Update in Diffuse Parenchymal Lung Disease 2007

Abstract

It is now widely accepted that the diagnostic reference standardfor interstitial lung disease is multidisciplinary evaluation, withthe integration of clinical, radiologic, and histologic data (1).However, published data generated by leading clinicians, radi-ologists, and histopathologists cannot be extrapolated to prac-tice outside tertiary centers. In a study from Michigan, a pre-viously reported multidisciplinary diagnostic algorithm (2) wascompared between expert and community practitioners usinga common dataset (3). Diagnostic agreement was moderatelygoodbetweenexpertparticipantsandimprovedsignificantlywhenafinaldiagnosiswasformulatedfromclinical,imaging,andbiopsydata. By contrast, agreement on the final diagnosis was only fairbetweencommunityphysicians,eventhoughdiagnosticagreementimprovedasdatawereintegrated.Thesefindingsunderlinewhatislikely to become an increasingly important issue: the need forregional clinics in diffuse lung disease, with ready access tospecialistskillsinkeydisciplines.Furtherstudiesarenowrequiredto quantify the value added by expert evaluation. In Flaherty andcolleagues’ studies (2, 3), the greatest disparity between the twogroups lay in the diagnosis of nonspecific interstitial pneumonia(NSIP),whichwasmoreoftenidentifiedbyacademicpractitionersthanbycommunity participants,who tendedtoprefer a diagnosisof idiopathic pulmonary fibrosis (IPF).NSIP, initially viewed as a ‘‘provisional’’ entity in thedeliberations of the American Thoracic Society/European Re-spiratory Society International Consensus Committee (4), isoften a source of diagnostic confusion. Idiopathic NSIP andNSIP secondary to disorders such as hypersensitivity pneumo-nitis (HP), drug-induced lung disease, and, especially, connec-tive tissue disease (CTD), are often subcategorized, but recentdata suggest that this distinction may be largely artificial. Ina cohort of patients with idiopathic NSIP, careful evaluationdisclosed a very high prevalence of clinical and serologic abnor-malities strongly suggestive of underlying undifferentiatedCTD or ‘‘CTD in waiting’’ (5), lending support to the hypothesisthat idiopathic NSIP is, in reality, CTD confined to the lungs inmany cases (6). Further important insights came from a study inwhich outcomes in biopsied patients were compared betweenidiopathic NSIP, IPF, and pulmonary fibrosis in CTD (7).Severity-adjusted survival in idiopathic NSIP was virtuallyidentical with that of NSIP in CTD, in keeping with a systemicautoimmune hypothesis for idiopathic NSIP. By contrast,mortality was higher in patients with IPF than in patients withCTD and usual interstitial pneumonia (UIP). Interestingly,mortality in UIP was higher in rheumatoid arthritis (RA) thanin other CTDs and was only marginally lower than in IPF,although the small size of the RA subgroup is an importantcaveat. UIP was more prevalent than NSIP in this RA cohort,and, based on earlier reports (8–10), appears to be much moreprevalent in RA than in other CTDs.On the basis of these and earlier observations, it appears thatthe entity of NSIP in the current classification of the idiopathicinterstitial pneumonias (IIPs) should now be reappraised. It hasalso been argued that the current classification of the IIPs createsdifficulties in the search for new therapies, by causing unduepolarization between competing pathogenetic hypotheses (11).InHP,fibroticNSIPandUIPbothoccur,andgeneexpressiondata are suggestive of underlying HP in a subset of patients withapparently idiopathic NSIP (12). The relative prognostic signif-icance of these patterns in HP is not yet known but, based ona recent report, it appears that computed tomography (CT)findings of extensive reticular disease, with or without honey-combing, and traction bronchiectasis are strongly indicative ofunderlying fibrotic disease (13). Further CT studies in largercohorts of patients with HP can be anticipated, comparing out-comes in relation to CT patterns of NSIP and UIP.