Tenax's Only Drug Fails Phase 3 LEVEL Trial; Stock Loses 90% of Its Value

Tenax Therapeutics (Nasdaq: TENX) announced Monday that TNX-103, its oral levosimendan program and the company’s only clinical asset, failed the primary endpoint of the Phase 3 LEVEL trial in pulmonary hypertension with preserved ejection fraction. The stock lost roughly 90% of its value intraday, shrinking the company’s market cap from about $500 million to approximately $50 million. TENX closed at $13.44 on Friday.

The numbers leave no room for interpretation. TNX-103 improved 6-minute walk distance by 14.0 meters versus 10.4 meters for placebo, a 3.5-meter treatment difference with a p-value of 0.63. The key secondary endpoint, the KCCQ symptom score, improved by 6.6 points versus 6.5 points for placebo, a difference of 0.1. Both primary and secondary endpoints missed cleanly in the 241-patient trial.

The subgroup story. Management highlighted prespecified subgroup data to argue the drug showed activity in sicker patients. Among the 119 patients with baseline walking distance below 333 meters, TNX-103 produced a 26.3-meter placebo-adjusted improvement (nominal p=0.0112). Patients over 71 years old showed a 27.1-meter benefit. The heart failure biomarker NT-proBNP fell 49% relative to placebo across the full population (nominal p<0.0001). But the company explicitly noted that subgroup and biomarker analyses do not establish efficacy, and the FDA had previously agreed that a single Phase 3 trial reaching p=0.01 could support an NDA. The actual primary endpoint p-value was 0.63, two orders of magnitude short.

What comes next. Tenax plans to request a Type C meeting with the FDA to discuss an enriched patient population strategy and revisions to the ongoing LEVEL-2 trial, the company’s second registrational study. Whether regulators will accept a protocol reboot after a cleanly negative trial is uncertain. Full LEVEL data was presented in a Late-Breaking session at ESC Congress 2026 in Munich on August 29, 2026, with additional analyses showing a 26.3-meter placebo-adjusted 6MWD benefit (p=0.0112), a 47% NT-proBNP reduction, and a 4.9 mmHg RVSP improvement in the subgroup with baseline walking distance below 333 meters. The overall primary endpoint missed cleanly at p=0.63.

Why this matters beyond Tenax. PH-HFpEF affects over 1.6 million people in North America and has no FDA-approved therapies. Merck’s sotatercept showed encouraging signals in the Phase 2 CADENCE trial in a related HFpEF population at ACC 2026, and that program is now the most advanced candidate in the class, but the field remains wide open. Tenax was the most advanced oral therapy in development.

For Tenax specifically, this is an existential outcome. TNX-103 is the company’s only drug. TNX-101 (the intravenous formulation) and TNX-102 are the same active ingredient, levosimendan, in different delivery forms. There is no pipeline diversification. The cash runway and any remaining value in the stock now depend entirely on whether the FDA sees a path forward in the LEVEL-2 enrichment strategy.

For a primer on how to interpret the numbers in a readout like this one, see our guide to clinical trial press releases.

What It Means for the Trade

The 90% intraday collapse is the market doing binary-event arithmetic correctly: a one-asset company whose only registrational trial missed at p=0.63 is worth its cash and its option value on a rescue, nothing more. From a $50 million market cap, the trade is now a lottery ticket on the Level-2 enrichment strategy — and the honest base rate for regulatory acceptance of a protocol reboot after a cleanly negative primary endpoint is low. The FDA agreed ex ante that a single trial at p=0.01 could support an NDA; the trial delivered p=0.63.

The Type C meeting is the entire catalyst calendar. If the FDA signals openness to an enriched design in Level-2 — restricted to the low-baseline-walk population where the 26.3-meter nominal signal lives — TENX gets a second life and the option reprices violently. If the FDA says no, the company is a cash shell.

Mechanism and Market Context

Levosimendan is a calcium sensitizer that enhances cardiac contractility without raising intracellular calcium — plus vasodilation through potassium-channel opening, which suits a population whose problem is a stiff, non-compliant ventricle rather than a weak one. In PH-HFpEF, the pulmonary hypertension is a consequence of left-sided filling pressures, so a drug that improves relaxation and unloads the ventricle is mechanistically sensible. The problem the Level trial exposed is trial-population heterogeneity, not necessarily biology: in a disease defined by a clinical syndrome rather than a discrete mechanism, the patients who entered the trial were heterogeneous enough that a modest real effect in a subgroup drowns in the full-population noise. The numbers show exactly that — both arms improved substantially, and the 3.5-meter difference is indistinguishable from zero, while the sicker third of patients showed a 26.3-meter placebo-adjusted gain and a 49% NT-proBNP drop, a biomarker effect too large to dismiss as noise even though it cannot support approval.

Risks and What Could Go Wrong

1. FDA declines the reboot. A Type C meeting that closes the enrichment door ends the story; remaining value is cash minus wind-down costs.
2. Level-2 repeats the pattern. Even if enrichment is allowed, the subgroup signals are nominal; a second miss at the enriched design is a true terminal event.
3. Dilution on the way to any readout. Funding a second registrational trial from a $50 million cap means brutal raise economics for current holders.
4. Field-level caution. Other PH-HFpEF programs just inherited Tenax’s lesson: full-population endpoints in a heterogeneous syndrome are high-variance.

Net: a terminal-adjacent event with one narrow path left. Position accordingly: option-sized or absent.