding the risk and benefits of this invasive procedure in the immunocompromised pediatric population. n Submitted for publication Mar 30, 2012; last revision received May 7, 2012; accepted Jun 7, Reprint requests: David H. Rothstein, MD, Department of Surgery, Ann and Robert H. Lurie Children s Hospital of Chicago, 225 East Chicago Ave, Box #63, Chicago, IL DRothstein@luriechildrens.org References 1. Chaoui D, Legrand O, Roche N, Cornet M, Lefebvre A, Peffault de Latour R, et al. Incidence and prognostic value of respiratory events in acute leukemia. Leukemia 2004;18: Ewig S, Glasmacher A, Ulrich B, Wilhelm K, Schafer H, Nachtsheim KH. Pulmonary infiltrates in neutropenic patients with acute leukemia during chemotherapy: outcome and prognostic factors. Chest 1998;114: Fanfulla F, Locatelli F, Zoia MC, Giorgiani G, Bonetti F, Spagnolatti L, et al. Pulmonary complications and respiratory function changes after bone marrow transplantation in children. Eur Respir J 1997;10: Armenian SH, La Via WV, Siegel SE, Mascarenhas L. Evaluation of persistent pulmonary infiltrates in pediatric oncology patients. Pediatr Blood Cancer 2007;48: Efrati O, Gonik U, Bielorai B, Modan-Moses D, Neumann Y, Szeinberg A, et al. Fiberoptic bronchoscopy and bronchoalveolar lavage for the evaluation of pulmonary disease in children with primary immunodeficiency and cancer. Pediatr Blood Cancer 2007;48: Snyder CL, Ramsay NK, McGlave PB, Ferrell KL, Leonard AS. Diagnostic open-lung biopsy after bone marrow transplantation. J Pediatr Surg 1990;25: Discussion Gluer S, Schwerk N, Reismann M, Metzelder ML, Nustede R, Ure BM, et al. Thoracoscopic biopsy in children with diffuse parenchymal lung disease. Pediatr Pulmonol 2008;43: Davies L, Dolgin S, Kattan M. Morbidity and mortality of open lung biopsy in children. Pediatrics 1997;99: Hayes-Jordan A, Benaim E, Richardson S, Joglar J, Srivastava DK, Bowman L, et al. Open lung biopsy in pediatric bone marrow transplant patients. J Pediatr Surg 2002;37: Mason WH, Siegel SE, Tucker BL. Diagnostic open lung biopsy for diffuse pulmonary disease in immunocompromised pediatric patients. Am J Pediatr Hematol Oncol 1982;4: Stefanutti D, Morais L, Fournet JC, Jan D, Casanova JL, Scheinmann P, et al. Value of open lung biopsy in immunocompromised children. J Pediatr 2000;137: Chuang ML, Lin IF, Tsai YH, Vintch JR, Pang LC. The utility of open lung biopsy in patients with diffuse pulmonary infiltrates as related to respiratory distress, its impact on decision making by urgent intervention, and the diagnostic accuracy based on the biopsy location. J Intensive Care Med 2003;18: Gulbahce HE, Pambuccian SE, Jessurun J, Woodard P, Steiner ME, Manivel JC, et al. Pulmonary nodular lesions in bone marrow transplant recipients: impact of histologic diagnosis on patient management and prognosis. Am J Clin Pathol 2004; 121: Kornecki A, Shemie SD. Open lung biopsy in children with respiratory failure. Crit Care Med 2001;29: Zihlif M, Khanchandani G, Ahmed HP, Soubani AO. Surgical lung biopsy in patients with hematological malignancy or hematopoietic stem cell transplantation and unexplained pulmonary infiltrates: improved outcome with specific diagnosis. Am J Hematol 2005; 78: Naiditch, Barsness, and Rothstein

5 January 2013 ORIGINAL ARTICLES Table IV. Factors associated with pneumothorax after chest tube removal following lung biopsy in immunocompromised children PPP (n = 24) No PPP (n = 26) P Female 8 (33.3%) 12 (46.1%).53 Age (y; mean SD) Primary diagnosis.68 Hematologic cancer 14 (58.4%) 18 (69.2%) Primary immunodeficiency 3 (11.5%) 3 (11.5%) Organ transplant recipient 4 (15.4%) 2 (7.7%) Solid cancer 2 (7.7%) 3 (11.5%) Chronic steroid use 1 (3.9%) 0 Stem cell transplant recipient 5 (19.2%) 7 (26.9%).86 Operative technique.66 Open 2 (8.3%) 2 (7.7%) VATS 22 (91.7%) 24 (92.3%) Number of biopsies taken (62.5%) 23 (88.5%) $2 9 (37.5%) 3 (11.5%) Chest tube duration Operative time PPP, post-pull pneumothorax. The Utility of Surgical Lung Biopsy in Immunocompromised Children 136.e1