Merck reports statistically significant efficacy and acceptable safety for efinopegdutide in a Phase 3 trial for MASH.
Following Phase 2b data on MASH and fibrosis, Merck advances efinopegdutide to later-stage testing. Success requires meeting primary efficacy endpoints in Phase 3.
Across 6 frontier models answering independently, the panel thinks this is unlikely: the median is 33%.
They broadly agree, 20 points between Gemini at 42% and DeepSeek at 22%.
Nothing is settled until 12 Nov 2029, when the outcome is checked against sources and every number above gets a permanent score.
What each one said
-
GeminiGoogle
Phase 3 MASH trials require histological endpoints that often suffer from high placebo responses and lengthy biopsy-driven timelines.
- Merck needs to initiate and read out a 52-week biopsy Phase 3 trial by late 2029, leaving almost zero margin for recruitment delays.
- GLP-1/GIP and dual agonists hit MASH resolution easily, but historic Phase 3 success on the dual fibrosis endpoint is under 50%.
- Gastrointestinal tolerability and competitive titration schedules from tirzepatide set a high bar for acceptable safety.
42% -
GPTOpenAI
The odds hinge on whether Phase 2 fibrosis/MASH effects survive a larger, longer biopsy-based Phase 3 program.
- Merck must select dose, enroll biopsy-confirmed MASH patients, and complete a multi-year histology trial before 12 November 2029.
- Late-stage MASH trials have high attrition; histologic endpoints and tolerability have sunk otherwise promising metabolic drugs.
- Confidence would rise sharply if Merck starts Phase 3 by 2027 with a clean dose-ranging package and FDA-aligned endpoints.
41% -
QwenAlibaba
Strong Phase 2b effect size vs. MASH Phase 3's historically brutal ~65-70% failure rate.
- Phase 2b showed ~65% MASH resolution at high dose vs ~20% placebo—a large delta, but MASH Phase 3 has killed obeticholic acid, selonsertib, cenicriviroc, and elafibranor…
- If Phase 3 initiated mid-2025 with a 48–72 week treatment period, a readout by late 2028 or 2029 is feasible, so the timeline clears.
- The dual GLP-1/glucagon mechanism is gaining validation (survodutide Phase 2 also strong), which nudges above the base MASH failure rate, but biopsy endpoint variability…
37% -
ClaudeAnthropic
Hinges on whether Merck even starts and completes a Phase 3 MASH trial for efinopegdutide by late 2029.
- Efinopegdutide's Phase 2b MASH data showed efficacy but GI tolerability issues that could limit dosing and complicate a cleanly positive Phase 3 readout.
- Phase 3 MASH trials typically require 2+ years of histology-confirmed enrollment plus analysis, making a trial start-to-readout within 3 years tight unless already…
- GLP-1/glucagon dual agonists face a competitive MASH landscape (resmetirom approved, survodutide and others ahead), so Merck may deprioritize or delay this program.
28% -
GrokxAI
Merck must start and finish a positive histology Ph3 for the dual agonist inside a 3-year window.
- Ph3 MASH trials need ~18 mo enrollment plus 52–72 wk treatment; a 2027 start barely fits the 2029-11 cutoff.
- Historic NASH Ph3 success is ~20-25%; dual-agonist fat-reduction data lift this modestly above the base rate.
- A 2027 delay, weak fibrosis signal, or stronger competitor readouts would make Merck drop or miss the date.
28% -
DeepSeekDeepSeek
Whether a Phase 3 MASH readout for efinopegdutide lands by late 2029, and clears both histology and tolerability bars.
- Merck started a Phase 3 MASH program only after 2024-2025 Phase 2b data; a 52-week biopsy endpoint plus enrollment typically needs 3-4 years from start.
- MASH Phase 3 precedent is brutal: resmetirom succeeded, but several GLP-1/other agents missed or showed borderline histology, so base rate for a clean win is well under…
- Efinopegdutide's dual GLP-1/glucagon profile may add weight loss but raises GI tolerability risk, the exact axis that sank prior MASH candidates.
22%
6 of 6 models answered · 20 points between the highest and lowest. None was shown the market price.
Question sourced from a news sweep on 13 Nov 2026. Forecast by google/gemini-3.8-flash, anthropic/claude-sonnet-5, openai/gpt-5.6-terra, x-ai/grok-4.6, deepseek/deepseek-v4.1-flash, qwen/qwen3.8-max-0902 via OpenRouter.