CSPC Ouyi Pharmaceutical has registered NCT07796477, a Phase 2 dose-finding trial of SYH2069 Injection in 240 adults with type 2 diabetes, with tirzepatide 15 mg as the active comparator. The double-blind, placebo- and active-controlled study plans four SYH2069 dose groups, four matching placebo…
A Phase 2 Head-to-Head With Tirzepatide in Type 2 Diabetes
CSPC Ouyi Pharmaceutical Co., Ltd. has posted a not-yet-recruiting Phase 2 trial, NCT07796477 , of SYH2069 Injection in 240 adults with type 2 diabetes mellitus T2DM , using tirzepatide 15 mg as the active comparator. The trial is designed as a multicenter, randomized, double-blind, placebo- and active-controlled, dose-finding study. The registry record was last updated on 2026-09-01. The posting appears to be the first disclosed Phase 2 dose-finding study of SYH2069 Injection in T2DM, and the record contains no prior efficacy or safety data for the molecule.
The choice of comparator is the most informative detail in the listing. Tirzepatide, the dual GIP and GLP-1 receptor agonist approved for T2DM, is the efficacy reference in the incretin class, and 15 mg is the highest dose in its approved maintenance regimen. Selecting that dose as an active control signals an intention to compete at the top of the glycemic control market rather than merely to beat placebo. A dose-finding study framed this way is designed to answer two questions at once: whether SYH2069 lowers glucose and body weight, and whether the effect sizes come within range of the current class leader. A showing anywhere in tirzepatide's effect-size neighborhood, even in a Phase 2 frame, would be a commercially meaningful result for a developer with no published clinical data on the molecule.
The stage matters. A trial that is not yet recruiting, with a record updated on 2026-09-01, may still be under regulatory or ethics review, and a registry posting is a declaration of intent rather than a result. Registration listings are not peer-reviewed evidence, and the absence of prior data on SYH2069 means the molecule's mechanism, its chemical identity, and even its status as a peptide are unconfirmed. The architecture of the trial, with its incretin-class comparator, titration design, and immunogenicity assessment, is consistent with a peptide-based metabolic agent. Consistency is not proof.
The trial allocates participants across nine arms in a 4:1:4:1:4:1:4:1:4 randomization ratio spanning four SYH2069 dose groups, four SYH2069 matching placebo groups, and one tirzepatide 15 mg active-control group. The planned group sizes are:
Total planned enrollment is 240 participants. In ratio terms, each unit of the allocation represents 10 participants: the four dose groups and the tirzepatide arm each account for four units, and each placebo group accounts for one. Pooled across the four placebo arms, the placebo sample is 40 participants, numerically equal to each SYH2069 dose group and to the active-control group. When placebo arms are pooled for analysis, the placebo-adjusted comparison at any given dose level is therefore 1:1, not 4:1.
The decision to run four separate placebo groups rather than one larger placebo arm is a blinding decision as much as a statistical one. Different dose levels of an injected product may require different injection volumes, titration schedules, or device presentations, and each needs a matching placebo to keep participants, investigators, and site staff blind to dose assignment. Matching placebo to each dose level is the standard way to preserve the blind across a range of administration requirements. The registry does not state whether the four placebo groups are intended to be pooled in the analysis, but the arithmetic of the design makes pooling the natural read.
Randomization will be stratified by baseline HbA1c , split at ≤8.5% or 8.5%, and by prior metformin use, yes or no. The chemistry of the HbA1c measurement explains why the split is placed there. Hemoglobin undergoes nonenzymatic glycation at the N-terminal valine of its beta chains, and the fraction of glycated hemoglobin is proportional to the average glucose concentration over the preceding two to three months, roughly the 120-day lifespan of the red blood cell. Because the measurement integrates fasting and postprandial glucose across that window, baseline HbA1c is both an eligibility gate and a predictor of response. Patients entering above 8.5% tend to carry more beta-cell dysfunction and a heavier glucotoxic burden, and they typically show larger absolute reductions when an effective agent is added. Stratifying at that cut point prevents an uneven distribution of these high-responder patients across the nine arms, which would otherwise bend the dose-response curve for reasons unrelated to the drug.
The metformin stratum is the second guard against imbalance. Metformin suppresses hepatic glucose production, improves peripheral insulin sensitivity to a smaller degree, and lowers HbA1c by roughly one to one and a half percentage points as monotherapy. A patient whose glycemia remains uncontrolled on metformin has usually progressed further in disease course, with lower endogenous insulin reserve, than a patient managed with diet and exercise alone. Randomizing within the yes/no metformin stratum distributes both the background drug effect and the underlying disease severity evenly across the four dose groups, the four placebo groups, and the tirzepatide arm, so the incremental effect of SYH2069 can be read within each stratum.
The two pre-randomization periods do analogous work. The 2-week screening period allows central laboratory confirmation of the HbA1c eligibility value, review of the inclusion and exclusion criteria, and informed consent. The 3-week run-in period that follows is longer than a single confirmatory visit would require, and the extra time has a purpose: washout or standardization of background glucose-lowering medication, a check of adherence, and a buffer against the behavior change that comes with being observed. Patients who join a diabetes trial often improve their diet and physical activity during the first weeks under observation, and a run-in lets some of that change occur before baseline is recorded, so the treatment effect measured from baseline is not inflated by the act of enrolling. The five-week interval from first contact to randomization is a standard trade in this class: a longer recruitment timeline in exchange for cleaner baselines.
The study has four periods: a 2-week screening period, a 3-week run-in period, a 28-week treatment period, and a 3-week safety follow-up period. The 28-week treatment period is long enough for HbA1c to reach a new steady state after dose titration: the measurement reflects the preceding two to three months of glycemia, and 28 weeks of treatment provides several months of observation after that plateau is reached. The 3-week safety follow-up captures early off-treatment changes, though it is short relative to the pharmacological washout of long-acting incretin agents.
All treatment groups will implement dose titration to achieve the target dose. For an incretin-based therapy, this is the standard mechanism for managing dose-dependent gastrointestinal adverse events, which are most prominent when treatment starts or escalates. The explicit requirement that every arm, including the tirzepatide comparator, titrate suggests the sponsor intends to compare agents under matched escalation conditions. Whether the titration schedules are identical across all nine arms is not stated in the registry, and this will matter when adverse event rates are read.
The design gives the study three analytical jobs. First, the four SYH2069 dose groups support a dose-response estimate across the selected dose range. Second, the tirzepatide 15 mg arm anchors those responses to a clinically meaningful reference, letting the sponsor judge whether any SYH2069 dose operates in the effect-size range of the class leader. Third, the placebo groups provide a within-trial adjustment for regression to the mean, the placebo effect, and the lifestyle changes that accompany trial participation, all of which inflate apparent treatment responses in uncontrolled diabetes cohorts.
The placebo groups deserve close reading. Ten participants per arm, 40 in total, is a modest control sample. With 40 placebo subjects followed for 28 weeks, the estimate of background change in HbA1c will carry wide confidence intervals. The active comparator therefore carries much of the interpretive weight in this design, and tirzepatide's well-characterized effect profile gives the sponsor a stable yardstick. This is a recurring pattern in diabetes development: the placebo group exists to satisfy regulatory expectations and to anchor safety comparisons, while the active control drives the go/no-go decision.
The double-blind design is only meaningful if the titration schedules preserve blinding, which is a genuine challenge when adverse event profiles can unblind investigators and participants. If SYH2069 and tirzepatide produce different rates of nausea, vomiting, or diarrhea during escalation, differential dropout could bias the completer population and distort the efficacy comparison. Gastrointestinal tolerability differences between an investigational agent and an established one are common, and the registry does not describe how the protocol handles these risks. The execution quality of the blinding and dropout strategy will only be visible when results appear.
The outcomes the Phase 2 study is designed to evaluate are:
The registry does not list the specific efficacy endpoints or their measurement timepoints. For a diabetes program of this size and duration, the efficacy readout typically centers on change from baseline in HbA1c, with body weight and fasting glucose as supporting measures, but the absence of detail means the analysis plan cannot yet be evaluated. The same applies to the safety, PK, and immunogenicity protocols: the domains are named, the specifics are not.
What the study cannot do is as important as what it can. A 28-week, 240-participant trial is far too small and too short to characterize the long-term safety of a chronic metabolic therapy. It is not powered to detect rare adverse events, cardiovascular outcomes, or durability of effect beyond the treatment period. The efficacy data from this trial will support dose selection and the decision to advance, not registration claims.
The regulatory frame for diabetes drugs has been fixed since 2008, when the FDA required sponsors to exclude an excess cardiovascular risk before a new agent reaches the market. The standard architecture uses a two-sided 95% confidence interval on the risk ratio of major adverse cardiovascular events, usually defined as cardiovascular death, nonfatal myocardial infarction, and nonfatal stroke. Under that framework, the upper bound of the interval must be below 1.8 for approval; if it falls between 1.3 and 1.8, the sponsor must complete a dedicated cardiovascular outcomes trial after marketing begins. NCT07796477 is not built to answer any of this. At 240 participants and 28 weeks of treatment, its cardiovascular contribution is limited to the systematic collection of adverse events for a later, program-level meta-analysis. The Phase 2 dataset will support a go/no-go decision, not a registration claim.
The efficacy standard in a dose-finding study is different. The question is whether the data produce a clear dose-response relationship, a tolerable titration profile, and pharmacokinetic evidence that the doses carried into Phase 3 are defensible. The 28-week treatment horizon matters because HbA1c reflects roughly three months of glycemia, so the measurement at week 28 describes a steady state reached several weeks after the final…
Peptides referenced: Tirzepatide, GLP-1.
Related reading: Quebec Court Bars Canlab Research From Selling Unauthorized Peptides, Seventh Victorian Case Links Counterfeit Retatrutide to Liver Toxicity, MHRA approves orforglipron, Lilly's oral GLP-1 drug, for UK use, AI-designed oral peptides get first commercial test in $100M deal.