BeyondSpring's second quarter was not a financial print. It was a clinical one. The 47-patient Phase 2 303 Study of plinabulin plus docetaxel and pembrolizumab in metastatic non-small cell lung cancer (NSCLC) patients who had already progressed on first-line immune checkpoint inhibitors (ICI) reported at the American Society of Clinical Oncology (ASCO) 2026 meeting a 58.0% two-year overall survival rate and a 78.1% one-year overall survival rate with median overall survival not yet reached at 28.8 months of follow-up, a long-tail survival number in a setting where the standard of care, single-agent docetaxel, has historically delivered median overall survival closer to ten months. That is the load-bearing data point of the quarter. Alongside it, the company seated a new chief executive officer, Min Qiu, effective July 1, 2026, and named a new chief financial officer, Na Li, with an explicit mandate to align the organization around the confirmatory DUBLIN-4 Phase 3 program in NSCLC post-ICI.
The financial print is the inverse of the clinical one. The company posted a consolidated net loss of $5.7 million in the second quarter, a $4.3 million net current liability position at quarter-end, and an explicit going-concern qualification in the same 10-Q. Cash plus short-term investments attributable to continuing operations stood at $6.5 million as of June 30, 2026, down from $12.6 million at year-end 2025; the company consumed $7.0 million of operating cash in the first half. DUBLIN-4 has not yet been initiated. The $0.72 share price at the August 21 close, against a December 2025 high of $2.35, reflects the gap between an investable clinical asset and a balance sheet that has not yet been financed for its pivotal trial.
The open question is not whether plinabulin works. The 303 Study reads as a confirmatory signal that the DUBLIN-3 result is real. The open question is whether the company can fund DUBLIN-4, a 442-patient randomized confirmatory Phase 3 with overall survival as the primary endpoint, before the runway ends. The falsification framework, accordingly, is anchored on the financing clock rather than the next data readout.