Enter Note Done


Decrease font size Increase font size


View Session Add to Schedule

Print Abstract

Marjorie Bateman, Fayez Kheir, Shigeki Saito, Gerald Berry, Ramsy Abdelghani, Philip Daroca, David Smith, David Weill, Shelby MacRae, Reinaldo RampollaContact Author(s)
Joseph LaskyPresenting Author(s)
Ala Alkhatib
Guidelines suggest surgical lung biopsy when a confident specific diagnosis of interstitial lung disease (ILD) is not achievable following careful history and exam, high-resolution computed tomography (HRCT), and a multidisciplinary discussion (MDD). However, invasive biopsies come at the price of morbidity, mortality and expense. Less-invasive biopsy methods, such as bronchoscopic lung cryobiopsy (BLC), have impacted the diagnostic impression and confidence of MDDs in the evaluation of ILDs. Recent reports indicate that a genomic classifier (GC) applied to lung tissue obtained by transbronchial biopsy (TBB) can distinguish usual interstitial pneumonia (UIP) from non-UIP. This study aimed to address the impact of sequentially presenting data from BLC and GC on the diagnostic confidence of the MDD in diagnosing ILD.

We enrolled 24 consecutive patients with fibrotic ILD without a typical UIP pattern on high-resolution computed tomography. All cases underwent BLC and TBB for GC. Two MDDs teams met to discuss the cases. Each MDD consisted of 2 pulmonologists, 1 radiologist and 1 pathologist. MDD1 sequentially reviewed clinical-radiologic findings, then biopsy results, and finally the GC result. At each step in the process the MDD diagnosis and confidence level was recorded. MDD2 sequentially reviewed the GC result prior to the BLC pathology report.

MDD1 had a significant increase in the diagnostic confidence after the addition of GC to BLC (from 35.7 to 71.4%, P = 0.03 regarding high confidence for IPF). There was no increase in high diagnostic confidence after the addition of BLC to GC in MDD2, albeit MDD2 had a higher confidence in the diagnosis of IPF than MDD1 following the clinical-radiographic presentation. Concordance coefficients and percentage agreement of UIP were: GC versus MDD1: κ = 0.92 (95% confidence interval [CI], 0.75-1), percentage agreement = 96% (95% CI, 79-100%); GC versus MDD2: κ = 0.75 (95% CI, 0.49-1), percentage agreement = 88% (95% CI, 77-97%); BLC1 versus MDD1: κ = 0.67 (95% CI, 0.38-0.97), percentage agreement = 83% (95% CI, 63-95%); BLC2 versus MDD2: κ = 0.66 (95% CI, 0.36-0.96), percentage agreement = 83% (95% CI, 63-95%).

The GC had a meaningful impact on the MDD diagnostic confidence for IPF. These data support the use of the GC alone in cases of possible or probable UIP radiographic pattern. BLC had the greatest impact regarding specific diagnosis in cases wherein the likelihood of UIP was considered low following clinical-radiographic review.

Clinical Implications:
We found that a GC that distinguishes usual interstitial pneumonia (UIP) from non-UIP in tissue obtained by transbronchial biopsy had a meaningful impact on the diagnostic confidence in the MDD diagnosis of IPF, and that BLC seldom added to the confidence following the review of the GC results.