RO7795081

Roche is testing RO7795081 head-to-head against semaglutide (the active ingredient in Novo's Ozempic and Wegovy) and placebo in a Phase 2 type 2 diabetes trial [1]. A separate Phase 2 obesity trial has already completed [2], and that readout is the next catalyst. RO7795081 is the development code for CT-996 (also known as tifeglipron / RG6652), the oral small-molecule GLP-1 receptor agonist Roche acquired through Carmot Therapeutics [6][12][13]. It is a mono-GLP-1 agonist, not a dual agonist. Carmot's other main asset, the injectable dual GLP-1/GIP agonist CT-388 (enicepatide), is a separate Roche obesity program that already reported 22.5% placebo-adjusted weight loss at 48 weeks in January 2026 [14]. For a company with no approved obesity drug and $72 billion in annual revenue [11], the Carmot pipeline (one injectable dual, one oral mono) is Roche's vehicle for catching up to Novo and Lilly in the biggest therapeutic category of the decade.

RO7795081 is a novel compound with no approvals anywhere. The clinical package includes two Phase 2 trials and three Phase 1 studies. The Phase 2 T2D trial (NCT07112872, n=330) is active not recruiting, meaning enrollment is closed and the trial is heading toward readout [1]. The Phase 2 obesity trial (NCT07081958, n=352) is marked completed [2]. Phase 1 work covers standard regulatory boxes: a renal impairment study still recruiting [3], a completed drug-drug interaction (DDI) study with itraconazole [4], and an active DDI study with statins [5]. A DDI study tests whether co-administered drugs change how the body processes RO7795081, which matters for the eventual label. No FDA designations have been granted, which is unsurprising since this is a crowded therapeutic category where regulatory shortcuts rarely apply. If the obesity Phase 2 data lands in late 2026 or early 2027, a Phase 3 start is the next decision point. The timing is tight. Roche needs these data to lock in a Phase 3 design before Lilly's orforglipron reads out its Phase 3 program and resets expectations for oral weight-loss drugs [10].

RO7795081 is CT-996 (tifeglipron, RG6652), a once-daily oral small-molecule GLP-1 receptor agonist from the Carmot Therapeutics acquisition that Roche closed in December 2023 for $2.7 billion upfront [6][12]. GLP-1 is the gut hormone that signals satiety and triggers insulin release after a meal. Semaglutide, tirzepatide, and liraglutide all hit this pathway as injectable peptides. CT-996 is specifically engineered as a signal-biased GLP-1 receptor agonist: it activates the cAMP arm of receptor signaling with minimal β-arrestin recruitment and little receptor internalization, a design intended to sustain signaling and reduce the dose-limiting GI toxicity that plagues the class [13]. The compound has a plasma half-life of roughly 17-22 hours, supporting once-daily oral dosing [13]. In the Phase 1 obesity study presented at ADA and EASD 2024, four weeks of CT-996 produced a placebo-adjusted weight loss of about 6.1% (7.3% absolute vs 1.2% on placebo) with mostly mild-to-moderate GI adverse events [13]. Carmot's lead injectable asset, CT-388 (enicepatide), is a separate GLP-1/GIP dual agonist Roche is advancing in parallel; it reported 22.5% placebo-adjusted weight loss at 48 weeks in January 2026 [14]. So Roche is running two distinct mechanisms (oral mono vs injectable dual) from the same acquisition. The GLP-1 pathway itself is validated as thoroughly as any target in modern medicine. Semaglutide pulled roughly $27 billion in 2024 revenue for Novo Nordisk across Ozempic, Wegovy, and Rybelsus, and tirzepatide approached $17 billion for Eli Lilly across Mounjaro and Zepbound [7]. The question is not whether GLP-1 works. The question is whether CT-996's biased-agonism design delivers meaningfully better tolerability or dosing than Lilly's orforglipron and Novo's oral semaglutide.

The Phase 2 T2D trial (NCT07112872) randomizes 330 patients with type 2 diabetes to RO7795081, placebo, or semaglutide, with change in HbA1c (a long-term blood sugar marker) from baseline to week 30 as the primary endpoint [1]. Including a semaglutide arm at this stage is aggressive. Most Phase 2 metabolic trials use placebo only, which makes efficacy look better and defers the head-to-head comparison with injectable peptides until Phase 3. By putting semaglutide in the Phase 2, Roche is either confident the drug can hold its own, or wants an early read on positioning. The obesity trial (NCT07081958) tested once-daily oral RO7795081 in 352 patients with obesity plus at least one weight-related comorbidity, with percent change in body weight as the primary endpoint over a treatment duration of roughly 38 weeks [2]. This is the trial that will drive the market reaction. The 38-week readout is shorter than Novo's STEP program (68 weeks) and Lilly's SURMOUNT program (72 weeks), which means weight loss will look smaller in absolute terms. Analysts will annualize or extrapolate, which creates interpretive noise. The semaglutide comparator in T2D is the cleanest data point to watch, since it anchors Roche's HbA1c reduction directly against the standard.

Our model estimates a 21% chance this drug is eventually approved. It starts from the historical base rate for Phase 2 drugs in this area (about 35%), then adjusts using ten facts about the trial and sponsor. What moves the number most: it is helped by an unusually multi-arm design (9 arms), larger-than-typical enrollment for this phase, and more secondary endpoints than usual; it is held back by weak or limited earlier-phase results. The other facts land near average for this stage, so they leave the estimate roughly where the base rate put it.

Oral GLP-1 agonists carry specific class risks. GI tolerability (nausea, vomiting, constipation) is universal and dose-limiting across the class. Liver toxicity killed Pfizer's danuglipron program in 2025 after transaminase elevations (markers of liver stress - enzymes released from damaged liver cells into the bloodstream) in Phase 2 [9]. Hepatic signals are the first thing to look for in Roche's Phase 2 safety data. In practice this means watching for ALT or AST elevations (alanine aminotransferase and aspartate aminotransferase - liver enzymes whose elevated blood levels signal potential liver damage). Efficacy risk: if weight loss at the primary endpoint lands below 10%, the drug is behind orforglipron (around 15% at week 36 in Phase 2 obesity data [10]) and will struggle to differentiate in a Phase 3 field with tirzepatide follow-ons and triple agonists coming. Commercial risk is real even if the drug works. By 2028, the oral weight-loss market will likely include Lilly's orforglipron, Novo's oral semaglutide at obesity dose (if approved for that indication), and possibly amycretin or retatrutide follow-ons. Payers will cover the leaders. A fourth-to-market oral GLP-1 with ordinary efficacy earns coverage limitations and prior authorization friction, which collapses net price. Execution risk is modest. Roche can run trials, and the Phase 2 package shows a thorough regulatory workup including renal and drug-drug interaction studies [3][4][5].

Worth watching, with specific triggers. The obesity Phase 2 (NCT07081958) is completed, which means data is already in Roche's hands and the next material disclosure is likely a top-line press release followed by a conference presentation [2]. EASD 2026 (September 2026 - already occurred) did not include a public CT-996/RO7795081 Phase 2 readout based on available coverage. The next anticipated venue is ObesityWeek 2026 in November, possibly preceded by a Q3 earnings-call top-line. The numbers to watch are percent weight loss at the primary endpoint and the proportion of patients hitting at least 10% weight loss. Above 12% body weight reduction puts RO7795081 in the conversation with orforglipron [10]. Below 8% is a problem, especially since Roche's own injectable CT-388 (enicepatide) hit 22.5% at 48 weeks [14] - an oral has to be competitive on efficacy or lean on the compliance/convenience advantage of pills to justify a Phase 3 commitment. Also watch for any ALT or AST elevations in the safety tables, which would echo danuglipron's failure mode [9]. For Roche, this is the most important pipeline readout outside oncology for 2026. The company's $72 billion revenue base [11] is overexposed to maturing franchises including Ocrevus and Hemlibra. A credible oral GLP-1, layered on the CT-388 injectable, would reshape the growth story. A miss on RO7795081 does not kill the obesity thesis but narrows Roche's oral strategy and raises pressure on the Carmot acquisition economics.