Gastric and Esophageal Cancer Registry Results Posted for Multi-Cohort Trial

원문 링크: WordPress 원문

ClinicalTrials.gov , the official trial registry , posted results on August 3, 2026 for participants with metastatic gastric or esophageal cancer , including a registry-reported 53.8 percent objective-response value in a 65-person esophageal regimen cohort . The record is a multi-cohort registry posting rather than a peer-reviewed comparative report , while it makes the reported cohort-level outcomes and safety record publicly inspectable .

The first public posting of this trial’s results on August 3, 2026 is the new event.

Participants: 219

Reported OR: 53.8% in 65 people

Response 95% CI: 41.03–66.30%

English | 한국어

AI-generated conceptual illustration; not study data.

The 65 people behind the 53.8% result

Design: interventional Phase Ib/II, open-label, randomized umbrella trial. Enrolled population: 219 participants with metastatic gastric or esophageal cancer. Analysis population: efficacy and safety populations were defined separately by cohort and treatment received. Primary endpoint: investigator-assessed objective response under RECIST v1.1.

This is a multi-cohort, open-label Phase Ib/II protocol: one overall study contains several treatment combinations and disease groups. That structure lets investigators examine more than one clinical question, but it also means that a number must stay attached to its own cohort rather than being treated as a result for every participant in the trial.

The 53.8 percent value belongs to the Stage 1 esophageal-cancer group given atezolizumab, cisplatin, and 5-FU. The registry participant-flow table identifies 65 people in this regimen’s efficacy population. The full protocol enrolled 219 people across several groups. The posted 95% confidence interval is 41.03 to 66.30 percent, which shows how much uncertainty surrounds this cohort estimate.

Here, objective response means that scans showed a tumor disappearing or shrinking by the study’s predefined rules, using RECIST v1.1.

Atezolizumab is an immune-checkpoint medicine; cisplatin and 5-FU are chemotherapy drugs. The same cohort’s posted median overall survival was 12.98 months (95% CI 11.17–15.70), but that describes the observed group. The registry lists other cohorts with their own results. The record contains other cohorts, including the esophageal cisplatin-plus-5-FU control group, with their own population sizes, follow-up, and outcomes.

AI-generated conceptual illustration; not study data.

Inside the registry record

Clinical trial registries can make results available before a conventional paper appears, and that timing is the central news event here. The practical value is the chance to examine how outcomes were framed, which populations were counted, and whether the data are divided across cohorts rather than summarized as one uniform experience.

In practical terms, the public record now lets a reader check the treatment, the 65-person analysis group, the endpoint, and the uncertainty interval in the same place. The percentage belongs to one treatment group, one analysis population, and one endpoint.

The result tables identify the outcome measure, cohort labels, and confidence intervals. They also distinguish participants enrolled in the trial from analysis populations such as the safety-evaluable and efficacy groups. For readers following gastric or esophageal cancer research, that separation is essential: an enrollment total tells who entered the protocol, whereas a particular outcome estimate belongs to a named analysis group and intervention context.

Public availability also permits clinicians, researchers, and patients to see why cautious wording is necessary. Because the record contains several cohorts and treatment combinations, the posting should be read as a collection of cohort-specific observations. It reports the cohorts separately. A registry result can document what the sponsor submitted. A journal report and independent review would add comparison with current care.

Limits of a small, terminated program

The official record describes an open-label, multicenter program with multiple cohorts, not one large confirmatory comparison. Several cohorts are small, and the reported estimates have differing uncertainty. Such variation makes cross-cohort comparisons especially fragile when background disease, line of therapy, regimen, and analysis set differ. A response estimate should therefore remain attached to its specific cohort rather than being generalized across gastric and esophageal cancers.

Safety deserves the same discipline. For the same 65-person regimen group, the registry’s group-level adverse-event summary lists serious adverse events in 34 of 65 people and other adverse events in 65 of 65 people. Those summaries record events in people who received study treatment; they do not establish that one drug caused every event, compare risk fairly with a different regimen, or predict an individual person’s risk.

The overall study status is terminated, and the registry says the sponsor closed it after deciding not to continue development of certain treatment combinations. Termination describes the program’s end; it cannot explain every cohort’s result on its own.

A peer-reviewed paper could add more methodological detail. It may later be supplemented, corrected, or interpreted in a publication with fuller methodological context. Until then, the responsible reading is to describe the posted results, their denominators, and their limits without converting them into a recommendation.

AI-generated conceptual illustration; not study data.

Details a full report could add

The next useful evidence would be a full report that specifies the prespecified analyses, missing-data handling, follow-up duration, and reasons participants discontinued within each cohort. Independent peer review can also clarify whether the registry entries align with a complete protocol and statistical analysis plan. Those details are particularly important in an umbrella design, where the same trial title can cover clinically different questions.

Readers can use the official registry page to examine the cohort labels, outcome definitions, and adverse-event fields directly. The linked PubMed record provides additional background, but the numerical statements in this report are locked to the ClinicalTrials.gov machine-readable record. That boundary prevents a registry posting from being presented as a journal conclusion or a regulatory action.

For now, the clearest description is narrow: official results for a terminated multi-cohort cancer trial became public on the registry. They give readers a record to examine while fuller reporting is still needed.

That distinction protects readers from two common errors. A registry entry can be an important public record while still leaving unanswered questions about comparators, data maturity, and clinical applicability. The appropriate next step is careful source review. Readers should also note that clinical decisions require disease details, prior treatment, health status, preferences, and professional assessment that a public registry table cannot supply.

About this report. The figures above come from a ClinicalTrials.gov results posting for one multi-cohort study. They are reported cohort outcomes, so they are best read alongside the named treatment group, analysis population, and endpoint. References: ClinicalTrials.gov official results record; PubMed background record. Disclaimer. This is not medical advice. Discuss personal care decisions with a qualified clinician.