A New Combination Strategy Shows Up to 31.3% Weight Loss in Preclinical Study

The modern obesity drug race has been defined by one word: incretins.

Blockbuster GLP-1 receptor agonists from companies like Novo Nordisk and Eli Lilly have reshaped both waistlines and Wall Street expectations. Drugs such as Semaglutide and Tirzepatide are no longer niche metabolic therapies - they are global economic forces.

Now, a newer entrant is attempting to carve out space in that arena with a fresh strategy.

Hong Kong and Boston based company Insilico Medicine has announced the nomination of ISM0676, an oral GIP receptor (GIPR) antagonist, as a preclinical candidate. In diet-induced obese humanized mice, the compound demonstrated up to 31.3% body weight loss when combined with semaglutide over 27 days. On its own, it achieved a 10.4% reduction, a modest one compared to GLP-1 standouts, but notable for a first-generation small molecule antagonist.

The headline number, 31.3%, is what grabs attention, but the real story lies beneath it...

The Strategic Pivot: GIPR Antagonism

The GIP receptor has become one of the most intriguing and debated targets in metabolic medicine. Tirzepatide, for example, activates both GLP-1 and GIP receptors. Insilico’s approach flips that script by focusing on blocking GIPR rather than stimulating it. 

This matters because GIP signaling is implicated in fat storage, insulin secretion, appetite regulation, and bone metabolism. Some researchers argue that antagonizing GIPR may counteract adipose deposition and improve metabolic efficiency, particularly when paired with GLP-1 therapies. In theory, this combination could mitigate several limitations of GLP-1 monotherapy, including lean mass loss, plateauing efficacy, and rapid weight regain after discontinuation.

For example, there was a review article back in 2020 where Jonathan Campbell, Associate Professor in Medicine at Duke University, and the single author of the paper, noted that both GIPR agonism and antagonism had been shown to reduce body weight in preclinical models, and that explaining this paradox requires a deeper understanding of GIP biology, a sign the field hasn’t fully agreed on mechanisms. 

That’s still theory. But the mouse data from Insilico Medicine suggest synergy worth investigating.

The AI Angle: Speed and Scale

Beyond pharmacology, the announcement is also a test case for AI-enabled drug design.

Insilico reports that ISM0676 was nominated just 14 months after project initiation, with fewer than 200 molecules synthesized. In a traditional drug discovery setting, that timeline would raise eyebrows — in a good way. The company credits its generative chemistry engine, Chemistry42, for accelerating hit identification and optimization.

Drug development remains brutally attritional. Preclinical success often evaporates in Phase I or II. But if AI platforms can consistently compress early discovery timelines while maintaining quality, that would mark a structural shift in how pipelines are built. But the proof, as always, will come in human data…

The Bigger Context: A $100+ Billion Question

The obesity treatment market is projected to grow up to $100 billion by 2030, as per Goldman Sachs’s estimates. Yet today’s leading therapies, despite dramatic weight loss results, are not perfect. Lean mass preservation, long-term sustainability, tolerability, and adherence remain active challenges, as well as side effects potentially linked to such drugs. 

If combination regimens become the next frontier, much like in oncology, companies that can design modular add-on therapies may find strategic leverage rather than direct competition. That appears to be Insilico’s positioning with this new move into cardiometabolic space: not directly competing with or replacing GLP-1s, but enhancing them in some ways.

It is important to emphasize: these are preclinical mouse data.

Human metabolism is way more complex, and GIP biology has produced conflicting clinical interpretations in the past. Combination metabolic therapies may introduce regulatory and safety complexities that extend well beyond weight reduction percentages.

Still, early signals matter especially when they suggest a potentially differentiated mechanism in a market dominated by a handful of incumbents.

An up to 31.3% preclinical result does signal that the incretin era may be entering its ‘second act,’ one where antagonists, combination logic, and AI-designed small molecules may all play meaningful roles.

Topic: AI in Bio