How Clinical Trials Work for Peptide Therapeutics

The short version

Before any peptide drug reaches a pharmacy shelf, it travels through a structured pipeline designed to answer three questions in order: is it safe enough to give to humans? does it work? and does it work well enough to justify the risks in everyday clinical use? Each phase of clinical development is organized around one of those questions, with increasingly large and diverse patient populations and increasingly rigorous controls on bias. Regulatory agencies such as the FDA and the EMA review the full package of evidence before granting a marketing authorisation, and post-market studies continue to fill gaps that pivotal trials cannot address.

Preclinical development

Preclinical work happens entirely outside human subjects. For a peptide therapeutic the work typically falls into three streams.

• In-vitro characterization. Receptor binding assays, functional assays (cAMP generation, beta-arrestin recruitment), cell-viability screens, and metabolic stability testing in plasma or liver microsomes. These establish whether the molecule has the desired pharmacological activity and a reasonable half-life before significant resources are committed to animal work.
• Animal pharmacology and efficacy models. Rodent models of the target disease (e.g., diet-induced obese mice for metabolic studies, STZ-induced diabetic rats for glucose-lowering work) test whether efficacy signals translate from a cell dish to a whole organism.
• IND-enabling toxicology. Good Laboratory Practice (GLP) toxicology studies in two species (typically rat and a non-rodent, often primate) characterize dose-limiting toxicities, the no-observed-adverse-effect level (NOAEL), and reproductive or genotoxic signals. These data, together with pharmacokinetic studies, form the core of the Investigational New Drug (IND) application submitted to the FDA before human trials begin.^[5]

Peptides present some specific challenges at this stage. Their susceptibility to proteolytic degradation means that bioavailability data from animals can be a poor predictor of human exposure if structural modifications (fatty acid conjugation, amino acid substitution, PEGylation) are not finalized before IND-enabling studies. Species differences in receptor homology also affect the translation of efficacy signals; a rodent GLP-1 receptor agonism study, for example, requires careful interpretation because rodent and human receptor pharmacology differ in detail.

Phase 1: first-in-human

Phase 1 is the first administration of the drug to humans. The primary goal is safety and pharmacokinetics (PK), not efficacy.^[3]

For most peptide metabolic drugs, Phase 1 trials enroll healthy adult volunteers rather than patients. This allows clean PK data without the confounders of active disease or polypharmacy. Exceptions apply in oncology, where patient populations are used from the outset.

A typical Phase 1 peptide trial structure:

• Single ascending dose (SAD). Starting dose derived from the animal NOAEL with a safety factor applied (commonly one-tenth of the NOAEL in the most sensitive species, adjusted for allometric scaling). Cohorts of 6-10 subjects receive one dose level; a safety review committee evaluates adverse events before escalating.
• Multiple ascending dose (MAD). Once the SAD data establish a safe single-dose range, subjects receive repeated doses to assess accumulation, trough / peak ratios, and whether tolerability changes with repeated exposure.
• Food effect and special populations. Dedicated sub-studies assess whether a meal changes absorption, and separate cohorts evaluate PK in subjects with renal or hepatic impairment if those populations are likely to use the drug.

Phase 1 trials are small (typically 10-80 subjects) and are not powered to detect efficacy signals. Their output is a safe dose range, a PK profile (Cmax, Tmax, AUC, half-life, clearance, volume of distribution), and an initial adverse-event characterization that guides Phase 2 dose selection.^[8]

Phase 2: dose-finding and proof of concept

Phase 2 trials enroll patients with the target condition, typically a few hundred subjects across multiple dose arms. The questions are: which dose produces the best efficacy-to-tolerability ratio? and does the drug work well enough in patients to justify a large Phase 3 program?

Common elements of a Phase 2 peptide metabolic trial:

• Dose-ranging design. Multiple active doses tested against placebo, sometimes with an active comparator (e.g., an approved GLP-1 agonist at a fixed dose) to calibrate the effect size.
• Biomarker and surrogate endpoints. HbA1c, fasting plasma glucose, body weight, and lipid panels are measured at 12-26 weeks. These are the fastest markers of biological activity and can be collected in a relatively short window.
• Mechanistic sub-studies. Continuous glucose monitoring, meal-tolerance tests, hyperinsulinaemic-euglycaemic clamps, or DEXA body-composition scans may be embedded in Phase 2 to understand the mechanism of weight loss or glycemic benefit.

Phase 2 results guide Phase 3 dose selection. A drug that fails to show a meaningful signal in Phase 2 almost always terminates before Phase 3.

Phase 3: pivotal efficacy

Phase 3 trials are the primary basis for regulatory approval. They are large (typically 1,000-10,000 subjects), randomized, often double-blind or double-dummy, and designed to demonstrate a statistically sound and clinically meaningful benefit against a prespecified primary endpoint. The FDA and EMA generally require at least two adequate and well-controlled Phase 3 trials, though a single large trial with a persuasive secondary endpoint package can sometimes suffice.^[1]

In metabolic peptide development, Phase 3 programs typically consist of multiple trials testing the same drug across populations that differ in disease severity, background therapy, or comorbidity. The SURPASS program for tirzepatide and the STEP program for semaglutide are examples: each comprised five to nine individual Phase 3 trials examining the drug in distinct sub-populations or against distinct comparators.

The regulatory threshold for success depends on the indication. For type 2 diabetes, a reduction in HbA1c that is statistically superior to placebo is the standard primary endpoint. For chronic weight management, the FDA expects both a mean weight reduction of at least 5% beyond placebo and a clinically meaningful proportion of patients achieving at least 5% weight loss.

Phase 4: post-marketing surveillance

Phase 4 studies begin after approval. They address questions that Phase 3 trials cannot: long-term safety signals, effectiveness in populations excluded from pivotal trials (the very elderly, those with severe renal impairment, pregnant women), and whether the drug's benefits translate outside the controlled environment of a randomized trial.

Post-marketing work takes several forms:

• Cardiovascular outcomes trials (CVOTs). Since 2008, the FDA has required manufacturers of new T2DM therapies to demonstrate cardiovascular safety in a dedicated trial. For GLP-1 receptor agonists this requirement produced the LEADER (liraglutide), SUSTAIN-6 (semaglutide), and AWARD-11 / SURPASS-CVOT programs. These trials typically run for 3-5 years and enroll tens of thousands of patients with elevated cardiovascular risk.
• Registry studies and pharmacovigilance. Spontaneous adverse-event reports from prescribers and patients feed regulatory databases (FDA FAERS, EudraVigilance). These are hypothesis-generating, not definitive, but they can trigger label updates or safety communications when signals are strong.
• Label expansion studies. A sponsor wishing to add an indication (for example, adding a cardiovascular risk-reduction claim or a pediatric indication) must conduct additional controlled studies, typically treated as a Phase 3 program for the new indication.
• Real-world evidence. Insurance claims databases, electronic health records, and disease registries can estimate effectiveness and safety in populations and settings far broader than pivotal trial enrollment criteria, but they carry confounding risks that require careful analytical adjustment.

Surrogate endpoints in peptide metabolic trials

A surrogate endpoint is a laboratory measurement or physiological variable used in place of the outcome that actually matters to patients (mortality, heart attack, blindness). Regulators accept surrogate endpoints when:

1. the surrogate is reliably measured and reproducible;
2. evidence shows the surrogate is on the causal pathway to the clinical outcome;
3. intervention-induced changes in the surrogate have been validated as predictive of changes in the clinical outcome across multiple trials.

HbA1c (glycated haemoglobin) is the canonical surrogate endpoint in type 2 diabetes trials. It reflects average plasma glucose over the preceding 8-12 weeks and is strongly associated with risk of microvascular complications (retinopathy, nephropathy, neuropathy). The FDA accepts HbA1c reduction as the primary endpoint for T2DM approval because decades of epidemiological and intervention data connect HbA1c to patient-relevant outcomes.^[4]

Percent body weight loss is the standard surrogate for obesity trials. The FDA thresholds (mean loss greater than 5% beyond placebo, and a responder proportion criterion) were established because weight loss at that magnitude predicts improvements in cardiometabolic risk factors.

The limitation of surrogate endpoints is that changes in the surrogate do not always predict changes in hard outcomes for every drug. A drug might lower HbA1c but increase cardiovascular events (the rosiglitazone experience is the historical example that drove the 2008 FDA CVOT requirement). Reading a peptide trial critically means noting whether the reported primary endpoint is a surrogate or a hard clinical outcome, and whether a CVOT has been completed.^[4]

Regulatory authorisation versus trial registration

A marketing authorisation (or "approval") is the regulatory decision that allows a company to sell a drug for a specific indication in a specific jurisdiction. It is granted by the FDA in the United States and by the EMA for the European Union. Approval means the agency has reviewed the full benefit-risk profile and concluded that the drug's benefits outweigh its risks for the stated indication and population.

Trial registration is a separate, earlier step. In the United States, the FDA Amendments Act of 2007 (FDAAA 801) requires sponsors to register applicable clinical trials on ClinicalTrials.gov before enrolment begins, and to post results within 12 months of primary completion. The purpose is transparency and the prevention of selective publication.^[6]

Key distinctions readers often conflate:

• A drug can have a completed Phase 3 trial (registered on ClinicalTrials.gov, results posted) but not yet be approved by any regulatory agency.
• FDA approval and EMA approval are independent decisions. A drug approved in the US may be under review, not submitted, or rejected in the EU.
• Accelerated Approval (FDA) and Conditional Marketing Authorisation (EMA) allow drugs to reach patients earlier, based on surrogate endpoints, with the requirement that confirmatory trials demonstrating clinical benefit are completed post-approval. A drug with Accelerated Approval is not provisionally approved; it is fully approved for the stated indication, but the approval can be withdrawn if confirmatory trials fail.^[7]