Radiographic Fibrosis Score Predicts Survival in Hypersensitivity Pneumonitis

Radiographic Fibrosis Score Predicts Survival in Hypersensitivity Pneumonitis

CHEST Original Research DIFFUSE LUNG DISEASE Radiographic Fibrosis Score Predicts Survival in Hypersensitivity Pneumonitis Joshua J. Mooney, MD; Bre...

245KB Sizes 0 Downloads 37 Views

CHEST

Original Research DIFFUSE LUNG DISEASE

Radiographic Fibrosis Score Predicts Survival in Hypersensitivity Pneumonitis Joshua J. Mooney, MD; Brett M. Elicker, MD; Thomas H. Urbania, MD; Misha R. Agarwal, PhD; Christopher J. Ryerson, MD; Michelle Linh T. Nguyen, BA; Prescott G. Woodruff, MD, MPH; Kirk D. Jones, MD; Harold R. Collard, MD, FCCP; Talmadge E. King Jr, MD; and Laura L. Koth, MD

Background: It is unknown if the radiographic fibrosis score predicts mortality in persistent hypersensitivity pneumonitis (HP) and if survival is similar to that observed in idiopathic pulmonary fibrosis (IPF) when adjusting for the extent of radiographic fibrosis. Methods: We reviewed records from 177 patients with HP and 224 patients with IPF whose diagnoses were established by multidisciplinary consensus. Two thoracic radiologists scored highresolution CT (HRCT) scan lung images. Independent predictors of transplant-free survival were determined using a Cox proportional hazards analysis. Kaplan-Meier survival curves were constructed, stratified by disease as well as fibrosis score. Results: HRCT scan fibrosis score and radiographic reticulation independently predicted time to death or lung transplantation. Clinical predictors included a history of cigarette smoking, auscultatory crackles on lung examination, baseline FVC, and FEV1/FVC ratio. The majority of HP deaths occurred in patients with both radiographic reticulation and auscultatory crackles on examination, compared with patients with only one of these manifestations (P , .0001). Patients with IPF had worse survival than those with HP at any given degree of radiographic fibrosis (hazard ratio 2.31; P , .01). Conclusions: Survival in patients with HP was superior to that of those with IPF with similar degrees of radiographic fibrosis. The combination of auscultatory crackles and radiographic reticulation identified patients with HP who had a particularly poor outcome. CHEST 2013; 144(2):586–592 Abbreviations: HP 5 hypersensitivity pneumonitis; HR 5 hazard ratio; HRCT 5 high resolution CT; ILD 5 interstitial lung disease; IPF 5 idiopathic pulmonary fibrosis; UCSF 5 University of California San Francisco; UIP 5 usual interstitial pneumonia

pneumonitis (HP) is a diffuse parenHypersensitivity chymal lung disease characterized by an immu-

nologic reaction to an inhaled organic antigen.1 Patients with HP, especially those with acute manifestations, generally have a favorable outcome, particularly when Manuscript received October 25, 2012; revision accepted January 10, 2013. Affiliations: From the Departments of Medicine (Drs Mooney, Agarwal, Woodruff, Collard, King, and Koth and Ms Nguyen), Radiology (Drs Elicker and Urbania), and Pathology (Dr Jones), University of California, San Francisco, CA; and the Department of Medicine (Dr Ryerson), University of British Columbia, Vancouver, BC, Canada. Drs Elicker, Urbania, and Agarwal contributed equally to this article. Part of this article was presented in abstract form at the American Thoracic Society International Conference, May 18-23, 2012, San Francisco, CA (abstract 4365).

the causative antigen is identified and removed.2-6 However, patients may alternatively develop a chronic form of disease that can be characterized by the presence of fibrosis on histopathologic7-9 or radiographic8,10-13 evaluation. Data from the National Center for Health Statistics show that age-adjusted mortality rates from HP have increased significantly (P , .0001) from 0.09 to 0.29 per million between 1980 and 2002 in US Funding/Support: This study was supported by departmental sources from the University of California San Francisco. Correspondence to: Laura L. Koth, MD, Department of Medicine, University of California San Francisco, Box 0111, 505 Parnassus Ave, San Francisco, CA 94143; e-mail: [email protected] © 2013 American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details. DOI: 10.1378/chest.12-2623

586

Downloaded From: http://journal.publications.chestnet.org/ by David Kinnison on 08/06/2013

Original Research

adults.14 Both radiographic and pathologic fibrosis are associated with an increase in all-cause mortality in HP,7,9-13 but it is unclear if these are independent risk factors when adjusting for other clinical and physiologic predictors. In this study, we sought to determine the clinical, physiologic, and radiographic features that were predictive of mortality in patients with HP who were referred to a tertiary care center. We assessed whether the radiographic fibrosis score—a composite score of the visual extent of reticulation and honeycombing on high-resolution CT (HRCT) scanning shown to be a predictor of mortality in patients with idiopathic pulmonary fibrosis (IPF)—was predictive of mortality in HP.15,16 Finally, we determined if HP had a similar survival compared with IPF when adjusting for severity of radiographic fibrosis. Addressing these questions will clarify the prognostic utility of clinical and radiographic findings in managing patients with advanced HP. Materials and Methods Study Population Patients with HP (n 5 190) and IPF (n 5 242) were prospectively enrolled from the University of California San Francisco (UCSF) Interstitial Lung Disease (ILD) Clinic from March 2000 to October 2010. Data from all patients were collected prospectively using standardized questionnaires and physician review. The diagnostic criteria for HP or IPF were based upon consensus agreement by experts at a multidisciplinary conference after thorough review of all available data in accordance with established American Thoracic Society/European Respiratory Society guidelines.17-20 Specifically, the diagnosis of HP was made based upon the following criteria: (1) consistent clinical history and features including chronic respiratory symptoms, (2) abnormal physiologic changes on pulmonary function tests, (3) a compatible HRCT scan as reviewed by an experienced thoracic radiologist (B. M. E.), (4) exclusion of other disease mimics, and (5) biopsy confirmation when a plausible antigen exposure could not be identified. The diagnosis of IPF was made by clinical history and physical examination to exclude other ILDs. In addition, we also required a consistent surgical lung biopsy specimen of usual interstitial pneumonia (UIP) and/or a radiographic pattern on CT scan that was considered to be “definite” UIP by an expert thoracic radiologist (B. M. E). The Committee on Human Research at UCSF approved the study design (institutional review board number 10-03488). Clinical Assessment and Measurements Among the patients with HP, 13 were excluded for missing clinical and/or pulmonary function variables, resulting in a clinical cohort of 177 patients. Forty-five of these patients did not have HRCT scans available for re-review and were excluded from the radiographic analysis cohort, which consisted of 132 patients. Among the cohort with IPF, 224 patients had complete clinical records. Date of diagnosis was defined as the date of the initial UCSF ILD clinic visit. Patient demographics, symptoms (cough, dyspnea score21), signs (auscultatory crackles), history of tobacco use, BMI, and pulmonary function values were recorded prospectively. The use of oxygen was dichotomously recorded based upon

use of long-term oxygen therapy or oxygen saturation , 88% with ambient air at the patient’s initial clinic visit. Antigen exposures, as determined by the initial evaluating clinician, were classified into avian, microbial, or unknown categories, as previously described.9 If the type or significance of the antigen was unclear, the exposure was classified as unknown. Serum precipitins or industrial-hygienist reports were not required for diagnosis or antigen confirmation given the lack of standardization and clinical utility.9,22,23 Vital status and all-cause mortality were ascertained for all patients by review of medical records and the Social Security Death Registry Index. UCSF’s lung transplantation database was cross-referenced from March 2000 to October 2010 with all patients with HP and IPF to ascertain lung transplantation status. Radiographic Assessment Baseline HRCT scans were re-reviewed by two experienced thoracic radiologists (B. M. E., T. H. U.) who were blinded to all clinical data. The mean extent of reticulation and honeycombing was scored to the nearest 5% in three zones in each lung as previously described15 to produce a semiquantitative CT fibrosis score. For the presence of ground-glass opacity, consolidation, mosaic perfusion, and traction bronchiectasis, each lung zone was scored on a four-point scale (0 5 no involvement, 1 5 1%-25% involvement, 2 5 26%-50% involvement, 3 5 51%-75% involvement, or 4 5 76%-100% involvement) as previously described.16,24 The average total score for each variable was calculated as the mean score of the six lung zones. Interobserver agreement for all variables was calculated by Spearman rank correlation coefficient. Joint review and consensus adjudication was used to resolve differences in eight CT scans from patients with HP with honeycombing difference . 5%. Statistical Analysis Patients’ demographic, clinical, and radiographic data were compared across groups using the t test or Mann-Whitney U test as appropriate for continuous variables, the x2 test for categorical variables, and the log-rank test for survival curves. A P value , .05 was considered significant in all cases. Predictors of time to death or transplant were determined for patients with HP using Cox proportional hazards analysis. Variables associated with time to death or transplant on unadjusted analysis were considered for inclusion in the multivariate model. The multivariate model was built using a stepwise model-building algorithm that finds the model with the smallest Akaike information criterion.25-27 This algorithm retains the subset of variables that lead to a model with the lowest Akaike information-criterion score. Kaplan-Meier survival curves were built, stratified by disease, and substratified by radiographic variables. For our survival analyses, our primary end point was time to death or lung transplantation. In secondary analyses, we developed competing risk models that assessed death and transplantation separately. To analyze the relationship between death or lung transplantation and presence or absence of crackles on lung examination and fibrosis on CT scan, a two-proportion z test with unpooled variance was used to test the difference in death proportions in each subgroup. All statistical analyses were performed using the statistical software R, version 2.15.1 (R Project for Statistical Computing; http://cran.r-project.org/bin/windows/base/). R package “survival” was used for survival analyses, package “mstate” for competing risk models, and base packages were used for all other analyses.

Results Study Population Baseline demographics are shown in Table 1. Compared with IPF, patients with persistent HP were

journal.publications.chestnet.org

Downloaded From: http://journal.publications.chestnet.org/ by David Kinnison on 08/06/2013

CHEST / 144 / 2 / AUGUST 2013

587

younger, more likely to be female, less likely to be ever smokers, had fewer pack-years of smoking, and had less crackles on auscultatory lung examination. Patients with HP had greater physiologic obstruction (based on FEV1 and FEV1/FVC), whereas patients with IPF had greater restriction and lower diffusing capacity. Patients with HP had a lower radiographic fibrosis score (median score, 11.04 for HP vs 15.00 for IPF). Patients with HP with and without biopsy specimens did not differ with respect to clinical characteristics, symptoms, physiologic function, antigen exposure, number of lung transplants, or number of deaths (e-Table 1). Sixty-five patients with HP had an identifiable antigen exposure (75% were avian, and 25% were microbial) (e-Table 1). These patients were instructed to perform specific abatement procedures (eg, removal of bird, extensive cleaning of home environment, removal of down products, remediation of sources of water-damaged structures, and changing residence if possible). Of these 65 patients, 25% underwent home industrial-hygienist evaluations. Treatment modalities at initial clinic visit included current use of prednisone in 45.2% of patients (80 of 177) and/or other immunomodulation therapy (ie, mycophenolate mofetil, azathioprine, and/or cyclophosphamide) in 7.9% of patients (14 of 177) and are summarized in e-Table 2. Clinical and Physiologic Predictors of Survival in HP Clinical predictors of time to death or lung transplantation identified on univariate analyses from the entire

HP cohort included dyspnea score, crackles on physical examination, need for oxygen therapy, and lung function measures (FVC, total lung capacity, diffusion capacity for carbon monoxide expressed as % predicted, and FEV1/FVC ratio) (Table 2). Independent predictors of time to death or lung transplantation included auscultatory crackles on lung examination (hazard ratio [HR], 4.7; P 5 .02), FVC (HR, 0.7; P 5 .01), FEV1/FVC ratio (HR, 1.58; P , .01), and a history of cigarette smoking (HR, 2.68; P 5 .01) (Table 2). Radiographic Predictors of Survival in HP For patients with HP who had HRCT scans available for scoring (n 5 132), we repeated univariate and multivariate modeling using all prior clinical variables and CT scan features of ground-glass opacities, consolidation, mosaic perfusion, reticulation, traction bronchiectasis, honeycombing, and fibrosis score. Interobserver correlation between radiologist scores was good to excellent for reticulation, traction bronchiectasis, honeycombing, and fibrosis score (e-Table 3). Radiographic honeycombing, reticulation, traction bronchiectasis, and fibrosis score were statistically significant univariate predictors of time to death or lung transplantation (Table 3). Using multivariate analyses, fibrosis score was independently associated with mortality and/or transplantation (HR, 1.35; P , .01) (Table 4). When fibrosis score was removed from the model and replaced by its individual components (reticulation and honeycombing), we found that reticulation was the radiographic feature independently

Table 1—Baseline Subject Characteristics Characteristics Clinical cohort No. Men Age, y Ever smoker Current smoker Pack-y Cough Auscultatory crackles Dyspnea score Long-term oxygen therapy BMI FEV1 % predicted FVC % predicted FEV1/FVC TLC % predicted Dlco % predicted Deceased Transplanted HRCT scan sub-cohort, No. CT scan fibrosis score

HP

IPF

177 54 (31) 60.76 ⫾ 11.3 84 (48) 1 (0.6) 8.64 ⫾ 15.2 149 (84) 100 (57) 9.97 ⫾ 5.51 27 (15) 29.16 ⫾ 6.15 71.32 ⫾ 20.78 68.92 ⫾ 20.48 80.29 ⫾ 11.68 71.91 ⫾ 18.27 49.20 ⫾ 19.41 21 (12) 10 (6) 132 17.82 ⫾ 18.14

224 165 (74) 69.96 ⫾ 8.65 174 (78) 4 (2) 24.01 ⫾ 25.61 185 (83) 212 (95) 9.82 ⫾ 5.71 54 (24) 28.2 ⫾ 4.72 77.58 ⫾ 18.59 69.74 ⫾ 17.48 82.23% ⫾ 7.74 67.39 ⫾ 15.32 44.01 ⫾ 16.73 96 (43) 15 (7) 182 18.42 ⫾ 12.42

P Valuea , .01 , .01 , .01 .28 .01 .82 , .01 .90 .03 .10 .03 .91 , .01 .01 .01 , .01 .53 .01

Data are expressed as mean ⫾ SD or No. (%) unless otherwise indicated. Dlco 5 diffusing capacity of the lung for carbon monoxide; HP 5 hypersensitivity pneumonitis; HRCT 5 high-resolution CT; IPF 5 idiopathic pulmonary fibrosis; TLC 5 total lung capacity. aValues in boldface are statistically significant (P ⱕ .05). 588

Downloaded From: http://journal.publications.chestnet.org/ by David Kinnison on 08/06/2013

Original Research

Table 2—HP Clinical Cohort Univariate and Multivariate HRs Variable Univariate variable Age, y Male sex Cough Dyspnea scoreb Ever smoker Pack-years Crackles Oxygen therapy Antigen identified BMI FVC % predictedc FEV1/FVCd TLC % predictedc Dlco % predictedc Multivariate variable Crackles FVC % predictedc FEV1/FVCd Ever smoker Oxygen therapy

Table 3—Univariate Analyses of HP Radiographic Cohort

HR

95% CI

P Valuea

1.24 1.10 0.99 1.20 1.73 1.07 10.70 3.70 1.08 1.01 0.61 1.66 0.63 0.62

0.91-1.70 0.53-2.31 0.38-2.57 1.06-1.37 0.85-3.54 0.94-1.23 3.23-35.21 1.94-8.26 0.50-2.29 0.95-1.07 0.49-0.75 1.28-2.15 0.52-0.76 0.50-0.78

.17 .79 .98 , .01 .13 .30 , .01 , .01 .85 .83 , .01 , .01 , .01 , .01

4.70 0.70 1.58 2.68 1.81

1.33-16.98 0.55-0.92 1.18-2.12 1.26-5.73 0.86-4.02

.02 .01 , .01 .01 .13

HR 5 hazard ratio. See Table 1 legend for expansion of other abbreviations. aValues in boldface are statistically significant (P ⱕ .05). bDefined as per 2-unit change. cDefined as per 10% change. dDefined as per 5% change.

associated with time to death or lung transplantation (e-Table 4). The vast majority of deaths in the HP radiographic cohort occurred when patients manifested reticulation and/or honeycombing as well as auscultatory crackles (P , .0001) (Table 5). To visualize how fibrosis score relates to survival in patients with HP, the radiographic cohort was divided into quartiles by their fibrosis score. Patients in the highest quartile for fibrosis score, Q4, had worse transplant-free survival than patients in the two lowest quartiles for fibrosis score (Q1, P , .01; Q2, P , .01) (Fig 1). Patients in the third quartile, Q3, had worse survival than Q1 (P , .05). Survival in the third quartile, Q3, and fourth quartile, Q4, were not statistically significantly different (P 5 .22). Survival in HP Compared With IPF On unadjusted analysis, our cohort of patients with HP had longer time to death or lung transplantation compared with patients with IPF (P , .0001) (Fig 2). The percentage of patients experiencing death or lung transplantation at 5 years was 25% for HP and 65% for IPF. However, since we and other groups have found that patients with HP with fibrosis have worse survival compared with those without fibrosis, we next determined whether patients with HP and those with IPF who had the same extent of radio-

Variable Univariate variable Age, y Male sex Cough Dyspnea scoreb Ever smoker Pack-years Crackles Oxygen therapy Antigen identified BMI Pulmonary function FVC % predictedc FEV1/FVCd TLC % predictedc Dlco % predictedc Radiographic Consolidation Ground-glass opacity Honeycombing Mosaic perfusion Reticulation Traction bronchiectasis Fibrosis score

HR

95% CI

P Valuea

1.27 1.18 0.84 1.23 1.83 1.12 12.73 4.08 0.92 1.00

0.87-1.84 0.52-2.68 0.29-2.45 1.06-1.43 0.82-4.07 0.96-1.30 2.98-54.39 1.78-9.35 0.38-2.20 0.94-1.06

.21 .69 .75 , .01 .14 .12 , .01 , .01 .84 .89

0.65 1.66 0.70 0.67

0.52-0.83 1.22-2.26 0.56-0.87 0.52-0.85

, .01 .01 , .01 , .01

0.55 0.91 10.26 1.40 3.02 4.46 1.54

0.07-4.24 0.51-1.64 1.32-79.23 0.73-2.71 1.80-5.10 2.08-9.57 1.25-1.88

.57 .76 .03 .31 , .01 , .01 , .01

See Table 1 and 2 legends for expansion of abbreviations. Values in boldface are statistically significant (P ⱕ .05). bDefined as per 2-unit change. cDefined as per 10% change. dDefined as per 5% change. a

graphic fibrosis had similar mortality. Controlling for potential confounders and predictors of survival or lung transplantation including fibrosis score, a diagnosis of IPF was associated with a shorter time to death or lung transplantation compared with HP (Table 6). We obtained similar results using a competing risk model (e-Table 5). Discussion The results from this study extend prior observations regarding the role of radiographic fibrosis in survival in patients with persistent HP and identified several novel findings. These include (1) that the radiographic fibrosis score is an independent predictor of Table 4—Multivariate Analyses of HP Radiographic Cohort Multivariate Variable

HR

95% CI

P Valuea

Fibrosis score Crackles Oxygen therapy FEV1/FVCb

1.35 5.20 2.93 1.35

1.08-1.70 1.14-23.85 1.22-7.04 0.96-1.89

, .01 .03 .02 .08

See Table 1 and 2 legends for expansion of abbreviations. Values in boldface are statistically significant (P ⱕ .05). bDefined as per 5% change. a

journal.publications.chestnet.org

Downloaded From: http://journal.publications.chestnet.org/ by David Kinnison on 08/06/2013

CHEST / 144 / 2 / AUGUST 2013

589

Table 5—Patients With HP Reaching Death or Transplant in Radiographic Cohort (n 5 132) Crackles

Reticulations and/or honeycombing on CT scan

No Yes

No

Yes

1/18 1/36

1/3 22/75a

See Table 1 legend for expansion of abbreviations. aP , .0001 compared with absence of crackles, with or without reticulations

mortality in persistent HP, (2) that patients with both fibrosis (reticulation and/or honeycombing) and auscultatory crackles on chest examination have a particularly poor outcome compared with patients with just one of these clinical findings, and (3) that patients with persistent HP have a better outcome compared with IPF, even after adjustment for radiographic fibrosis severity. Radiographic fibrosis has been shown to be a predictor of survival in HP.10,11 Tateishi et al12 found the presence of radiographic honeycombing independently associated with decreased survival in avian HP. Walsh et al13 studied only chronic HP and found that traction bronchiectasis was the strongest independent predictor of mortality. In contrast, an underpowered study by Sahin and colleagues8 found that radiographic fibrosis was not associated with worse survival. Our study shows that the extent of reticulation on CT scan was an independent predictor of

Figure 2. Kaplan-Meier plot comparing survival in patients with HP (solid line) and those with IPF (dotted line) (n 5 177 patients with HP and n 5 224 patients with IPF). HP 5 hypersensitivity pneumonitis; IPF 5 idiopathic pulmonary fibrosis.

mortality in HP. In addition, we show that the composite score of reticulation and honeycombing (ie, fibrosis score) was also an independent predictor of mortality in HP. An unexpected result was the finding that patients with HP who presented with both reticulations on CT scan and auscultatory crackles on examination had significant mortality compared with patients with HP who had only one (or neither) of these findings (Table 5). Crackles have recently been proposed as a valuable screening tool for detecting populations with early or subclinical IPF and our study supports its utility in prognosticating advanced HP.28,29 Because crackles on lung examination is a subjective and nonspecific physical examination finding that depends on the knowledge and skill of the physician doing the examination, it could be considered unreliable and noninformative in assessing disease severity. However, two independent studies have found crackles to be associated with mortality on univariate analyses in patients with HP.10,11 The prognostic utility of this finding may be different in less-experienced practitioners Table 6—HP and IPF Multivariate Analyses

Figure 1. Kaplan-Meier plot stratifying patients with hypersensitivity pneumonitis by quartiles of fibrosis scores. Scores were generated as described in the “Materials and Methods” section and divided into quartiles where Q1 represents the lowest range of scores and Q4 represents the highest. Only significant P values , .05 are shown. (Sample size per quartile of reticulation score: Q1, 36 patients; Q2, 30 patients; Q3 and Q4, 33 patients). P , .05 for Q1 compared with Q3; P , .01 for Q1 compared with Q4; and P , .01 for Q2 compared with Q4.

Multivariate Variable

HR

95% CI

P Valuea

IPF diagnosis Fibrosis score Age, y Ever smoker FVC % predictedb Oxygen therapy Crackles

2.31 1.03 1.21 1.61 0.83 2.37 2.57

1.37-3.90 1.02-1.05 0.96-1.52 1.01-2.57 0.73-0.94 1.52-3.71 0.89-7.36

, .01 , .01 .10 .05 , .01 , .01 .08

See Table 1 and 2 legends for expansion of abbreviations. aValues in boldface are statistically significant (P ⱕ .05). bDefined as per 10% change.

590

Downloaded From: http://journal.publications.chestnet.org/ by David Kinnison on 08/06/2013

Original Research

and external validation is required to confirm these results. Despite these limitations, the combination of reticulations on CT scan and presence of crackles may provide synergistic information in predicting poor outcome in the clinical setting. Another goal of this study was to compare survival of patients with HP to that of those with IPF with similar degrees of radiographic fibrosis. Pérez-Padilla et al7 found that mortality was indistinguishable between patients with a UIP-like pattern on lung histopathology and bird exposure compared with those without bird exposure. We found that overall mortality was better in HP compared with IPF, even after controlling for degree of radiographic fibrosis; however, our cohort was not solely composed of patients with avian HP. We speculate that the differences in prognosis associated with fibrosis in HP and IPF could be due to responsiveness of HP to treatment (antigen avoidance or drug therapy), lead time bias, differences in the rates of disease progression or exacerbations, or different frequencies of associated comorbidities (eg, higher rates of pulmonary hypertension, emphysema, and lung cancer in patients with IPF). Conclusions We found that radiographic fibrosis, specifically the extent of fibrosis (reticulations and honeycombing), predicts worse outcome in patients with HP. The combination of reticulations on CT scan and crackles on lung examination may identify patients with a particularly poor outcome and deserves further evaluation in future HP studies. Transplant-free survival in patients with HP remains superior to that of patients with IPF with similar degrees of radiographic fibrosis. These findings highlight the importance of an accurate diagnosis when conveying prognostic information to patients, directing treatment, and considering referral for lung transplant. Acknowledgments Author contributions: Dr Mooney had full access to the data and takes responsibility for the integrity of the data and the accuracy of the data analysis. Dr Mooney: contributed to study design; data collection and analysis; and writing, review, and approval of the final manuscript and served as principal author. Dr Elicker: contributed to CT scan data collection and analysis and review and approval of the final manuscript. Dr Urbania: contributed to CT scan data collection and analysis and review and approval of the final manuscript. Dr Agarwal: contributed to the writing, review, and approval of the final manuscript and performed all statistical analyses. Dr Ryerson: contributed to data collection of all IPF subjects and writing, review, and approval of the final manuscript. Ms Nguyen: contributed to data collection and review and approval of the final manuscript. Dr Woodruff: contributed to statistical analysis and writing, review, and approval of the final manuscript.

Dr Jones: contributed to histopathology slide review in determination of diagnoses and review and approval of the final manuscript. Dr Collard: contributed to oversight of data collection and review and approval of the final manuscript. Dr King: contributed to the writing, review, and approval of the final manuscript. Dr Koth: contributed to study design; data collection and statistical analysis; and the writing, review, and approval of the final manuscript. Financial/nonfinancial disclosures: The authors have reported to CHEST that no potential conflicts of interest exist with any companies/organizations who products or services may be discussed in this article. Role of sponsors: The sponsor had no role in the design of the study, the collection and analysis of the data, or the preparation of the manuscript. Other contributions: We thank all the patients and members of the UCSF ILD clinic who made this study possible. Additional information: The e-Tables can be found in the “Supplemental Materials” area of the online article.

References 1. Fink JN. Hypersensitivity pneumonitis. J Allergy Clin Immunol. 1984;74(1):1-10. 2. Zacharisen MC, Schlueter DP, Kurup VP, Fink JN. The longterm outcome in acute, subacute, and chronic forms of pigeon breeder’s disease hypersensitivity pneumonitis. Ann Allergy Asthma Immunol. 2002;88(2):175-182. 3. de Gracia J, Morell F, Bofill JM, Curull V, Orriols R. Time of exposure as a prognostic factor in avian hypersensitivity pneumonitis. Respir Med. 1989;83(2):139-143. 4. Braun SR, doPico GA, Tsiatis A, Horvath E, Dickie HA, Rankin J. Farmer’s lung disease: long-term clinical and physiologic outcome. Am Rev Respir Dis. 1979;119(2):185-191. 5. Cormier Y, Bélanger J. Long-term physiologic outcome after acute farmer’s lung. Chest. 1985;87(6):796-800. 6. Moran JV, Greenberger PA, Patterson R. Long-term evaluation of hypersensitivity pneumonitis: a case study follow-up and literature review. Allergy Asthma Proc. 2002;23(4):265-270. 7. Pérez-Padilla R, Salas J, Chapela R, et al. Mortality in Mexican patients with chronic pigeon breeder’s lung compared with those with usual interstitial pneumonia. Am Rev Respir Dis. 1993;148(1):49-53. 8. Sahin H, Brown KK, Curran-Everett D, et al. Chronic hypersensitivity pneumonitis: CT features comparison with pathologic evidence of fibrosis and survival. Radiology. 2007;244(2): 591-598. 9. Vourlekis JS, Schwarz MI, Cherniack RM, et al. The effect of pulmonary fibrosis on survival in patients with hypersensitivity pneumonitis. Am J Med. 2004;116(10):662-668. 10. Hanak V, Golbin JM, Hartman TE, Ryu JH. High-resolution CT findings of parenchymal fibrosis correlate with prognosis in hypersensitivity pneumonitis. Chest. 2008;134(1): 133-138. 11. Lima MS, Coletta EN, Ferreira RG, et al. Subacute and chronic hypersensitivity pneumonitis: histopathological patterns and survival. Respir Med. 2009;103(4):508-515. 12. Tateishi T, Ohtani Y, Takemura T, et al. Serial high-resolution computed tomography findings of acute and chronic hypersensitivity pneumonitis induced by avian antigen. J Comput Assist Tomogr. 2011;35(2):272-279. 13. Walsh SL, Sverzellati N, Devaraj A, Wells AU, Hansell DM. Chronic hypersensitivity pneumonitis: high resolution computed tomography patterns and pulmonary function indices as prognostic determinants. Eur Radiol. 2012;22(8):1672-1679. 14. Bang KM, Weissman DN, Pinheiro GA, Antao VC, Wood JM, Syamlal G. Twenty-three years of hypersensitivity pneumonitis

journal.publications.chestnet.org

Downloaded From: http://journal.publications.chestnet.org/ by David Kinnison on 08/06/2013

CHEST / 144 / 2 / AUGUST 2013

591

15.

16.

17.

18.

19.

20.

mortality surveillance in the United States. Am J Ind Med. 2006;49(12):997-1004. Best AC, Meng J, Lynch AM, et al. Idiopathic pulmonary fibrosis: physiologic tests, quantitative CT indexes, and CT visual scores as predictors of mortality. Radiology. 2008;246(3): 935-940. Lynch DA, Godwin JD, Safrin S, et al; Idiopathic Pulmonary Fibrosis Study Group. High-resolution computed tomography in idiopathic pulmonary fibrosis: diagnosis and prognosis. Am J Respir Crit Care Med. 2005;172(4):488-493. American Thoracic Society. American Thoracic Society. Idiopathic pulmonary fibrosis: diagnosis and treatment. International consensus statement. American Thoracic Society (ATS), and the European Respiratory Society (ERS). Am J Respir Crit Care Med. 2000;161(2 pt 1):646-664. American Thoracic Society; European Respiratory Society. American Thoracic Society/European Respiratory Society International Multidisciplinary Consensus Classification of the Idiopathic Interstitial Pneumonias. This joint statement of the American Thoracic Society (ATS), and the European Respiratory Society (ERS) was adopted by the ATS board of directors, June 2001 and by the ERS Executive Committee, June 2001. Am J Respir Crit Care Med. 2002;165(2):277-304. Flaherty KR, King TE Jr, Raghu G, et al. Idiopathic interstitial pneumonia: what is the effect of a multidisciplinary approach to diagnosis? Am J Respir Crit Care Med. 2004; 170(8):904-910. Ohtani Y, Saiki S, Kitaichi M, et al. Chronic bird fancier’s lung: histopathological and clinical correlation. An applica-

21.

22. 23. 24.

25. 26. 27. 28. 29.

tion of the 2002 ATS/ERS consensus classification of the idiopathic interstitial pneumonias. Thorax. 2005;60(8):665-671. Watters LC, King TE, Schwarz MI, Waldron JA, Stanford RE, Cherniack RM. A clinical, radiographic, and physiologic scoring system for the longitudinal assessment of patients with idiopathic pulmonary fibrosis. Am Rev Respir Dis. 1986;133(1): 97-103. Burrell R, Rylander R. A critical review of the role of precipitins in hypersensitivity pneumonitis. Eur J Respir Dis. 1981;62(5):332-343. Cormier Y, Bélanger J. The fluctuant nature of precipitating antibodies in dairy farmers. Thorax. 1989;44(6):469-473. Arakawa H, Johkoh T, Honma K, et al. Chronic interstitial pneumonia in silicosis and mix-dust pneumoconiosis: its prevalence and comparison of CT findings with idiopathic pulmonary fibrosis. Chest. 2007;131(6):1870-1876. Akaike H, ed. Information theory and an extension of the maximum likelihood principle. Budapest: Academiai Kiado; 1973. Akaike H. A new look at statistical model identification. IEEE Trans Automat Contr. 1974; 19(6):716-723. Bozdogan H. Model selection and Akaike’s Information Criterion (AIC): the general theory and its analytical extensions. Psychometrika. 1987;52(3):345-370. Cottin V, Cordier JF. Velcro crackles: the key for early diagnosis of idiopathic pulmonary fibrosis? Eur Respir J. 2012;40(3): 519-521. Cottin V, Cordier JF. Subclinical interstitial lung disease: no place for crackles? Am J Respir Crit Care Med. 2012;186(3): 289.

592

Downloaded From: http://journal.publications.chestnet.org/ by David Kinnison on 08/06/2013

Original Research