Introduction
Despite guideline evolution across NCCN, ESMO, and CSCO 131415, gastric and gastroesophageal junction (GEJ) adenocarcinoma management remains burdened by pervasive practice gaps spanning detection, staging, biomarker testing, treatment delivery, and monitoring. These gaps translate directly into measurable harms: advanced-stage presentation, preventable metastatic disease, missed targeted therapy eligibility, and inferior survival. The following review synthesizes current evidence to delineate discrete deficiencies across the care continuum and identifies pragmatic opportunities for improvement.
Diagnostic Delay and Endoscopy Quality
A foundational gap lies at the endoscopy level. In a community-hospital cohort study (2010–2019, n=18,635 EGDs), biopsies were obtained in only 13% of procedures, and premalignant condition detection rates varied nearly threefold across endoscopists (1.8%–5.8%) 6. This variability carried direct clinical consequences: a negative correlation was observed between individual premalignant lesion detection rate and subsequent gastric cancer diagnosis within three years (rs = −0.65 for all premalignant conditions; rs = −0.69 for dysplasia, p=0.037) 6. These findings confirm that endoscopist skill and protocol adherence—not simply imaging technology—determine early detection yield.
Compounding this, H. pylori eradication programs, while efficacious (quadruple therapies achieving >90% eradication rates in high-incidence regions), suffer from inconsistent success verification using noninvasive testing, poorly defined surveillance intervals for patients with atrophy or intestinal metaplasia, and regional antimicrobial resistance monitoring gaps 3. Organized gastric cancer screening is most established at the national level in Japan and South Korea, while China has implemented screening in selected high-risk regions; comparable infrastructure remains limited in most Western settings and many low-middle income countries (LMICs) 9.
An emerging but not yet systematized strategy is personalized risk-based screening that stratifies individuals beyond age alone. Current programs predominantly apply uniform age cutoffs, despite accumulating evidence supporting individualized thresholds—a documented gap between guideline evolution and real-world implementation 2.
Pathology Quality and Risk Stratification
Even when biopsies are obtained, pathological risk stratification suffers from interobserver variability. Among 835 patients, OLGA-based atrophy scoring showed only moderate agreement between expert gastrointestinal pathologists (kappa 0.41–0.57), while general pathologist agreement was poor 5. In contrast, intestinal metaplasia assessment (OLGIM) yielded near-perfect agreement (kappa 0.80–0.82) 5. Critically, OLGA identified 33.4% of patients as stage I–IV, whereas OLGIM captured only 19.9%—meaning sole reliance on OLGIM would miss approximately 14% of high-risk individuals 5. The recommended approach of combining both systems is not yet standardized in most Western centers, creating missed surveillance opportunities and a structural underestimation of at-risk populations.
Staging Accuracy and Laparoscopy Underutilization
Accurate pre-treatment staging remains compromised by the underuse of staging laparoscopy. In an Australian retrospective cohort, staging laparoscopy detected occult metastases in 25.6% of patients considered non-metastatic on imaging, prompting treatment intent changes in 37 of 199 cases 1. Notably, survival for laparoscopy-detected metastatic patients (median OS 8.3 months) was comparable to those with imaging-confirmed M1 disease (6.7 months), confirming equivalent prognostic weight 1. Data from Pakistan demonstrate even higher detection yields, with staging laparoscopy identifying metastases in 40% of apparently non-metastatic patients—48% in gastric versus 28% in GEJ tumors (p=0.01) 7.
A Scandinavian multidisciplinary team conference (MDTC) study of 222 patients found that peritoneal carcinomatosis was missed in 16.7% of gastric cancer cases deemed resectable by MDTC—confirming that multidisciplinary review alone is insufficient without laparoscopic confirmation 12. Yet staging laparoscopy is not universally integrated into preoperative protocols across the US, EU, or LMICs. This gap has direct downstream consequences: patients who undergo resection with undetected peritoneal disease experience early relapse and unnecessary surgical morbidity.
Biomarker Testing: Uptake, Quality, and Tissue Challenges
Comprehensive biomarker profiling—HER2, PD-L1 CPS, MSI-H/dMMR, EBV, CLDN18.2, and FGFR2b—is essential for contemporary treatment selection, yet testing remains inconsistently implemented.
| Biomarker | Prevalence | Key Testing Challenge |
|---|---|---|
| CLDN18.2 (≥75% cells, 2+/3+) | 38.4% globally 19 | 61.1% concordance between archival and fresh tissue 19 |
| FGFR2b (any 2+/3+) | 15.4% overall; 29.8% in recent samples 18 | Specimen age critically affects detection; older samples show 10.4% prevalence 18 |
| PD-L1 (CPS ≥1) | 62.0% in Japanese advanced GC 22 | Assay concordance (28-8 vs. 22C3): Kendall's tau 0.94; test failure rate 3–6% 20 |
| MSI-H | 6.3–6.9% 2122 | Higher PD-L1 CPS co-expression in MSI-H tumors (CPS ≥10: 66.7% vs. 24.8% in MSS, p<0.0001) 22 |
| HER2-positive | 17.7% in surgical cohort 23 | — |
A critical finding is that standard PD-L1 testing alone misses 5.2% of patients with PD-L1-negative tumors who harbor high tumor mutational load (TML) or MSI-H and may still benefit from immunotherapy 21. Furthermore, approximately 53.9% of FGFR2b-overexpressing tumors are negative for all other actionable biomarkers (HER2, dMMR, PD-L1 CPS ≥5, CLDN18.2), identifying a distinct population with no current standard targeted option 18. Tissue specimen aging disproportionately impairs FGFR2b detection, underscoring the need for fresh biopsy protocols and reflexive retesting at progression.
The October 2024 FDA approval of zolbetuximab-clzb for CLDN18.2-positive HER2-negative advanced gastric/GEJ cancer 31—demonstrating median OS of 18.2 versus 15.5 months (HR 0.750, p=0.0053) in SPOTLIGHT—makes CLDN18.2 testing a clinical imperative. However, global testing infrastructure for this biomarker is nascent.
Perioperative Therapy and Surgical Quality
Real-world perioperative FLOT delivery demonstrates significant attrition from trial to practice. In the Italian RealFLOT study (n=206), only 60.2% of patients received full preoperative cycles, and merely 13.6% completed four full-dose postoperative FLOT cycles—versus approximately 81% completion in the FLOT4-AIO trial 17. The real-world pCR rate of 7.3% was markedly lower than trial-reported rates of 16–20% 17, attributable to more advanced baseline disease (87% cT3–4, 85% cN+) and dose attrition. D2 lymphadenectomy was achieved in 82.1% and R0 resection in 92.1% of operated patients 17, indicating acceptable surgical quality in tertiary centers.
In a randomized Chinese trial (n=312), adjuvant chemoradiation provided a statistically significant 3-year DFS benefit over chemotherapy alone specifically in high-risk patients (pN ≥N2: 71.0% vs. 53.0%; p<0.05), with acceptable grade 3–4 toxicity rates (18.5% CRT vs. 26.1% chemotherapy) 24. This confirms that risk-stratified adjuvant strategies are feasible but underutilized outside Asia.
Metastatic Disease: Treatment Uptake and Toxicity
In Japan, a temporal analysis of 949 HER2-negative advanced gastric cancer patients (2011–2023) showed only modest population-level OS improvement across immunotherapy eras (median OS: 16.2, 15.2, 21.3 months across periods; log-rank p=0.50), with any-line immunotherapy use significantly associated with improved prognosis (HR 0.54, p<0.0001) 29. Critically, elderly patients (≥75 years) received significantly lower rates of second- and third-line therapy, highlighting age-related undertreatment.
In Korea, third-line ICI versus irinotecan demonstrated comparable median OS (5.5 vs. 6.0 months, p=0.786) overall, but ICI showed superior OS in patients without peritoneal metastasis (HR 0.54, p=0.047) and dramatically better PFS in MSI-H/EBV-positive tumors (12.7 vs. 2.8 months, HR 0.27, p=0.012) 30. Development of immune-related adverse events with nivolumab correlates with superior OS (HR 0.4, p<0.05) and doubled objective response rates (24.7% vs. 6.4%) 26, reinforcing the value of proactive irAE monitoring as a surrogate efficacy signal.
A meta-analysis of six pembrolizumab RCTs confirmed that combination therapy reduces mortality risk (OS HR 0.72, p<0.01), with the greatest benefit observed at PD-L1 CPS ≥10 33. Yet real-world first-line immunotherapy adoption remains geographically uneven, particularly in LMICs where payer coverage for novel agents (zolbetuximab, trastuzumab deruxtecan, nivolumab) is absent or limited.
Surveillance, Monitoring, and Health System Coordination
Post-resection surveillance protocols lack standardization globally. No high-quality evidence from retrieved materials supports intensified imaging-based surveillance improving OS in resected gastric cancer patients. The role of ctDNA-based minimal residual disease (MRD) detection and early relapse identification remains investigational and is not yet incorporated into any major guideline 1314. H. pylori infection status as a predictor of immunotherapy outcomes represents an emerging unmet need bridging prevention and precision oncology that is not yet standardized in clinical practice 16.
The NCCN (Version 2.2026) multilevel guideline framework—acknowledging variable implementation capacity—implicitly recognizes that guideline adherence and infrastructure standardization remain unresolved 13.
Conclusion
Across the care continuum, measurable gaps persist in endoscopy biopsy rates, pathological risk stratification (OLGA/OLGIM), staging laparoscopy utilization, comprehensive biomarker testing, perioperative therapy dose intensity, and equitable access to approved targeted agents. Prioritized actions include: (1) standardizing biopsy protocols and OLGA/OLGIM combined staging; (2) mandating staging laparoscopy for all non-metastatic T3/T4 or N+ disease; (3) implementing reflex multiplex biomarker panels (HER2, PD-L1, MSI-H, CLDN18.2, FGFR2b) at diagnosis and progression; (4) adopting risk-stratified perioperative regimens with structured dose-intensity monitoring; and (5) establishing equitable access frameworks for immunotherapy and targeted agents across all practice settings. Closing these gaps requires coordinated investment in endoscopist training, centralization of complex surgery, pathology standardization, and health system policy engagement across all geographies.