Transbronchial needle aspiration combined with cryobiopsy in the diagnosis of mediastinal diseases: a multicentre, open-label, randomised trial

Background Transbronchial mediastinal cryobiopsy is a novel sampling technique for mediastinal disease. Despite the possibility of lung cancer misdiagnosis, the improved diagnostic yield of this approach for non-lung-cancer lesions compared with standard endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) highlights its diagnostic potential as a complementary technique to conventional biopsy. We aimed to evaluate the safety profile and added value of the combined use of transbronchial mediastinal cryobiopsy and standard EBUS-TBNA for the diagnosis of mediastinal diseases. Methods We conducted an open-label, randomised trial at three hospital sites in Europe and Asia. Eligible patients were aged 15 years or older, with at least one mediastinal lesion of 1 cm or longer in the short axis that required diagnostic bronchoscopy. Participants were randomly assigned (1:1) using a block randomisation scheme generated by a computer (block size of four participants based on a random table from an independent statistician) to the combined use of EBUS-TBNA and transbronchial mediastinal cryobiopsy (combined group) or EBUS-TBNA alone (control group). Because of the nature of the intervention, neither participants nor investigators were masked to group assignment. The coprimary outcomes were differences in procedure-related complications and diagnostic yield (defined as the proportion of participants for whom mediastinal biopsy led to a definitive diagnosis), assessed in the full analysis set, including all the patients who met the eligibility criteria and had a biopsy. A fully paired, intraindividual diagnostic analysis in participants who had both needle aspiration and mediastinal cryobiopsy was conducted, in addition to interindividual comparisons. This trial is now complete and is registered with ClinicalTrials.gov, NCT04572984. Findings Between Oct 12, 2020, and Sept 9, 2021, 297 consecutive patients were assessed for eligibility and 271 were enrolled and randomly assigned to the combined group (n=136) or the control group (n=135). The addition of cryobiopsy to standard sampling significantly increased the overall diagnostic yield for mediastinal lesions, as shown by both interindividual (126 [93%] of 136 participants in the combined group vs 109 [81%] of 135 in the control group; risk ratio [RR] 1·15 [95% CI 1·04–1·26]; p=0·0039) and intraindividual (126 [94%] of 134 vs 110 [82%] of 134; RR 1·15 [95% CI 1·05–1·25]; p=0·0026) analyses. In subgroup analyses in the intraindividual population, diagnostic yields were similar for mediastinal metastasis (68 [99%] of 69 participants in the combined group vs 68 [99%] of 69 in the control group; RR 1·00 [95% CI 0·96–1·04]; p=1·00), whereas the combined approach was more sensitive than standard needle aspiration in benign disorders (45 [94%] of 48 vs 32 [67%] of 48; RR 1·41 [95% CI 1·14–1·74]; p=0·0009). The combined approach also resulted in an improved suitability of tissue samples for molecular and immunological analyses of non-small-cell lung cancer. The incidence of adverse events related to the biopsy procedure did not differ between trial groups, as grade 3–4 airway bleeding occurred in three (2%) patients in the combined group and two (1%) in the control group (RR 0·67 [95% CI 0·11–3·96]; p=1·00). There were no severe complications causing death or disability. Interpretation The addition of mediastinal cryobiopsy to standard EBUS-TBNA resulted in a significant improvement in diagnostic yield for mediastinal lesions, with a good safety profile. These data suggest that this combined approach is a valid first-line diagnostic tool for mediastinal diseases. Funding National Natural Science Foundation of China.