Prognostic Significance of Histopathologic Subsets in Idiopathic Pulmonary Fibrosis

Transcription

1 Prognostic Significance of Histopathologic Subsets in Idiopathic Pulmonary Fibrosis JULIE A. BJORAKER, JAY H. RYU, MARK K. EDWIN, JEFFREY L. MYERS, HENRY D. TAZELAAR, DARRELL R. SCHROEDER, and KENNETH P. OFFORD Division of Pulmonary and Critical Care Medicine, Mayo Graduate School of Medicine, Division of Anatomic Pathology, and the Section of Biostatistics, Mayo Clinic and Mayo Foundation, Rochester, Minnesota Idiopathic pulmonary fibrosis (IPF) is a generally fatal disorder with a reported median survival of 3 to 6 yr. This has been based on relatively few studies with diagnoses inconsistently confirmed by adequate lung biopsy. Retrospective analysis of 104 patients with IPF who had open lung biopsy (OLB) at Mayo Medical Center from 1976 to 1985 was performed to establish the overall survival rate, the spectrum of histopathological subgroups and their associated prognostic significance. The study group consisted of 54 men and 50 women with a median age of 63 yr. Median survival was 3.8 yr after diagnosis by OLB with an estimated 10 yr survival of 27%. Current histopathologic review showed a heterogeneous group including usual interstitial pneumonia (UIP), desquamative interstitial pneumonia (DIP), nonspecific interstitial pneumonia/fibrosis (NSIP), acute interstitial pneumonia (AIP), bronchiolitis, bronchiolitis obliterans organizing pneumonia (BOOP), and others. Median survival of the UIP group was 2.8 yr which is significantly worse (p 0.001) than for other subgroups of chronic interstitial pneumonias. IPF includes several histopathologic subgroups with significantly different survival rates. Patients with UIP have worse survival than patients with other types of idiopathic chronic interstitial pneumonias including NSIP. Accurate histopathologic classification is essential for prognostication in patients with IPF. Bjoraker JA, Ryu JH, Edwin MK, Myers JL, Tazelaar HD, Schroeder DR, Offord KP. Prognostic significance of histopathologic subsets in idiopathic pulmonary fibrosis. AM J RESPIR CRIT CARE MED 1998;157: (Received in original form April 29, 1997 and in revised form August 8, 1997) Correspondence and requests for reprints should be addressed to Dr. Jay H. Ryu, Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Mayo Clinic, Rochester, MN Am J Respir Crit Care Med Vol 157. pp , 1998 Idiopathic pulmonary fibrosis (IPF), also called cryptogenic fibrosing alveolitis, is a clinicopathologic term referring to a generally fatal disorder of unknown cause characterized by varying degrees of inflammation and fibrosis in the lung parenchyma (1 5). Mean survival in IPF has been estimated to be 3 to 6 yr but with a variable clinical course (2, 6 10). This estimate has been based on relatively few studies with diagnoses inconsistently established by adequate, i.e., open or thoracoscopic, lung biopsy (2, 6, 8, 9). Pathologic review of lung specimens from IPF patients may show a variety of histologic patterns (6 8, 10, 11). Usual interstitial pneumonia (UIP) is a specific histologic pattern of fibrosing interstitial pneumonia seen in the majority of patients with IPF. In this sense, UIP and IPF are considered synonyms. Over the past 15 yr, several lung injury patterns have been delineated as potential mimics of IPF. Bronchiolitis obliterans organizing pneumonia (BOOP) (12), acute interstitial pneumonia (AIP) (13), and respiratory bronchiolitis associated interstitial lung disease (RB/ILD) (14) have been separated as unique clinicopathologic subgroups with a distinctly different natural history than IPF. Recently, Katzenstein and Fiorelli (15) described a pattern of chronic interstitial pneumonia referred to as nonspecific interstitial pneumonia/fibrosis (NSIP) in patients with a variety of clinical syndromes and with histology distinct from UIP. With the exception of AIP, these other entities are associated with a greater likelihood of responding to corticosteroid therapy and better prognosis than UIP. We reviewed our experience over a 10-yr period to further investigate the significance of histopathologic findings in patients with IPF. METHODS Patient Selection Surgical and medical index information on 152 patients previously diagnosed with IPF was linked to that of 479 patients who underwent open lung biopsy (OLB) at Mayo Medical Center over a 10-yr period. The years were chosen to allow for adequate follow-up. Medical records were reviewed and patients with evidence of connective tissue disease or environmental exposures known to cause interstitial lung disease were excluded. There were 104 patients with IPF who met the study criteria. These patients manifested clinical, physiologic, and radiographic features of IPF as previously described (1, 2). Data Collection The medical records were reviewed for information related to the patient s age, sex, smoking history, occupation, symptoms, signs, environmental exposures, associated disorders, medications, laboratory tests, pulmonary function tests, arterial blood gas studies, treatment, and outcome. The date of OLB at the Mayo Medical Center was used as the date of diagnosis. A mailed survey was used to obtain vital sta-

2 200 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL tus information for all not known to be deceased as of April 15, Date of death was obtained from death certificates or next of kin. Pathologic Review All 104 patients had previously undergone OLB and had the diagnosis of IPF made based on the biopsy results. Pathologic material was available for current rereview from 102 of the 104 patients (98%) and was reviewed by two of us (H.D.T., J.L.M.) in a blinded fashion and classified into five histopathologic categories: UIP, desquamative interstitial pneumonia (DIP), NSIP, AIP, and others. Usual interstitial pneumonia (7) is characterized by a heterogeneous appearance at low magnification with alternating zones of normal lung, collagen fibrosis, and honeycomb change. Scattered foci of fibroblast proliferation are also usually present. The fibrotic changes preferentially affect subpleural and paraseptal parenchyma. Patchy chronic inflammation accompanies the interstitial fibrosis, and consists mainly of lymphocytes, plasma cells, and histiocytes with a minor component of neutrophils and eosinophils. Desquamative interstitial pneumonia (7) is less variegated than UIP. Alveolar septa are thickened in a more uniform fashion, by a combination of connective tissue and an inflammatory infiltrate that often includes plasma cells and occasional eosinophils. The most striking feature is the presence of numerous pigmented macrophages within affected airspaces. Nonspecific interstitial pneumonia/ fibrosis (15) represents a pattern of chronic interstitial pneumonia that lacks characteristic features of other specific entities such as UIP, DIP, hypersensitivity pneumonitis, BOOP, Langerhans cell granulomatosis, or chronic eosinophilic pneumonia. Lung biopsies in this group are characterized by varying proportions of chronic interstitial inflammation and fibrosis which is temporally uniform (i.e., the process appears to have occurred over a single time span). This is in sharp contrast to the temporally heterogeneous pattern of UIP, related in part to an admixture of recent organization ( fibroblast foci ) and old collagen-rich scars. In the past such cases have been reported as cellular interstitial pneumonia, not otherwise specified (16, 17). Acute interstitial pneumonia (13), also referred to as Hamman-Rich syndrome, is characterized histologically by changes typical of acute and organizing diffuse alveolar damage. Remaining cases included several unrelated entities such as chronic eosinophilic pneumonia, hypersensitivity pneumonitis, nonspecific chronic bronchiolitis, RB/ILD, and BOOP. Statistical Analysis Cumulative survival probabilities were estimated using the Kaplan- Meier method (18) and compared with the general life expectancy for the white population of the West North Central region of the United States (19) for persons of like age, sex, and calendar year of birth using a one-sample log-rank test (20). The log-rank test (21) was used to compare survival of groups of patients. Three-year expected survival probability was determined for each patient (based on age, sex, and calendar year of birth using West North Central population survival tables), and a Cox proportional hazards regression model was used to compare survival between histopathologic subgroups after adjusting for 3-yr expected survival probability. In all cases two-sided tests were used with p values 0.05 used as evidence of findings not attributable to chance. RESULTS Table 1 shows the characteristics of the entire IPF group at the time of diagnosis. There were 54 men and 50 women with a median age of 63 yr. Forty-two percent had never smoked. Most patients had experienced exertional dyspnea (90%) and cough (81%) with a median duration of 6 and 5 mo, respectively, prior to OLB. Twenty-two percent had been treated with corticosteroids prior to lung biopsy. On exam, 80% had inspiratory crackles and 25% displayed digital clubbing. Parenchymal changes on chest radiograph were present in 98%. Most patients had a restrictive pattern on pulmonary function testing and exercise-induced desaturation. Following OLB, 89% were treated with prednisone. TABLE 1 CLINICAL FINDINGS AT THE TIME OF DIAGNOSIS IN 104 PATIENTS WITH IDIOPATHIC PULMONARY FIBROSIS Characteristic* % n Median Mean SD Range Sex Male Female Age, yr Smoking status Current Former Never Symptoms Exertional dyspnea (n 97) Duration, mo Cough (n 97) Duration, mo Signs Crackles Clubbing (n 100) Radiographic infiltrates (n 103) Pulmonary function (n 84) Normal 2 2 Obstructed Restricted Mixed 6 5 Nonspecific 6 5 Oxygen saturation, % Rest (n 88) Exercise (n 55) Corticosteroid treatment At the time of diagnosis Immediately following diagnosis * Number of patients with information available is 104, except where indicated. The histopathologic diagnoses and number of cases in each following rereview of the OLB slides are shown in Table 2. Only 62% of the cases were found to have UIP. The next most common diagnosis was NSIP. For the purpose of analyzing the prognostic significance, the cases were divided into three main subgroups: UIP, NSIP, and Others. The Others subgroup consisted of several chronic interstitial lung diseases (DIP, bronchiolitis, BOOP, RB/ILD, chronic eosinophilic pneumonia, hypersensitivity pneumonia) that have historically been associated with a higher response rate to corticosteroid therapy and a TABLE 2 HISTOPATHOLOGIC DIAGNOSES IN 102 PATIENTS* WITH IDIOPATHIC PULMONARY FIBROSIS Histopathological Diagnosis % n Usual interstitial pneumonia Nonspecific interstitial pneumonia/fibrosis Desquamative interstitial pneumonia 8 8 Bronchiolitis 4 4 Bronchiolitis obliterans organizing pneumonia 4 4 Respiratory bronchiolitis associated interstitial lung disease 2 2 Chronic eosinophilic pneumonia 1 1 Hypersensitivity pneumonitis 1 1 Acute interstitial pneumonia 2 2 Honeycomb change only 1 1 Hemorrhage 1 1 Scarring and pneumonia 1 1 * A total of 104 patients were diagnosed with IPF via open lung biopsy at the Mayo Medical Center during the 10-yr period from 1976 through Pathologic material was available for current histopathologic review from 102 of these patients.

3 Bjoraker, Rhu, Edwin, et al.: Prognosis in Pulmonary Fibrosis 201 TABLE 3 CLINICAL CHARACTERISTICS ACCORDING TO HISTOPATHOLOGIC SUBGROUP Characteristic UIP (n 63)* NSIP Others (n 14) (n 20) p Value Sex, % Male Female Age, yr Mean SD Range Smoking status, % Current Former Never Symptoms Exertional dyspnea, % NS Duration, mo (mean SD) NS Cough, % NS Duration, mo (mean SD) NS Signs Crackles, % Clubbing, % NS Radiographic infiltrates, % NS Oxygen saturation, % Rest (mean SD) NS Exercise (mean SD) NS Corticosteroid treatment, % At the time of diagnosis NS Immediately following diagnosis NS * Usual interstitial pneumonia. n 63 for all characteristics except: exertional dyspnea (n 57); cough (n 60); clubbing (n 62); radiographic infiltrates (n 62); oxygen saturation at rest (n 52); and oxygen saturation with exercise (n 35). Nonspecific interstitial pneumonia/fibrosis. n 14 for all characteristics except: cough (n 13); oxygen saturation at rest (n 11); and oxygen saturation with exercise (n 7). Includes desquamative interstitial pneumonia (n 8), bronchiolitis (n 4), bronchiolitis obliterans organizing pneumonia (n 4), respiratory bronchiolitis associated interstitial lung disease (n 2), chronic eosinophilic pneumonia (n 1), and hypersensitivity pneumonitis (n 1). Overall n 20 for all characteristics except: exertional dyspnea (n 19); cough (n 17); clubbing (n 18); oxygen saturation at rest (n 18); and oxygen saturation with exercise (n 12). p Value associated with the overall comparison across histopathologic subgroups using analysis of variance for continuous variables and exact test for categorical variables. NS not significant (p 0.05). better prognosis than UIP. Two patients with AIP, three patients with nondiagnostic histopathologic findings (honeycomb change only, hemorrhage, scarring and pneumonia, respectively), and two patients whose biopsy slides were unobtainable for current histopathologic review were excluded from this analysis. Table 3 describes the clinical characteristics of the three main histopathological subgroups, UIP, NSIP, and Others at Characteristic TABLE 4 PULMONARY FUNCTION ACCORDING TO HISTOPATHOLOGIC SUBGROUP UIP NSIP Others* n Mean SD n Mean SD n Mean SD p Value TLC, % pred FVC, % pred NS FEV 1, % pred NS DL CO, % pred NS Definition of abbreviations: UIP usual interstitial pneumonia; NSIP nonspecific interstitial pneumonia/fibrosis; DL CO diffusing capacity of the lungs for carbon monoxide. * Includes desquamative interstitial pneumonia (n 8), bronchiolitis (n 4), bronchiolitis obliterans organizing pneumonia (n 4), respiratory bronchiolitis associated interstitial lung disease (n 2), chronic eosinophilic pneumonia (n 1), and hypersensitivity pneumonitis (n 1). Number of patients with information available for the given characteristic. p Value associated with the overall comparison across histopathologic subgroups using analysis of vaiance. NS not significant (p 0.05). the time of the OLB-based diagnosis of IPF. Patients with UIP were older with a mean of 65 yr compared with 57 yr and 54 yr in the NSIP and Others subgroups, respectively. Current smoking was more frequent in the Others group likely due to cases of bronchiolitis, RB/ILD, and DIP (14). Table 4 outlines the pulmonary function data obtained for the three subgroups. UIP patients had a significantly lower total lung capacity (% predicted normal) compared with the Others category. There were no other significant differences in other physiologic parameters including arterial blood gas results (data not shown). Seventy-six of the 104 patients are known to be deceased. Recent follow-up was obtained for 23 of the 28 remaining patients, all with more than 8.5 yr of follow-up. Five patients were lost to follow-up at 11 d, 7, 8, 23, and 32 mo. Observed versus expected survival for the IPF patients is shown in Figure 1. There is a significant (p 0.001) decrease in survival for IPF patients. Comparison of survival in the three main histopathologic subgroups is shown in Figure 2. UIP has significantly worse survival compared with NSIP and Others (p 0.001). In a multivariate survival analysis, UIP was still found to have significantly (p 0.001) worse survival than NSIP and Others, after adjusting for 3-yr expected survival probability. Two patients with AIP both died within 1 mo of their diagnosis as did the patient with diffuse pulmonary hemorrhage and the patient with honeycomb change. The patient with nonspecific scarring and bronchopneumonia survived 2.8 yr. One patient for whom pathologic material was unavailable for rereview survived 3.3 yr, and the other was still alive after 8.8 yr.

4 202 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL Figure 1. Overall survival of 104 patients diagnosed with idiopathic pulmonary fibrosis (IPF) via open lung biopsy: observed versus expected rate. For the entire group of patients with IPF, survival was significantly worse (p 0.001) than the expected rate based on all cause mortality for persons in the general population of the West North Central region of the U.S. of like age, sex, and calendar year of birth. p Value is from the one-sample log rank test comparison of the observed and expected survival curves. DISCUSSION Our findings demonstrate the heterogeneity of histopathologic findings in patients with the clinical syndrome of IPF. The survival rate for our group as a whole is very similar to that reported in previous studies (2, 6, 8 10). Our patients also had similar clinical findings in that they were middle-aged or older, presented with exertional dyspnea and chronic cough, and had diffuse bilateral radiographic infiltrates. Histologically confirmed UIP was associated with a significantly worse prognosis than other histologic patterns of chronic interstitial pneumonia including NSIP. Our findings are consistent with the results of the study by Katzenstein and Fiorelli (15) who demonstrated a better prognosis for patients with NSIP than that associated with UIP. Although our UIP patients were older, survival differences remained significant even after adjusting for 3-yr expected survival based on age, sex, and calendar year of birth. Liebow and Carrington (22) recognized 30 yr ago that patients with IPF were a heterogeneous group and emphasized the value of separating histologic patterns to predict natural history. Several frequently quoted studies analyzing survival in patients with IPF have been hampered by potentially inadequate pathologic analysis (2, 6, 8 10). For example, Turner- Warwick and colleagues (2) reported on 220 patients with IPF in whom biopsy material was available for review in only 42 cases, including trephine and other needle biopsies. Autopsy specimens were examined in 26 cases. Histopathologic confirmation of UIP is now known to be difficult in small lung biopsy specimens including those obtained by bronchoscopy and needle methods (23). Our study included only patients who had open lung biopsy and a minimum potential for 8 yr of follow-up. Classification of interstitial pneumonias has evolved in recent years. Although Hamman and Rich (24) are sometimes credited with the initial description of IPF, their patients contrast sharply with classic descriptions of IPF in that they pursued a fulminant and rapidly fatal course. AIP is a more recently proposed synonym for Hamman-Rich syndrome to underscore clinical and morphologic differences from IPF (13, 25). Indeed, our two patients with AIP died within 1 mo of diagnosis. In 1978, Carrington and colleagues delineated the differences between UIP and DIP and reported a higher response rate to corticosteroid therapy and a significantly better survival rate in patients with DIP (7). Since then, other interstitial lung disorders or histologic patterns have been described which also need to be distinguished from UIP. These include BOOP, RB-ILD, and most recently, NSIP (12 15). These entities may resemble IPF clinically, physiologically, and radiographically, but are associated with a higher response rate to corticosteroid therapy and a better prognosis. These observations raise questions about the extent to which historical observations of a 15 to 30% objective response rate to corticosteroid therapy reported in patients with IPF apply to patients with carefully defined UIP (2, 6, 8 10). We suspect that responders to corticosteroid therapy reported in previous studies may include patients with interstitial lung diseases other than UIP. Given the complications associated with long-term corticosteroid therapy, there is a need to clarify the role of corticosteroid therapy in patients with UIP. Our data suggest that DIP and NSIP represent separate and distinct entities rather than earlier forms of UIP (5, 7, 11). This conclusion is supported not only by differences in radiographic and histologic findings but also significant differences in survival curves that cannot be explained by differences in age at diagnosis (26 28). Our study reiterates that patients with IPF demonstrate heterogeneous lesions. Histopathological subtypes represent a dominant prognostic factor in patients with IPF. Histologically confirmed UIP is associated with a substantially worse survival than other categories of chronic interstitial pneumonia. These findings have important implications for advising IPF patients about the usefulness of corticosteroid therapy, prognosis, and timing of lung transplantation. Figure 2. Observed survival according to histopathological subgroups. Patients with UIP had a significantly worse survival (p 0.001, rank sum test) compared with nonspecific interstitial pneumonia/fibrosis and Others subgroups. Median survival for UIP patients was 2.8 yr. References 1. Crystal, R. G., J. D. Fulmer, W. C. Roberts, M. L. Moss, B. R. Line, and H. Y. Reynolds Idiopathic pulmonary fibrosis: clinical, histologic, radiographic, physiologic, scintigraphic, cytologic, and biochemical aspects. Ann. Intern. Med. 85: Turner-Warwick, M., B. Burrows, and A. Johnson Cryptogenic fi-

5 Bjoraker, Rhu, Edwin, et al.: Prognosis in Pulmonary Fibrosis 203 brosing alveolitis: clinical features and their influence on survival. Thorax 35: Raghu, G Idiopathic pulmonary fibrosis: a rational clinical approach. Chest 92: Cherniak, R. M., R. G. Crystal, and A. R. Kalica Current concepts in idiopathic pulmonary fibrosis: a road map for the future. Am. Rev. Respir. Dis. 143: King, T. E., Jr Idiopathic pulmonary fibrosis. In M. I. Schwarz and T. E. King, Jr., editors. Interstitial Lung Disease, 2nd ed. Mosby Year Book, St. Louis Stack, B. H. R., Y. F. J. Choo-Kang, and B. E. Heard The prognosis of cryptogenic fibrosing alveolitis. Thorax 27: Carrington, C. B., E. A. Gaensler, R. E. Coutu, M. X. FitzGerald, and R. G. Gupta Natural history and treated course of usual and desquamative interstitial pneumonia. N. Engl. J. Med. 298: Turner-Warwick, M., B. Burrows, and A. Johnson Cryptogenic fibrosing alveolitis: response to corticosteroid treatment and its effect on survival. Thorax 35: Tukiainen, P., E. Taskinen, P. Holsti, O. Kohola, and M. Valle Prognosis of cryptogenic fibrosing alveolitis. Thorax 38: Wright, P. H., B. E. Heard, S. J. Steel, and M. Turner-Warwick Cryptogenic fibrosing alveolitis: assessment by graded trephine lung biopsy compared with clinical, radiographic, and physiologic features. Br. J. Dis. Chest 75: Meier-Sydow, J., S. M. Weiss, R. Buhl, M. Rust, and G. Raghu Idiopathic pulmonary fibrosis: current clinical concepts and challenges in management. Sem. Respir. Crit. Care Med. 15: Epler, G. R., T. V. Colby, T. C. McCloud, C. B. Carrington, and E. A. Gaensler Bronchiolitis obliterans organizing pneumonia. N. Engl. J. Med. 312: Katzenstein, A. A., J. L. Myers, and M. Mazur Acute interstitial pneumonia: a clinicopathologic, ultrastructural, and cell kinetic study. Am. J. Surg. Pathol. 10: Myers, J. L., C. F. Veal, Jr., M. S. Shin, and A. A. Katzenstein Respiratory bronchiolitis causing interstitial lung disease: a clinicopathologic study of six cases. Am. Rev. Respir. Dis. 135: Katzenstein, A. A., and R. F. Fiorelli Nonspecific interstitial pneumonia/fibrosis: histologic features and clinical significance. Am. J. Surg. Pathol. 18: Tazelaar, H. D., R. W. Viggiano, J. Pickersgill, and T. V. Colby Interstitial lung disease in polymyositis and dermatomyositis: clinical features and prognosis as correlated with histologic findings. Am. Rev. Respir. Dis. 141: Yousem, S. A., T. V. Colby, and C. B. Carrington Lung biopsy in rheumatoid arthritis. Am. Rev. Respir. Dis. 131: Kaplan, E. L., and P. Meier Non-parametric estimation from incomplete observations. J.A.S.A. 53: Berstralh, E. J., and K. P. Offord. Conditional Probabilities Used in Calculating Cohort Expected Survival. Section of Biostatistics, Mayo Clinic. Technical Report Series No. 37, January Fleming, T. R., and D. P. Harrington A class of hypothesis tests for one and two sample censored survival data. Commun. Stat. A10 (8): Mantel, N Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother. Rep. 50: Liebow, A. A., and C. B. Carrington Alveolar diseases. The interstitial pneumonias. In M. Simon, editor. Frontiers of Pulmonary Radiology. Grune and Stratton, New York. 23. Wall, C. P., E. A. Gaensler, C. B. Carrington, and J. A. Hayes Comparison of transbronchial and open biopsies in chronic infiltrative lung diseases. Am. Rev. Respir. Dis. 123: Hamman, L., and A. R. Rich Acute diffuse interstitial fibrosis of the lungs. Bull. Johns Hopkins Hosp. 74: Olson, J., T. V. Colby, and C. G. Elliott Hamman-Rich syndrome revisited. Mayo Clin. Proc. 65: Park, J. S., K. S. Lee, J. S. Kim, C. S. Park, Y.-L. Suh, and D. L. Choi, and K. J. Kim Nonspecific interstitial pneumonia with fibrosis: radiographic and CT findings in seven patients. Radiology 195: Hartman, T. E., S. L. Primack, E.-Y. Kang, S. J. Swensen, D. M. Hansell, and G. McGuiness, and N. L. Müller Disease progression in usual interstitial pneumonia compared with desquamative interstitial pneumonia: assessment with serial CT. Chest 110: Akira, M., S. Yamamoto, H. Hara, M. Sakatani, and E. Ueda Serial computed tomographic evaluation in desquamative interstitial pneumonia. Thorax 52: