Importance of Open Lung Biopsy in the Diagnosis of Invasive Pulmonary Aspergillosis in Patients With Hematologic Malignancies

Transcription

1 American Journal of Hematology 71:75 79 (2002) Importance of Open Lung Biopsy in the Diagnosis of Invasive Pulmonary Aspergillosis in Patients With Hematologic Malignancies Kihyun Kim, 1 Mark H. Lee, 1 * Jingook Kim, 2 Kyung Soo Lee, 3 Sung Min Kim, 1 Man Pyo Jung, 1 Joungho Han, 4 Ki Woong Sung, 5 Won Seog Kim, 1 Chul Won Jung, 1 Sung Soo Yoon, 1 Young-Hyuck Im, 1 Won Ki Kang, 1 Keunchil Park, 1 and Chan Hyung Park 1 1 Department of Medicine, Sungkyunkwan University School of Medicine, Samsung Medical Center, Seoul, Korea 2 Department of Thoracic Surgery, Sungkyunkwan University School of Medicine, Samsung Medical Center, Seoul, Korea 3 Department of Radiology, Sungkyunkwan University School of Medicine, Samsung Medical Center, Seoul, Korea 4 Department of Diagnostic Pathology, Sungkyunkwan University School of Medicine, Samsung Medical Center, Seoul, Korea 5 Department of Pediatrics, Sungkyunkwan University School of Medicine, Samsung Medical Center, Seoul, Korea Invasive pulmonary aspergillosis is a serious infectious complication in immunocompromised patients. Recent reports indicate its favorable clinical outcomes by early diagnosis with chest computed tomography scan. We retrospectively analyzed our experiences with histopathological evaluation by open lung biopsy in 31 patients (32 cases) with hematologic malignancies, suspected of having invasive pulmonary aspergillosis clinically and radiologically. Although the initial computed tomography findings of all cases were highly indicative of invasive pulmonary aspergillosis by demonstrating nodules or masses with a halo sign (16 cases), segmental area of consolidation with ground-glass attenuation (7 cases), both nodules or masses with a halo sign and segmental areaof consolidation with ground-glass attenuation (7 cases) and poorly defined centrilobular nodules (2 cases), we could histopathologically confirm invasive fungal infections only in 17 cases (53.1%) by open lung biopsy. There were 13 cases of invasive pulmonary aspergillosis, two cases of aspergilloma, and two cases of mucormycosis. No fungal hyphae were found in the other 15 cases: organizing pneumonia in seven cases, pulmonary hemorrhage in three cases, brochiolitis obliterans with organizing pneumonia in two cases, and CMV pneumonia, pulmonary tuberculosis, candida pneumonia in one case each, respectively. We could perform open lung biopsy without mortality and significant morbidity. In view of the low positive predictive value of chest computed tomography scan and the very low morbidity of open lung biopsy, this procedure is recommendable for the diagnosis of invasive pulmonary aspergillosis and determination of its treatment. Am. J. Hematol. 71:75 79, Wiley-Liss, Inc. Key words: aspergillosis; open lung biopsy; computed tomography INTRODUCTION Invasive pulmonary aspergillosis (IPA) is a serious infectious complication in patients with hematologic malignancies. The incidence of aspergillosis is increasing due to widespread use of intensive chemotherapy and immunosuppressive agents and due to increasing numbers of organ transplantations and AIDS patients [1]. The mortality rate of IPA in patients with neutropenia resulting from chemotherapy for leukemia was %, and moreover, it reached up to % in bone marrow transplantation (BMT) patients in recent reports of metaanalysis [2,3] Wiley-Liss, Inc. Recently, there are several studies indicating that early detection of IPA with computed tomography (CT) scan and aggressive medical treatment in combination with surgery improved the survival in these patients [4 9]. *Correspondence to: Mark H. Lee, M.D., Division of Hematology- Oncology, Department of Medicine, Sungkyunkwan University School of Medicine, Samsung Medical Center, 50 Ilwon-Dong, Kangnam-Ku, Seoul , Korea. mlee@smc.samsung.co.kr Received for publication 15 March 2002; Accepted 15 May 2002 Published online in Wiley InterScience ( com). DOI: /ajh.10168

2 76 Kim et al. A halo sign on the CT scan, nodule surrounded by an area of ground-glass attenuation, is an important indicator for diagnosing IPA in neutropenic patients and it appears during the early stage of IPA [10,11]. Aircrescent sign is also a key indicator of IPA but less useful for early diagnosis as it usually appears with bone marrow recovery [12]. Although some surgical series reported a high positive predictive value of CT scan in the diagnosis of IPA [4 9], a preliminary study from our center showed a low positive predictive value [13]. Here, we evaluated our further experiences with histopathological confirmation by open lung biopsy in patients suspected of IPA on thin-sectioned CT scan. PATIENTS AND METHODS Patients Between May 1995 and January 2001, 325 patients with hematologic malignancies including severe aplastic anemia were treated in our institution. Among them, 74 and 54 patients underwent allogeneic BMT and autologous peripheral blood stem cell transplantation (PBSCT), respectively. Chest CT scans for the evaluation of lung infiltrates during the period of neutropenic fever or post- BMT immunosuppression were performed for 114 patients. The primary radiologic diagnoses were as follows: IPA for 51 patients (52 cases), bacterial pneumonia for 22 patients, cytomegalovirus (CMV) or Pneumocystis carinii pneumonia for 19 patients, brochiolitis obliterans with organizing pneumonia (BOOP) for 7 patients, pulmonary tuberculosis for 7 patients, and other diagnoses for 8 patients. Among 51 patients with hematologic malignancies suspected of having IPA on thin-sectioned CT scan, open lung biopsy was performed for 31 patients (32 cases) to confirm the diagnosis. Two separate episodes occurred in one patient at a 7-month interval, and biopsies were done at each time. The reasons why open lung biopsy was not performed for the other 20 patients were as follows: 7 expired before biopsy due to underlying disease progression or other causes, 6 were medically treated without pathological confirmation according to the attending physicians preference, 3 refused open lung biopsy, 2 developed Aspergillus rhinosinusitis simultaneously, and Aspergillus species were detected in sputum culture for other 2 patients. All patients were nursed in the hematology-oncology wards of Samsung Medical Center built in October The wards had ducted ventilation facilities with a ventilating fan producing positive air flow to each room. The volume of air ventilating each room per hour was 3 times as much as the volume of the room, and the air ventilating the room was exhausted without further circulation. The air pressure among the rooms was equal. The inflow air was filtered twice by dust filter. The elimination efficiency of dust in the atmosphere by dust filters was 90% or more. There was extension work, which could provoke an epidemic of IPA, done to the main building adjacent to the wards from July 1997 through October Diagnosis and Treatment Strategy When a febrile neutropenic episode occurred in patients with hematologic malignancies, broad-spectrum antibiotics (usually -lactam plus aminoglycoside) were administered intravenously as an empirical treatment. Severe neutropenia was defined as absolute neutrophil count (ANC) less than 500/mm 3. If fever persisted for 48 hrs or more, vancomycin was added intravenously. Amphotericin B was added empirically (0.5 mg/kg/day) when fever persisted for another 48-hr period after the beginning of vancomycin. When pulmonic lesions on chest roentgenogram showed no improvements with empirical antibiotics or if pulmonary fungal infections were radiologically suspected, a chest high-resolution CT (HRCT) scan was planned. CT scan technique used has been previously described (1-mm thin section at 10-mm interval during full inspiration) [13]. On CT scan, nodules or masses with halo sign, air-crescent sign, and consolidation with ground-glass attenuation were considered to be lesions indicative of IPA. Once HRCT scan revealed findings consistent with IPA, the dose of amphotericin B was increased to mg/kg/day. A thoracotomy or a video-assisted thoracoscopic surgery was performed when feasible, depending on the patient s clinical situation for pathological diagnosis. Although open lung biopsy was primarily planned for pathological diagnosis, complete resection was done if possible. Sputum panculture and smear were done in all the patients, but a bronchoscopy with bronchoalveolar lavage (BAL) was not routinely planned. In addition to routine hematoxylin and eosin (H&E) staining on the surgical specimens, para-aminosalicylic acid (PAS) and methenamine silver staining were done to detect fungal hyphae, and fungus culture was done for all surgical specimens. Statistical Analysis The positive predictive value of CT scan was calculated by the percentage of IPA cases including aspergilloma and mucormycosis confirmed by histopathology divided by total open lung biopsy cases. The patient characteristics and clinical findings between group A and group B were compared by Fisher s exact test and Mann Whitney test. Prognostic factors associated with clinical outcome were analyzed by Fisher s exact test, and survival curves were plotted according to the Kaplan Meier method.

3 Open Lung Biopsy for Invasive Pulmonary Aspergillosis 77 TABLE I. Patient Characteristics a RESULTS Patient Characteristics Among 51 patients with hematologic malignancies suspected of having IPA on thin-sectioned CT scan, open lung biopsy was performed for 31 patients (32 cases) to confirm the diagnosis as described. Among those undergoing open lung biopsy, 16 patients (17 cases) were confirmed to have invasive pulmonary fungal infection. Patient characteristics of the histopathologically confirmed invasive pulmonary fungal infection cases (group A) and the false-positive cases (group B) which were indicative of IPA on HRCT but histologically negative for IPA, were detailed in Table I. There was no statistically significant difference between the two groups (P > 0.05). Among 63 patients whose primary radiologic diagnoses were not IPA, three subsequently developed pulmonic fungal infection. Two of them were diagnosed to have Aspergillus and mucormycosis tracheobronchitis by endobronchial biopsy of their pseudomembrane during brochoscopy. Brochoscopy was performed 15 and 58 days after the initial CT scans due to persistence of centrilobular nodules on follow-up CT scans in both cases. The third case in immunosuppression due to chronic GVHD was radiologically diagnosed to have pulmonary tuberculosis by CT scan but proved to be chronic IPA after fine-needle aspiration of the lung lesion, which was performed 3 days after the CT scan. Clinical and Radiological Features Group A (N 17) b Group B (N 15) Age (years, range) 40 (14 58) 35 (22 58) Male/Female 12/5 12/3 Diagnosis Acute leukemia NHL 1 1 CML 1 MDS 1 1 Treatment Chemotherapy Allo-BMT 3 2 Steroid (>1 mg/kg/day) 4 2 Duration of Neutropenia (days, range) 24 (0 43) 18 (0 59) a Group A, histologically confirmed IPA cases; Group B, false-positive group (the cases indicative of IPA on HRCT but histopathologically negative for IPA). Abbreviations: NHL, non-hodgkin lymphoma; CML, chronic myeloid leukemia; MDS, myelodysplastic syndromes; allo-bmt, allogeneic bone marrow transplantation. b A patient who has had two episodes was counted twice. The most common initial symptoms were fever and cough. Blood-tinged sputum and pleuritic chest pain were more common in group A than in group B without statistical significance (8 and 9 cases vs. both 3 cases, P and 0.076). The median duration from the initial symptom to the radiologic diagnosis was 12 days in group A and 7 days in group B. At the time of the radiological diagnosis, 10 cases in group A and 12 cases in group B had severe neutropenia. The median duration from severe neutropenia to the radiologic diagnosis was 16 days (range, 0 36) in group A and 11 days (range, 0 30) for group B. Median values of fibrinogen in group A and B were 5.63 g/l (range, ) and 5.52 g/l (range, ), respectively, without significant difference (P 0.90). The initial CT findings of 17 cases in group A were as follows: (i) nodules or masses with halo sign in 9 cases; (ii) segmental area of consolidation with ground-glass attenuation in 4 cases; (iii) both segmental area of consolidation with ground-glass attenuation and nodules or masses with a halo sign in 2 cases; and (iv) poorly defined centrilobular nodules in 2 cases (Fig. 1). Those of 15 cases in group B were as follows: (i) nodules or masses with a halo sign in 7 cases; (ii) segmental area of consolidation with ground-glass attenuation in 3 cases; and (iii) both segmental area of consolidation with ground-glass attenuation and nodules or masses with a halo sign in 5 cases. In group A, 11 cases showed multiple lesions and 6 cases showed single lesions. In group B, 11 and 2 cases demonstrated multiple lesions and single lesions, respectively. Cavitary lesions were detected from the follow-up CT scans in 3 cases in group A and 1 case in group B. These radiological findings were not statistically different between the two groups (P > 0.05). Open Lung Biopsy and Histopathologic Findings Open lung biopsies were done in the 32 cases on 31 patients. Three of them were video-assisted thoracic surgery (VATS). The timing of open lung biopsy was dependent on the patient s condition. There was no emergency case for the treatment of hemoptysis. Although surgical procedures were primarily performed for pathological diagnosis, complete resection of fungal lesion was possible in 6 of the 8 cases with single lesions. Except for 1 case of lobectomy, most of the surgical procedures were segmental or subsegmental wedge resections. The interval from the radiologic diagnosis to the surgery was 9 days (range, 1 37) in group A and 12 days (range, 1 29) in group B (P > 0.05). Open lung biopsies were done during severe neutropenic status (ANC < 500/ mm 3 ) in 9 cases and severe thrombocytopenia (platelet count <20,000/mm 3 ) in 2 cases with platelet transfusion. The median amount of transfused packed red blood cells and platelets concentrate was 1 pint (range, 0 6) and 12 units (range, 0 40), respectively. Most of the complications were minor problems such as pain and minor

4 78 Kim et al. Fig. 1. Initial CT findings that suggest invasive pulmonary aspergillosis: (left) nodule with halo sign; (middle) segmental area of consolidation with groundglass attenuation; and (right) poorly defined centrilobular nodule. All of these lesions were histopathologically proven to be invasive aspergillosis. wound problems; however, one patient experienced cardiac arrest with subsequent acute renal failure on the day of the surgery but recovered fully without any sequelae. Only one case of Aspergillus flavus was found in sputum pan-culture, which was done in all the patients. Fungal hyphae were found in 17 of the 32 cases, and the histopathologic diagnoses were as follows (group A): IPA in 13 cases, aspergilloma in 2 cases, and invasive pulmonary mucormycosis in 2 cases. From the cultures of surgical specimens, 1 case of Aspergillus fumigatus and 1 case of Aspergillus species were obtained. On the contrary, we could not find any fungal hyphae in the other 15 cases, and their histopathologic diagnoses were as follows (group B): organizing pneumonia in 7 cases, pulmonary hemorrhage in 3 cases, BOOP in 2 cases, and CMV pneumonia, pulmonary tuberculosis, and candida pneumonia (neutrophilic abscess with yeast) in one case each, respectively. Thus the positive predictive value of chest HRCT scan in the diagnosis of fungal infections was 53.1% in this study. In addition, brochoscopy with BAL fluid analysis including cytology, mycoscopy, and fungus culture was done in eight of 51 patients radiologically suspected of having IPA; however, none of 8 showed the evidence of fungal infection. DISCUSSION As Aspergillus is ubiquitous in the air and environment, isolation with high-efficiency particulate air filtration or laminar air flow has been the only way to reduce the risk of aspergillosis [1]. Itraconazole prophylaxis has been eagerly tried, but erratic absorption of oral form has limited its routine use. Although itraconazole solution with better pharmacokinetic profile was reported effective for aspergillosis prophylaxis, it has not yet been firmly established [14]. At the present day, early diagnosis with aggressive treatment is the best way to reduce mortality [1]. IPA is suspected when fever persists with pulmonary infiltrates in immunocompromised patients in spite of the administration of broad-spectrum antibiotics. If nodules with halo sign or air-crescent sign can be detected, the possibility of IPA becomes even higher. The strategy for early detection of IPA using chest CT scan has been tried successfully [12]. The standard treatment of IPA has been the administration of amphotericin B with or without itraconazole. Several recent studies of surgical approaches in combination with medical treatment were encouraging in selective cases with favorable results [4 9]. Although no randomized prospective study has been reported, they suggested that surgical approach could be helpful in certain circumstances. In these reports, excluding emergency operation for massive hemoptysis, the objectives of surgery were mainly for either resection of the localized masses to prevent massive pulmonary hemorrhage that might be fatal or eradication of residual fungal foci prior to further cytotoxic chemotherapy or BMT. These surgical series reported favorable results with minor surgical complications in spite of neutropenia or thrombocytopenia in these patients. In these reports the histopathologic diagnosis revealed a high positive predictive value of chest CT scan ranging from 81.5% to 100% to confirm the preoperative diagnosis. On the contrary, our previously published data reporting a positive predictive value of 58.3% [13] were so different from those of other studies that we pursued a new strategy to confirm the histopathologic diagnosis for all patients with suspected IPA. We performed open lung biopsy in 32 cases on 31 patients who were clinically and radiologically indicative of having IPA. In this study only 17 of the 32 cases (53.1%, 95% confidence interval %) had IPA including two cases with mucormycosis, essentially identical to our previous rate of 58.3%. Although this study is small in terms of patient size, our cases are less selective than those of other surgical series in which most of the enrolled patients were

5 Open Lung Biopsy for Invasive Pulmonary Aspergillosis 79 selected not for diagnostic purposes but for therapeutic purposes. Although open lung biopsies were done during severe neutropenic status and severe thrombocytopenia, no serious complication occurred in our study. Open lung biopsy is a safe and tolerable procedure in the diagnosis of IPA. Earlier histopathologic confirmation by open lung biopsy is important in determining whether patients continue to be treated, especially when we keep in mind the high toxicity profile of amphotericin B for patients who are scheduled for further cytotoxic chemotherapy or BMT in the future. Although open lung biopsy was safely performed in our study, more convenient diagnostic methods for IPA should be explored. As described in other reports [1,4], fever, pleuritic chest pain, and hemoptysis were common symptoms for IPA patients and there was a trend for higher incidences of pleuritic chest pain in patients with IPA compared to patients without IPA (P 0.076). Sputum smear and culture are neither sensitive nor specific, and serum antigenemia test has a low sensitivity [15]. Fibrinogen level was correlated with diagnosis and prognosis in a study, but we could not confirm the results [4]. Although we did not perform bronchoscopy with BAL routinely as it has variable sensitivity but a high specificity, recent reports showed that antigenemia test with BAL fluid could enhance the sensitivity [4,15]. CTguided percutaneous fine needle aspiration and biopsy can be a useful diagnostic method for peripheral lesions but it can hardly be performed for central lesions or for patients with severe thrombocytopenia or coagulopathy [16,17]. Recently new techniques based on PCR or sandwich ELISA for the detection of antigen such as Aspergillus galactomannan has been reported and would be helpful [15,18 20]. In conclusion, in view of low positive predictive value of chest CT scan and very low morbidity of open lung biopsy, this procedure is recommendable for the diagnosis of IPA and determination of its treatment. REFERENCES 1. Denning DW. Invasive aspergillosis. Clin Infect Dis 1998;26: Denning DW. Therapeutic outcome in invasive aspergillosis. Clin Infect Dis 1996;23: Lin S-J, Schranz J, Teutsch SM. Aspergillosis case fatality rate: systemic review of the literature. Clin Infect Dis 2001;32: Caillot D, Casasnovas O, Bernard A, Couaillier JF, Durand C, Cuisenier B, Solary E, Piard F, Petrella T, Bonnin A, Couillault G, Dumas M, Guy H. Improved management of invasive pulmonary aspergillosis in neutropenic patients using early thoracic computed tomographic scan and surgery. J Clin Oncol 1997;5: Yeghen T, Kibbler CC, Prentice HG, Berger LA, Wallesby RK, McWhinney PH, Lampe FC, Gillespie S. Management of invasive pulmonary aspergillosis in hematology patients: a review of 87 consecutive cases at a single institution. Clin Infect Dis 2000;31: Wong K, Waters CM, Walesby RK. Surgical management of invasive pulmonary aspergillosis in immunocompromised patients. Eur J Cardiothorac Surg 1992;6: Reichenberger F, Habicht J, Kaim A, Dalquen P, Bernet F, Schlapfer R, Stulz P, Perruchoud AP, Tichelli A, Gratwohl A, Tamm M. Lung resection for invasive pulmonary aspergillosis in neutropenic patients with hematologic diseases. Am J Respir Crit Care Med 1998;158: Salerno CT, Ouyang DW, Pederson TS, Larson DM, Shake JP, Johnson EM, Maddaus MA. Surgical therapy for pulmonary aspergillosis in immunocompromised patients. Ann Thorac Surg 1998;65: Pidhorecky I, Urschel J, Anderson T. Resection of invasive pulmonary aspergillosis in immunocompromised patients. Ann Surg Oncol 2000; 7: Kuhlman JE, Fishman EK, Siegelman SS. Invasive pulmonary aspergillosis in acute leukemia: characteristic findings on CT, the CT halo sign, and the role of CT in early diagnosis. Radiology 1985;157: Caillot D, Couaillier JF, Bernard A, Casasnovas O, Denning DW, Mannone L, Lopez J, Couillault G, Piard F, Vagner O, Guy H. Increasing volume and changing characteristics of invasive pulmonary aspergillosis on sequential thoracic computed tomography scans in patients with neutropenia. J Clin Oncol 2001;19: Blum U, Windfuhr M, Buitrago-Tellez C, Sigmund G, Herbst EW, Langer M. Invasive pulmonary aspergillosis. MRI, CT, and plain radiographic findings and their contribution for early diagnosis. Chest 1994;106: Won HJ, Lee KS, Cheon JE, Hwang JH, Kim TS, Lee HG, Han J. Invasive pulmonary aspergillosis: prediction at thin-section CT in patients with neutropenia: a prospective study. Radiology 1998;208: Morgenstern GR, Prentice AG, Prentice HG, Ropner JE, Schey SA, Warnock DW on behalf of the U.K. Multicentre Antifungal Prophylaxis Study Group. A randomized controlled trial of itraconazole versus fluconazole for the prevention of fungal infections in patients with haematological malignancies. Br J Hematol 1999;105: Latge J-P. Aspergillus fumigatus and aspergillosis. Clin Microbiol Rev 1999;12: Hwang S, Kim H, Park S, Jung J, Jang H. The value of CT-guided percutaneous needle aspiration in immunocompromised patients with suspected pulmonary infection. Am J Radiol 2000;175: Denining D, Evans E, Kibbler C, Richardson M, Roberts M, Rogers T, Warnock D, Warren R. Guidelines for the investigation of invasive fungal infection in hematologic malignancy and solid organ transplantation. Eur J Clin Microbiol Infect Dis 1997;16: Hebart H, Loffler J, Meisner C, Serey F, Schimid D, Bohme A, Martin H, Engel A, Bunjes D, Kern WV, Schumacher U, Kanz L, Einsele H. Early detection of Aspergillus infection after allogenic stem cell transplantation by polymerase chain reaction screening. J Infect Dis 2000; 181: Maertens J, Verhaegen J, Demuynck H, Brock P, Verhoef G, Vandenverghe P, Van Eldere J, Verbist L, Boogaerts M. Autopsy-controlled prospective evaluation of serial screening for circulating galactomannan by sandwich enzyme-linked immunosorbent assay for hematologic patients at risk for invasive aspergillosis. J Clin Microbiol 1999; 37: Maertens J, Verhaegen J, Lagrou K, Van Eldere J, Boogaerts M. Screening for circulating galactomannan as a noninvasive diagnostic tool for invasive aspergillosis in prolonged neuropenic patients and stem cell transplantation recipients: a prospective validation. Blood 2001;97: