Idiopathic interstitial pneumonias and the concept of the trump card.

O ver the last several years, the clinical-radiologicpathologic classification of idiopathic interstitial pneumonias (IIPs) has undergone fine-tuning and some remodeling to define discrete identities for these diseases. Publication in 2002 of the American Thoracic Society/European Respiratory Society International Multidisciplinary Consensus Classification scheme1 represented an important milestone in this process. In this scheme, provisional criteria for diagnosis of idiopathic nonspecific interstitial pneumonia (NSIP) were provided, with recognition that additional investigation was necessary to solidly establish the characteristics of this process and its relationships to other IIPs. At about the same time, the existence of a relationship between NSIP and usual interstitial pneumonia (UIP) was suggested by Flaherty et al.2 These investigators discovered that 26% of patients receiving multilobar surgical lung biopsy for suspected idiopathic pulmonary fibrosis (IPF) had “discordant UIP” (NSIP in at least one lobe and UIP in at least one lobe), and that patients with a discordant UIP pattern had survival more closely resembling concordant UIP (UIP pattern in all lobes) than concordant NSIP (NSIP pattern in all lobes). Although survival for the discordant UIP group was slightly better than that of the concordant UIP group for the first 5 years, the survival curves intersected between 5 years and 6 years. In this issue, the article by Monaghan and colleagues (see page 522) confirms these findings and further emphasizes the advantages of obtaining surgical lung biopsy samples from multiple sites. In this retrospective analysis of a cohort of 64 patients undergoing multiple biopsies of single or multiple lobes for suspected cryptogenic fibrosing alveolitis (CFA)/IPF, which were histologically reclassified using the methods of Flaherty et al,2 25 patients showed concordant UIP, 8 patients showed discordant UIP, and 31 patients showed concordant NSIP. At 5 years after biopsy, survival rates for the concordant NSIP, discordant UIP, and concordant UIP groups were 75%, 37%, and 17%, respectively, and the survival difference between the NSIP and UIP groups remained significant after controlling for all other variables assessed. These results agree with those of other studies3–7 that have documented superior survival for NSIP as compared to UIP. Monaghan et al also show differences in mean patient age between groups, with discordant UIP occupying the intermediate position between concordant UIP and concordant NSIP, as was also noted by Flaherty et al.2 These findings raise interesting questions about the natural history and independence vs relatedness of these IIP patterns although, histologically, patterns of UIP and NSIP are generally distinct and identifiable. UIP demonstrates patchy, temporally heterogeneous lung involvement by dense fibrosis and mild or moderate interstitial lymphoplasmacytic infiltrates, architectural remodeling, honeycomb change, and fibroblastic foci.1 Fibroblastic foci are a key feature of UIP and are believed to represent zones of disease activity whose extensiveness has been linked to survival.8,9 NSIP includes a spectrum from cellular to fibrosing patterns.1 While distinguishing the cellular NSIP pattern from UIP is usually straightforward, separation of the fibrosing pattern of NSIP from UIP can be more challenging. Fibrosing NSIP demonstrates dense or loose interstitial fibrosis that is temporally homogeneous, may have a more diffuse distribution histologically, and usually lacks the more prominent fibroblastic foci, architectural remodeling, and honeycomb change characteristic of UIP.1,10 Occasional cases with overlapping features occur and can be difficult to classify. In patients with clinical features of CFA/IPF, it has been suggested that NSIP may represent an early phase of the disease process that is followed later by UIP,2 and that NSIP may represent relatively “inactive” disease in CFA/IPF.11 A histologic study10 of surgical lung biopsy and subsequent lung explant specimens confirms the high frequency of NSIP-like areas in UIP, but includes several cases with NSIP-like areas in explants but not in the earlier biopsies showing UIP, interpreted by the authors as CHEST editorials