Acid Suppression After Gastric ESD Favors Trends, Not Clear Winners
Gastric endoscopic submucosal dissection leaves behind a controlled but sizable iatrogenic ulcer, and the central question in everyday practice is how best to accelerate mucosal repair without assuming an unproven advantage. In this frequentist network meta-analysis of 31 randomized controlled trials, seven pharmacological regimens were compared for post-ESD ulcer healing and delayed bleeding prevention. The results were more restrained than many clinicians might hope: several approaches ranked numerically well, but no regimen established statistically significant superiority over PPI therapy for healing, and no strategy clearly reduced delayed bleeding or perforation across the treatment course.
A single ulcer created by ESD is not a routine peptic lesion. It is larger, deeper, and exposed to a healing environment shaped by acid load, local ischemia, lesion size, location, and Helicobacter pylori status, which helps explain why the field has kept moving beyond one-dimensional acid suppression toward potassium-competitive acid blockers and mucosal protectants.
Early Healing Signals Favored P-CABs and Combination Therapy
At 4 weeks, P-CAB monotherapy carried the highest numerical ranking and produced a point estimate that leaned toward better healing than PPI, with an odds ratio of 1.61 and a 95% confidence interval from 0.96 to 2.68; the comparison did not cross into statistical significance. The same early window also showed favorable numerical performance for combination regimens, including PPI + MP and P-CAB + MP. By 8 weeks, PPI + MP held the highest numerical ranking, yet no regimen achieved statistically superior healing relative to PPI.
The biologic logic behind those early trends is easy to understand. Stronger, more sustained acid suppression may stabilize the gastric environment around an artificial ulcer, while a mucosal protective agent can support epithelial restitution and barrier repair at the wound surface. In a lesion that is deliberately created during cancer resection, both pharmacologic pressure points matter: acid control reduces ongoing injury, and mucosal protection may improve the quality of re-epithelialization.
The ranking pattern is useful, but it should not be mistaken for a decisive treatment hierarchy. Numerical superiority in a network model can reflect a modest effect size, sparse direct comparisons, or uncertainty carried across the evidence network. Here, the confidence intervals and the overall certainty ratings kept the early signal from becoming a clinical mandate.
Bleeding and Perforation Were Not Meaningfully Shifted by Drug Choice
Across delayed bleeding and perforation outcomes, all regimens performed similarly. H2-receptor antagonist–based strategies showed a non-significant trend toward higher complication risk, but the analysis did not identify a statistically significant safety advantage for any regimen. In practical terms, the main postoperative hazards after gastric ESD remained difficult to move with medication alone.
That finding is clinically important because delayed bleeding after ESD is not purely a question of luminal pH. Vessel caliber, defect size, technical factors, and local tissue vulnerability all contribute to the risk. Even the best acid-suppressive strategy cannot fully offset the mechanics of a large resection bed or the vascular anatomy beneath it, which helps explain why the safety analysis stayed flat despite differences in ranking.
Evidence Certainty Limited How Far the Rankings Could Be Trusted
Most pairwise comparisons were supported by low-to-moderate GRADE evidence. Heterogeneity, inconsistency, sensitivity analyses, and publication bias were assessed, and transitivity was evaluated qualitatively, but the evidence base still did not support a strong comparative claim. This matters because network meta-analysis can extend inference across regimens that have not been directly compared head-to-head in large, definitive trials, yet the strength of the conclusion always depends on the quality and balance of the underlying studies.
The protocol was prospectively registered in PROSPERO, and the review followed PRISMA guidance for network meta-analysis. Adult patients who underwent gastric ESD were eligible, and the interventions included PPIs, P-CABs, H2RAs, mucosal protectants, and combinations of these agents. Outcomes were assessed at 4 and 8 weeks for ulcer healing, with delayed bleeding and perforation tracked over the treatment course. The framework was rigorous; the outcome signal remained modest.
What Gastroenterologists Can Use at the Bedside
For now, PPI therapy remains the reference standard for many post-ESD patients, not because it eclipsed every alternative, but because no comparator clearly displaced it. P-CAB-based regimens and combinations with mucosal protectants looked numerically attractive for early ulcer healing, and those results may justify selective use when clinicians want maximal acid suppression or are managing a patient with a large defect and a concern for delayed healing. Still, the data stop short of proving that one regimen should replace another as routine care.
This is the kind of evidence that changes tone more than it changes protocol. It suggests that the field is not stuck, but it also makes clear that pharmacology alone has not solved the postoperative ulcer problem after gastric ESD.
The trial network leaves several practical questions open: whether certain lesion subtypes benefit disproportionately from combination therapy, whether the apparent early advantage of P-CABs persists in larger and more diverse cohorts, and whether any regimen can reliably lower bleeding risk in a way that changes endoscopic practice. For now, the most defensible reading is cautious. Early healing may be nudged by drug choice, but delayed bleeding and perforation still look governed by factors beyond acid suppression, and the evidence remains too uncertain to declare a new standard.