Design: International, open-label, parallel-group randomised trial
Population: 627 adults with peripheral lesions ≤30 mm on radial EBUS, Japan, Malaysia, South Korea, Taiwan
Primary outcome: Histology-based diagnostic yield
Effect: Eccentric/adjacent 77% vs 65% (difference 12.4%, 2.8–22.0); overall 83% vs 75%

Cryo-RCT, published in July 2026, randomised 627 patients at six centres in Japan, Malaysia, South Korea and Taiwan with peripheral lesions up to 30 mm seen on radial endobronchial ultrasound. They had stand-alone cryobiopsy with a single-use probe (one to three specimens) or conventional sampling with forceps, needles or both (five or more specimens where feasible).

Where the probe sat adjacent to or eccentric to the lesion — the harder cases — histological yield was 77% with cryobiopsy against 65% (difference 12.4 percentage points, 95% CI 2.8–22.0). Across all patients it was 83% against 75%. Moderate bleeding was twice as common with cryobiopsy (38% vs 19%), though severe bleeding was rare in both (1%). Pneumothorax occurred in 2% versus under 1%, with five chest drains in the cryobiopsy group and none in the other.

The gain is largest exactly where conventional sampling struggles, and it may save repeat procedures. The trade-off is more bleeding and a small pneumothorax risk, and cryoprobes add cost.

Consider cryobiopsy when radial EBUS shows an eccentric or adjacent probe position.
Prepare for moderate bleeding, which occurred in over a third of cryobiopsies.
Warn patients of a small pneumothorax risk, about 2%, which may need a chest drain.
For concentric lesions, the advantage over conventional sampling is smaller.

Why it matters

Better tissue from the first bronchoscopy can avoid a second procedure and speed treatment.

The statistics, in plain English

A 12.4-point difference with an interval from 2.8 to 22.0 means cryobiopsy is very likely better in eccentric lesions, though the true gain may be modest. The trial was open label, but histological yield is an objective endpoint.