Concordance between transbronchial lung cryobiopsy and surgical lung biopsy for interstitial lung disease in the same patients

BACKGROUND: The diagnostic accuracy and safety of transbronchial lung cryobiopsy (TBLC) via a flexible bronchoscope under sedation compared with that of surgical lung biopsy (SLB) in the same patients is unknown. METHODS: Retrospectively the data of fifty-two patients with interstitial lung diseases (median age: 63.5 years; 21 auto-antibody positive) who underwent TBLC followed by SLB (median time from TBLC to SLB: 57 days) was collected. The samples from TBLC and SLB were randomly labelled to mask the relationship between the two samples. Diagnosis was made independently by pathologists, radiologists, and pulmonary physicians in a stepwise manner, and a final diagnosis was made at multidisciplinary discussion (MDD). In each diagnostic step the specific diagnosis, the diagnostic confidence level, idiopathic pulmonary fibrosis (IPF) diagnostic guideline criteria, and treatment strategy were recorded. RESULTS: Without clinical and radiological information, the agreement between the histological diagnoses by TBLC and SLB was 42.3% (kappa [κ] = 0.23, 95% confidence interval [CI]: 0.08–0.39). However, the agreement between the TBLC-MDD and SLB-MDD diagnoses and IPF/non-IPF diagnosis using the two biopsy methods was 65.4% (κ = 0.57, 95% CI: 0.42–0.73) and 90.4% (47/52), respectively. Out of 38 (73.1%) cases diagnosed with high or definite confidence at TBLC-MDD, 29 had concordant SLB-MDD diagnoses (agreement: 76.3%, κ = 0.71, 95% CI: 0.55–0.87), and the agreement for IPF/non-IPF diagnoses was 97.4% (37/38). By adding the pathological diagnosis, the inter-observer agreement of clinical diagnosis improved from κ = 0.22 to κ = 0.42 for TBLC and from κ = 0.27 to κ = 0.38 for SLB, and the prevalence of high or definite diagnostic confidence improved from 23.0% to 73.0% and from 17.3% to 73.0%, respectively. Of all 383 TBLC performed during the same period, pneumothorax occurred in 5.0% of cases, and no severe bleeding, acute exacerbation of interstitial lung disease, or fatal event was observed. CONCLUSIONS: TBLC via a flexible bronchoscope under deep sedation is safely performed, and the TBLC-MDD diagnosis with a high or definite confidence level is concordant with the SLB-MDD diagnosis in the same patients. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12890-023-02571-9.