Gastric Cancer

Gastric cancer, predominantly gastric adenocarcinoma , is a major global malignancy. According to 2008 data, it was the fourth most common cancer worldwide and the second leading cause of cancer death ^[1]A1c. By 2022, it had become the fifth leading cause of cancer mortality, reflecting declining incidence in some regions ^[15]A1b. The disease carries a high mortality-to-incidence ratio, with most patients presenting at an incurable stage and a median survival of less than 1 year ^[1]A1c[13]D5.

Geographic and Temporal Patterns

Approximately three-quarters of new cases arise in Asian countries, with East Asia bearing the highest burden ^[1]A1c. In Western populations, the overall incidence has declined, but a shift toward tumors involving the proximal stomach and gastroesophageal junction has been observed ^[14]D5.

The table below summarizes key epidemiological parameters.

The marked geographic variation in incidence points to environmental and dietary risk factors, which are discussed in the next section.

Pearl: Despite declining incidence in many regions, gastric cancer's high mortality-to-incidence ratio and late presentation make it a persistent global health challenge; clinicians should maintain a high index of suspicion in patients from high-incidence areas.

After understanding the global burden of gastric cancer, the clinician must identify modifiable and non-modifiable risk factors that drive incidence and guide prevention strategies. The strongest and most tractable risk factor is Helicobacter pylori infection. Using immunoblot, which detects antibodies that persist even after the infection clears, nearly all noncardia gastric cancer cases are classified as H. pylori positive, with a multivariable odds ratio of 21.4 (95% CI 7.1-64.4) compared with ELISA-based detection (OR 6.8, 95% CI 3.0-15.1) ^[24]C4. This suggests that H. pylori infection is a necessary condition for noncardia gastric cancer. Diet and lifestyle also matter: a Western/unhealthy dietary pattern (rich in starchy foods, meat, and fats) carries an OR of 1.51 (95% CI 1.21-1.89) for the highest versus lowest consumption, while a Prudent/healthy pattern (rich in fruits and vegetables) is protective (OR 0.75, 95% CI 0.63-0.90) ^[2]B2a. High vegetable intake specifically reduces distal gastric cancer risk among men (HR 0.78, 95% CI 0.63-0.97 for highest versus lowest quintile) ^[12]B2b. Heavy alcohol consumption (≥4 drinks/day) increases risk by 20% (RR 1.20, 95% CI 1.01-1.44), but moderate drinking shows no association ^[23]B2a. Regular NSAID use is independently protective: each incremental year of use reduces risk by 21% (HR 0.79, P < 0.001), and among H. pylori-infected patients the reduction is 48% per year, with a number needed to treat of 50 ^[34]B3b.

Non-modifiable risk factors

Genetic predisposition accounts for a substantial fraction of cases. Carriers of a mismatch repair (MMR) gene mutation have a standardized incidence ratio of 9.78 (95% CI 1.18-35.3) for gastric cancer ^[25]C4. Germline CDH1 loss-of-function variants cause hereditary diffuse gastric cancer; endoscopic surveillance in these individuals detects invasive signet ring cell carcinoma in 63% of those undergoing two or more surveillance endoscopies, and prophylactic total gastrectomy reveals multifocal stage IA carcinoma in 93% of patients with negative biopsies ^[31]C4. A polygenic risk score derived from 112 single-nucleotide polymorphisms identifies Chinese individuals with high genetic risk (HR 2.08, 95% CI 1.61-2.69) compared with low genetic risk ^[30]B2a. First-degree family history of gastric cancer is also a risk factor for incidence, but paradoxically, patients with stage III/IV disease and a first-degree family history have better survival after curative-intent surgery (HR 0.49 for disease-free survival, 95% CI 0.29-0.84) ^[33]B3b.

Prevention and screening

Modifiable risk factors are actionable. A healthy lifestyle, defined as not smoking, never consuming alcohol, low intake of preserved foods, and frequent consumption of fresh fruits and vegetables, reduces gastric cancer risk by 47% (HR 0.53, 95% CI 0.29-0.99) even among individuals with high genetic risk, with an absolute risk reduction of 1.12% ^[30]B2a. H. pylori screening and eradication is recommended in high-incidence regions. In Asia, resource-stratified guidelines recommend endoscopic screening in populations with high incidence, such as those in East Asia ^{[1]A1c[22]A1c}. For individuals with hereditary syndromes, genetic testing and surveillance are indicated: in CDH1 carriers, endoscopic surveillance is an acceptable alternative to prophylactic gastrectomy for those who decline surgery, with a low rate of incident tumours beyond T1a ^[31]C4.

Pearl: H. pylori infection is a necessary condition for noncardia gastric cancer, nearly all cases are immunoblot-positive, so eradication in high-risk populations is the single most impactful preventive intervention, and regular NSAID use may further reduce risk in H. pylori-infected patients (NNT = 50).

Beyond modifiable risk factors, a subset of gastric cancers arises from inherited germline mutations that confer high lifetime risk and mandate distinct management strategies. The most well-characterized hereditary syndrome is hereditary diffuse gastric cancer (HDGC), but gastric cancer also occurs in Lynch syndrome , hereditary breast-ovarian cancer, Li-Fraumeni syndrome, and other polyposis syndromes ^[45]D5.

Hereditary Diffuse Gastric Cancer (HDGC)

HDGC is an autosomal dominant syndrome caused predominantly by inactivating germline mutations in CDH1, the gene encoding E-cadherin, and less frequently by pathogenic variants in CTNNA1 ^[36]A1c. Penetrance estimates vary by ascertainment: using strict HDGC clinical criteria, the cumulative risk of gastric cancer by age 80 years is 70% for males and 56% for females ^[36]A1c. However, in families ascertained through multigene panel testing without strict HDGC criteria, penetrance is lower, 42% for males and 33% for females, suggesting that historical estimates may overestimate risk for some carriers ^[49]B3b. Female breast cancer risk (lobular type) is more consistent, ranging from 39% to 55% ^{[36]A1c[49]B3b}.

Genetic testing criteria have been relaxed in the 2020 International Gastric Cancer Linkage Consortium (IGCLC) guidelines: testing is now recommended for isolated diffuse gastric cancer (DGC) at age <50 years (previously <40), bilateral lobular breast cancer (LBC) at age <70 years, and all Māori individuals with confirmed DGC ^[36]A1c. Testing should include sequencing and copy number analysis of CDH1; if negative, CTNNA1 analysis is considered ^[36]A1c.

Management centers on prophylactic total gastrectomy (PTG), recommended in early adulthood (generally 20-30 years) for CDH1 pathogenic variant carriers from HDGC families, regardless of endoscopic findings ^[36]A1c. For those declining or postponing surgery, annual endoscopic surveillance at expert centers with systematic random biopsies is an acceptable alternative. In a prospective cohort of 270 CDH1 carriers, surveillance detected signet ring cell carcinoma in 63% of patients over a median 31 months, and no participant developed advanced (stage III/IV) gastric cancer during follow-up ^[31]C4. Random biopsies are essential: omitting them would lead to under-diagnosis in 42% of cases ^[42]B2b.

Other Hereditary Syndromes

Gastric cancer risk is elevated in several other hereditary cancer syndromes (Table 1).

Common Genetic Variants and Polygenic Risk

Beyond monogenic syndromes, common genetic variants contribute to gastric cancer susceptibility. A polygenic risk score derived from 112 single-nucleotide polymorphisms was significantly associated with gastric cancer risk in a Chinese cohort (HR 2.08 for high vs low genetic risk) ^[30]B2a. Importantly, a healthy lifestyle (no smoking, no alcohol, low preserved foods, high fresh fruit/vegetable intake) reduced risk by 47% among those with high genetic risk (absolute risk reduction 1.12%) ^[30]B2a. Host genetic variants in immune-related genes (e.g., IL1RN, IL8) also modify risk, particularly in the context of Helicobacter pylori infection ^{[38]B2b[48]C4}.

Pearl: For any patient with diffuse-type gastric cancer diagnosed before age 50, bilateral lobular breast cancer before age 70, or a family history of both, refer for genetic counseling and CDH1 testing, identifying a pathogenic variant changes management from surveillance to consideration of prophylactic total gastrectomy .

Beyond germline predisposition, the histologic and molecular heterogeneity of gastric cancer defines distinct prognostic groups and guides therapeutic selection. Two complementary frameworks, histologic classification and molecular subtyping, are now essential for risk stratification and treatment planning.

Histologic Classification

The Lauren classification divides gastric adenocarcinoma into intestinal, diffuse, and mixed types. Intestinal-type tumors form glandular structures and are more common in high-incidence regions, whereas diffuse-type tumors are poorly cohesive, infiltrate as single cells or small clusters, and often lack gland formation. Diffuse histology carries independent prognostic significance: in a real-world cohort of 608 patients, diffuse-type histology was independently associated with worse overall survival (OS) ^[64]B3b. The prognostic impact extends to treatment response. In a meta-analysis of six phase III trials (n=5,410), patients with intestinal-type tumors derived greater OS benefit from first-line chemoimmunotherapy than those with diffuse-type tumors (HR 0.78 vs 0.87) ^[66]A1a. Similarly, in the INT-0116 trial of adjuvant chemoradiation, patients with diffuse histology showed minimal nonsignificant treatment effect, whereas intestinal-type patients benefited substantially ^[55]A1b. Anatomic location also correlates with histology: esophagogastric junction (EGJ) tumors have a lower incidence of diffuse-type histology than gastric tumors, but multivariate analysis confirms that histologic subtype, not location, is the dominant prognostic factor ^[65]B3b.

The World Health Organization (WHO) classification further subdivides gastric cancer into tubular, papillary, mucinous, and poorly cohesive (including signet-ring cell) carcinomas. Poorly cohesive/diffuse-type histology is a component of the modified JCOG prognostic index, which stratifies patients into good (median OS 20.5 months), moderate (13.5 months), and poor (10.2 months) risk groups ^[64]B3b.

Table 1. Lauren Classification of Gastric Cancer

Molecular Classification

The Cancer Genome Atlas (TCGA) proposed four molecular subtypes based on comprehensive genomic profiling: Epstein-Barr virus (EBV)-positive (~9%), microsatellite instability (MSI) (~22%), genomically stable (GS) (~20%), and chromosomal instability (CIN) (~50%) ^[60]D5. Each subtype has distinct molecular features and therapeutic implications. EBV-positive tumors show recurrent PIK3CA mutations, extreme DNA hypermethylation, and immune checkpoint activation, making them candidates for immunotherapy. MSI tumors have high mutation burden and respond to PD-1 blockade. GS tumors are enriched for diffuse histology and CDH1/RHOA mutations, and currently lack validated targeted therapies. CIN tumors show marked aneuploidy and amplification of receptor tyrosine kinases (e.g., EGFR, HER2, MET) ^[60]D5.

An alternative classification from the Asian Cancer Research Group (ACRG) identifies four subtypes: MSI, microsatellite stable with epithelial-to-mesenchymal transition (MSS/EMT), MSS/TP53+, and MSS/TP53- ^[67]D5. The MSS/EMT subtype overlaps with TCGA GS and is enriched for diffuse histology and poor prognosis.

Table 2. TCGA Molecular Subtypes of Gastric Cancer

Key Molecular Biomarkers

HER2 is the only validated predictive biomarker for targeted therapy in gastric cancer. Amplification by FISH occurs in 10.9% of cases, with 90% concordance between FISH and immunohistochemistry (IHC) ^[59]B2b. In the INT-0116 trial, patients with HER2-non-amplified cancers derived significant survival benefit from adjuvant chemoradiation (median OS 44 vs 24 months; P=0.003), whereas those with HER2 amplification did not ^[59]B2b. Trastuzumab is approved for HER2-positive advanced disease.

EGFR is overexpressed in 15-45% of gastric cancers, but lapatinib, a dual EGFR/HER2 inhibitor, showed only modest single-agent activity (response rate 9%, median OS 4.8 months) in a phase II trial ^[58]B2b.

ERCC1 mRNA expression is an independent prognostic factor: high expression was associated with worse OS (HR 1.37; 95% CI 1.08-1.75) in the JCOG9912 trial, but it did not predict chemotherapy benefit ^[54]A1b.

Claudin 18.2 (CLDN18.2) is an emerging target for zolbetuximab. An artificial intelligence model using H&E-stained slides predicted CLDN18.2 expression with an AUROC of 0.752-0.856 and identified subgroups with differential benefit from nivolumab plus chemotherapy ^[63]B3b.

PD-L1, MSI, and EBV status guide immunotherapy selection. The tumor microenvironment also plays a role: M2 tumor-associated macrophages (TAMs) promote immunosuppression and angiogenesis, particularly in diffuse-type tumors, and represent a potential therapeutic target ^[68]D5.

Gene Expression Signatures for Risk Stratification

A seven-gene signature (GC-RiskAssigner: CDH1, ELOVL5, EGFR, PIP5K1B, FGF1, CD44v8.10, TBCEL) derived from the MAGIC trial stratified perioperative chemotherapy-treated patients into high-risk (median OS 10.2 months) and low-risk (80.9 months) groups (HR 5.1; P<0.0001), independent of nodal status ^[52]A1b. A four-gene classifier (GZMB, WARS, SFRP4, CDX1) validated in the CLASSIC trial predicted adjuvant chemotherapy benefit: in the predicted-benefit group, 5-year OS improved from 64.5% to 80% with chemotherapy (HR 0.47; P=0.0015), whereas no benefit was seen in the no-benefit group ^[32]B2b.

Pearl: Diffuse-type histology and the GS/EMT molecular subtype identify patients with poor prognosis and limited benefit from current chemoimmunotherapy and adjuvant chemoradiation, these patients urgently need novel therapeutic strategies targeting the immunosuppressive tumor microenvironment.

The histologic subtypes and molecular classifications described above shape the clinical behavior of gastric cancer, but the presentation ultimately depends on stage at diagnosis. Early gastric cancer is frequently asymptomatic; when symptoms do occur, they are non-specific and often attributed to benign conditions. By the time patients seek medical attention, many already have locally advanced or metastatic disease.

Symptoms

Weight loss is the most common symptom and a marker of poor prognosis. In a multicenter cohort of 406 patients with esophagogastric cancer, 48% had pretreatment cachexia defined as >5% half-year weight loss or >2% weight loss in patients with BMI <20 kg/m². Cachexia was independently associated with decreased overall survival (HR 1.52; 95% CI 1.11-2.09) ^[5]B2b. Despite its prevalence, more than one-third of patients with cachexia were not referred for dietetic consultation at baseline ^[5]B2b.

Other symptoms include epigastric pain, early satiety, dysphagia, nausea, vomiting, and hematemesis or melena from tumor bleeding. Obstructive symptoms (dysphagia for proximal tumors, gastric outlet obstruction for distal tumors) suggest advanced local disease. The evidence does not report exact prevalence rates for these individual symptoms.

Signs

Physical examination may reveal pallor from chronic blood loss, a palpable epigastric mass, or left supraclavicular (Virchow) lymphadenopathy. Ascites , hepatomegaly, or peritoneal nodules indicate peritoneal carcinomatosis. Laboratory abnormalities include anemia (iron deficiency from chronic bleeding) and elevated inflammatory markers. The modified Glasgow prognostic score (mGPS), derived from C-reactive protein and albumin, is strongly prognostic for overall survival in advanced gastric cancer and correlates with sarcopenia ^[70]A1b.

Onset and Progression

Gastric cancer typically progresses insidiously over months. Alarm features that demand urgent upper endoscopy include dysphagia, persistent vomiting, unexplained weight loss, GI bleeding, and a palpable abdominal mass.

Atypical Presentations

Autoimmune gastritis is an emerging cause of gastric cancer, particularly in younger women. In a nested case-control study within the Finnish Maternity Cohort, antiparietal cell antibody (APCA) seropositivity was associated with a 5.5-fold increased risk of gastric cancer among H. pylori-seronegative women (OR 5.52; 95% CI 3.16-9.64), and the association was strongest for tumors in the fundus and corpus (OR 24.84; 95% CI 8.49-72.72) ^[83]B3b. This pattern aligns with rising incidence trends in young females and supports a model where autoimmune gastritis replaces H. pylori as the driving factor.

In hereditary diffuse gastric cancer (CDH1 mutation carriers), endoscopic surveillance with systematic random biopsies detects intramucosal signet ring cell carcinoma in 40% of individuals over a median follow-up of 51 months, often before symptoms develop ^[42]B2b.

Pearl: Cachexia is present in nearly half of patients at diagnosis and independently predicts worse survival; early dietetic referral can mitigate weight loss ^[5]B2b. In young women without H. pylori infection, consider autoimmune gastritis as an underlying etiology, especially for corpus-predominant tumors ^[83]B3b.

The clinical presentation of gastric cancer, whether dyspepsia, weight loss, or bleeding, mandates tissue acquisition for histologic confirmation and molecular profiling. Endoscopic biopsy remains the cornerstone, but the technique and tissue handling directly determine diagnostic accuracy and the feasibility of downstream biomarker testing.

Tissue Acquisition Techniques

Endoscopic biopsy is the first-line approach. Multiple biopsies (≥6-8) from the tumor edge and center increase yield, especially for diffuse-type cancers where deeper invasion may require repeated sampling. Endoscopic ultrasound (EUS)-guided fine-needle aspiration adds value for subepithelial lesions or when nodal metastases are suspected. Sentinel node mapping, using a standardized dual-tracer endoscopic injection technique, has been prospectively validated in a multicenter phase II trial: among 397 eligible patients with cT1-T2 tumors <4 cm, the sentinel node detection rate was 97.5% (387 of 397), and the accuracy for nodal metastasis was 99% (383 of 387), with a sensitivity of 93% (53 of 57) ^[94]B2b. The procedure is safe and effective for superficial, small tumors, but false-negative results increase in pT2 or larger tumors (>4 cm) ^[94]B2b.

Histologic Assessment and Biomarker Testing

All gastric adenocarcinomas should undergo systematic biomarker testing to guide therapy. The NCCN guidelines recommend testing for HER2, PD-L1, microsatellite instability (MSI), and - increasingly - FGFR2b overexpression ^[84]A1c.

HER2 testing follows a sequential algorithm: immunohistochemistry (IHC) first, then in situ hybridization (ISH) for equivocal cases (IHC 2+). In the INT-0116/SWOG9008 trial, the HER2 amplification rate by FISH was 10.9% and the IHC overexpression rate was 12.2%, with 90% agreement between the two methods ^[59]B2b. The MAGIC trial reported a concordance of 96% between IHC3+ and brightfield dual ISH ^[86]A1b. However, the VARIANZ study revealed a 22.7% deviation rate between central and local HER2 testing, and patients with centrally confirmed HER2+ disease had significantly longer survival on trastuzumab than those with discordant results (median OS 20.5 vs. 10.9 months; HR 0.42) ^[96]C4. These findings underscore the need for rigorous, centralized testing with quantified thresholds. Optimized thresholds for trastuzumab benefit were a minimum of 40% HER2+ tumor cells and a HER2 amplification ratio of ≥3.0 ^[96]C4.

FGFR2b overexpression is an emerging biomarker. In a pooled analysis of two phase III trials, the prevalence of FGFR2b any 2+/3+ was 36.5% (95% CI 35.4-37.6) and of FGFR2b ≥10% 2+/3+ was 16.6% (95% CI 15.7-17.4) using a validated IHC assay ^[95]B2b. Testing is performed on archival or fresh tumor tissue, with membrane staining scored by a pathologist. The assay is not yet approved by health authorities but is increasingly used to select patients for bemarituzumab trials ^[95]B2b.

PD-L1 and MSI testing are also standard. In the KEYNOTE-059 trial, which used pembrolizumab , PD-L1 positivity was defined as expression in ≥1% of tumor cells or stroma by IHC ^[98]B2b; the umbrella K-Umbrella trial used PD-L1, MSI-high, and Epstein-Barr virus status to allocate patients to nivolumab ^[27]A1b.

Diagnostic Test Performance

The table below summarizes key performance metrics for the diagnostic tests described in the evidence.

Stepwise Diagnostic Algorithm

1. Endoscopic biopsy - obtain ≥6 biopsies from tumor edge and center; fix in formalin for histology and molecular testing.
2. Histologic confirmation - hematoxylin and eosin staining, classify as intestinal or diffuse type per Lauren classification.
3. HER2 testing - IHC first; if 2+, order ISH. Report quantification of percentage of positive cells and amplification ratio.
4. PD-L1 testing - IHC (e.g., 22C3 or SP142 antibody) with ≥1% threshold.
5. MSI testing - PCR or IHC for mismatch repair proteins.
6. FGFR2b testing - IHC for membrane 2+/3+ staining if clinical trial enrollment is considered.
7. Sentinel node biopsy - consider for cT1-T2 tumors <4 cm at experienced centers.

Tissue handling must preserve DNA/RNA integrity. Fresh biopsies for research protocols should be snap-frozen; archival paraffin-embedded tissue is acceptable for standard clinical assays. The pathologist’s report should include the percentage of tumor cells and the intensity of staining for each biomarker to guide therapeutic decisions.

Pearl: When a biopsy specimen shows only IHC 2+ for HER2, reflex ISH is mandatory, but even with a positive result, the survival benefit of trastuzumab is significantly attenuated if <40% of tumor cells are HER2-positive or if the amplification ratio is <3.0 ^[96]C4. Centralized testing with quantified thresholds reduces the 22.7% discordance rate seen between local and central laboratories.

Once histologic diagnosis is confirmed, imaging defines the anatomic extent of disease and guides treatment decisions. Staging determines whether curative-intent resection is feasible and which neoadjuvant or perioperative strategy applies. The NCCN guidelines mandate a staging CT of the chest and abdomen for all patients ^{[100]A1c[105]B2a}. Modality choice and sequence follow the clinical scenario: local (T) staging, nodal (N) assessment, detection of distant metastases (M), and, increasingly, functional characterization for treatment response.

Computed Tomography

Contrast-enhanced multidetector CT (MDCT) of the chest, abdomen, and pelvis is the first-line imaging study for initial staging. It evaluates the primary tumor, regional lymph nodes, and distant metastases (liver, peritoneum, lung, distant nodes). Diagnostic accuracy for overall T staging ranges from 77.1% to 88.9% ^[102]B2a. Sensitivity for serosal involvement is 82.8% to 100%, specificity 80% to 96.8% ^[102]B2a. CT is less reliable for early T1 tumors and for detecting peritoneal carcinomatosis, where sensitivity drops below 50% ^[105]B2a. For nodal staging, clinical assessment by CT is inaccurate compared with final pathology ^[111]D5. Despite these limitations, CT remains the essential initial staging tool because it is widely available, reproducible, and provides a comprehensive survey.

Endoscopic Ultrasound

Endoscopic ultrasound (EUS) is the preferred modality for locoregional (T and N) staging. A systematic review reported diagnostic accuracy for overall T staging of 65% to 92.1% ^[102]B2a. Sensitivity for serosal involvement is 77.8% to 100%, specificity 67.9% to 100% ^[102]B2a. EUS is superior to CT for distinguishing T1 from T2 tumors and for assessing depth of invasion, which is critical for deciding between endoscopic resection and surgery. EUS also guides fine-needle aspiration of suspicious lymph nodes. The NCCN guidelines recommend EUS as part of the initial staging workup when available ^[100]A1c.

Positron Emission Tomography

FDG-PET/CT is not routinely recommended for initial staging because of limited sensitivity for diffuse-type, mucinous, and signet-ring cell adenocarcinomas, which often show low FDG avidity. However, PET has a role in response assessment. In a prospective study of 64 patients with advanced gastric cancer, an early FDG uptake reduction of 40% (δ-SUV) at day 14 predicted clinical response with 70% sensitivity and 83% specificity ^[113]B2b. A reduction of 30% predicted disease control with 58% sensitivity and 100% specificity ^[113]B2b. FDG uptake reduction in liver metastases was also prognostic for overall survival ^[113]B2b. PET is therefore used selectively, primarily to evaluate response to neoadjuvant or palliative chemotherapy and to detect occult metastases when CT findings are equivocal.

Diagnostic Laparoscopy

Diagnostic laparoscopy is recommended for patients at risk for stage IV disease, particularly those with T3/T4 tumors, bulky nodal disease, or suspected peritoneal involvement ^[105]B2a. Laparoscopy detects occult peritoneal carcinomatosis and positive peritoneal cytology that CT misses, thereby avoiding unnecessary laparotomy. The NCCN guidelines advise staging laparoscopy before curative-intent resection in all patients with locally advanced disease ^[100]A1c. Peritoneal lavage for cytology is performed concurrently; positive cytology (even without visible implants) is considered M1 disease.

Emerging Imaging Biomarkers

A CT-based scoring model for tertiary lymphoid structures (ctTLS) has been developed and validated in a multicenter study of 3,155 patients. The ctTLS model predicted TLS status with AUCs of 0.727-0.809 in internal validation and 0.704-0.807 in external validation ^[112]B2b. Higher ctTLS classes were associated with improved disease-free and overall survival (HR range 0.206-0.634) and with benefit from immunotherapy: objective response rates were 16.7% for ctTLS-0, 35.5% for ctTLS-1, 45.8% for ctTLS-2, and 53.8% for ctTLS-3 ^[112]B2b. This noninvasive approach may refine patient selection for immunotherapy in the future.

Surveillance Imaging

After curative-intent treatment, surveillance imaging is performed to detect recurrence. The NCCN guidelines recommend a CT scan of the chest and abdomen every 6 to 12 months for the first 2 years, then annually ^[100]A1c. PET/CT is reserved for suspected recurrence when CT findings are equivocal or when metabolic confirmation is needed. No randomized data support a survival benefit from intensive surveillance, but early detection of oligometastatic recurrence may allow salvage therapy.

Pearl: EUS remains the first choice for T staging, but diagnostic laparoscopy is essential to rule out occult peritoneal disease before proceeding with curative-intent surgery, CT alone misses up to 30% of peritoneal metastases.

Imaging defines the anatomic extent of disease, but molecular profiling of the tumor specimen determines the therapeutic strategy. The NCCN Guidelines recommend biomarker testing for all patients with newly diagnosed advanced gastric cancer, including HER2, PD-L1, microsatellite instability (MSI)/mismatch repair (MMR) status, and, in selected cases, tumor mutational burden (TMB) and NTRK fusions ^[114]A1c. These biomarkers stratify patients for targeted therapy and immunotherapy, directly influencing first- and later-line treatment decisions.

HER2 Testing and Targeted Therapy

HER2 overexpression or amplification is the first actionable biomarker in gastric cancer. Testing follows a validated algorithm: immunohistochemistry (IHC) as the initial screen, with reflex fluorescence in situ hybridization (FISH) for IHC 2+ cases ^[84]A1c. IHC 3+ or IHC 2+/FISH+ defines HER2 positivity. Trastuzumab plus chemotherapy is the standard first-line therapy for HER2-positive advanced disease ^[84]A1c. The KEYNOTE-811 trial demonstrated that adding pembrolizumab to trastuzumab and chemotherapy further improves progression-free survival and overall survival without compromising health-related quality of life ^[116]A1b.

In the second-line setting, trastuzumab deruxtecan (T-DXd) 6.4 mg/kg every 3 weeks has emerged as the preferred agent. In the phase 3 DESTINY-Gastric04 trial, T-DXd significantly improved overall survival compared with ramucirumab plus paclitaxel (median 14.7 vs. 11.4 months; HR 0.70; 95% CI 0.55-0.90; P=0.004) ^[93]B2b. The objective response rate was 44.3% with T-DXd versus 29.1% with ramucirumab-paclitaxel. Adjudicated drug-related interstitial lung disease or pneumonitis occurred in 13.9% of T-DXd recipients, mostly low-grade ^[93]B2b. These results confirm earlier findings from DESTINY-Gastric01, where T-DXd produced an objective response in 51% of patients versus 14% with chemotherapy (P<0.001) and improved overall survival (median 12.5 vs. 8.4 months; HR 0.59; 95% CI 0.39-0.88) ^[121]B2b.

For patients with HER2-positive disease that has progressed after trastuzumab, anbenitamab (a bispecific antibody) plus chemotherapy showed clinically meaningful benefit in the phase 3 KC-WISE trial: median progression-free survival 7.1 versus 2.7 months (HR 0.25; 95% CI 0.17-0.39; P<0.0001) and median overall survival 19.6 versus 11.5 months (HR 0.29; 95% CI 0.17-0.50; P<0.0001) ^[85]A1b. Grade 3 or higher treatment-related adverse events occurred in 60% of the anbenitamab group versus 45% of controls ^[85]A1b.

Other HER2-directed agents have not succeeded in gastric cancer. Lapatinib plus paclitaxel did not significantly improve overall survival in the TyTAN trial (11.0. 8.9 months), although the response rate was higher (odds ratio 3.85; P<0.001) ^[87]A1b. Trastuzumab emtansine (T-DM1) was not superior to taxane in the GATSBY study (median overall survival 7.9. 8.6 months; HR 1.15; 95% CI 0.87-1.51) ^[120]B2b.

Microsatellite Instability and Immunotherapy

MSI-high (MSI-H) status is present in approximately 7.8% of resectable gastric cancers ^[118]B2a. In a meta-analysis of individual patient data from the MAGIC, CLASSIC, ARTIST, and ITACA-S trials, MSI-H was a strong favorable prognostic marker: 5-year overall survival was 77.5% (95% CI 70.0%-85.8%) for MSI-H versus 59.3% (95% CI 56.6%-62.1%) for microsatellite stable (MSS) tumors (HR 1.78; 95% CI 1.17-2.73) ^[118]B2a. Critically, patients with MSI-H tumors did not benefit from perioperative chemotherapy (HR for overall survival 1.50; 95% CI 0.55-4.12), whereas MSS patients did (HR 0.75; 95% CI 0.60-0.94) ^[118]B2a. This finding has led to recommendations that MSI-H status be considered when deciding on perioperative chemotherapy and that immune checkpoint blockade be investigated in this subgroup.

PD-L1 and Immune Checkpoint Inhibitors

PD-L1 combined positive score (CPS) is used to select patients for anti-PD-1 therapy. In the PLATFORM trial, maintenance durvalumab after first-line chemotherapy did not improve progression-free survival in the overall HER2-negative population, but exploratory analyses suggested a favorable overall survival effect in patients with CPS ≥5 (HR 0.63; 95% CI 0.32-1.22) and in those with an immune biomarker-positive tumor microenvironment (HR 0.60; 95% CI 0.29-1.23) ^[119]B2b. The K-Umbrella trial allocated patients with PD-L1-positive, MSI-H, or Epstein-Barr virus-related tumors to nivolumab plus paclitaxel and observed a durable survival benefit (median overall survival 12.0 vs. 7.6 months; P=0.08) ^[27]A1b.

Tumor Mutational Burden and Emerging Biomarkers

High TMB (TMB-H) has been associated with improved overall survival in chemo-refractory gastric cancer treated with the PD-1 inhibitor toripalimab: median overall survival 14.6 months for TMB-H versus 4.0 months for TMB-low (HR 0.48; 96% CI 0.24-0.96; P=0.038) ^[123]C4. PD-L1 overexpression alone did not correlate with significant survival benefit in that study ^[123]C4.

Other molecular markers have prognostic or predictive value. High ERCC1 expression was an independent poor prognostic factor for overall survival in the JCOG9912 trial (HR 1.37; 95% CI 1.08-1.75; P=0.010) ^[54]A1b. Low ATM protein expression identified patients who derived greater overall survival benefit from olaparib plus paclitaxel (HR 0.35; 80% CI 0.22-0.56) ^[35]B2b. A seven-gene signature (GC-RiskAssigner) derived from the MAGIC trial stratified perioperative chemotherapy-treated patients into high-risk (median overall survival 10.2 months) and low-risk (80.9 months) groups (HR 5.1; P<0.0001), independent of nodal status ^[52]A1b.

Controversies and Guideline Disagreement

Pearl: HER2 testing by IHC with reflex FISH for 2+ is mandatory for all advanced gastric cancer; trastuzumab deruxtecan is the preferred second-line therapy for HER2-positive disease, and MSI-H status should prompt consideration of immunotherapy and caution against perioperative chemotherapy.

Building on the molecular classification that now refines prognosis, the anatomic extent of disease, captured by the TNM staging system, remains the central determinant of treatment strategy. Staging in gastric cancer is a two-step process: clinical staging (cTNM) guides initial treatment decisions, and pathologic staging (pTNM) after surgery provides definitive prognostic information and directs adjuvant therapy.

TNM Staging System

The American Joint Committee on Cancer ( AJCC ) and Union for International Cancer Control (UICC) TNM staging system is universally used. The T category describes depth of tumor invasion through the gastric wall, the N category reflects the number of regional lymph node metastases, and the M category indicates the presence or absence of distant metastases. Clinical trials consistently enroll patients based on these categories: for example, the PRODIGY trial included patients with clinical T2-3N+ or T4Nany disease ^[103]A1b, while the CLASSIC trial enrolled patients with pathologic stage II-IIIB after D2 gastrectomy ^[135]A1b. The ACTS-GC trial included stage II or III disease ^[129]A1b, and the ARTIST 2 trial specifically required node-positive stage II or III disease after D2 resection ^[124]A1b.

Clinical Staging Modalities

Accurate clinical staging is essential to avoid unnecessary surgery and to select patients who may benefit from neoadjuvant therapy. Current staging modalities include high-quality CT scan, endoscopic ultrasound (EUS), PET, and diagnostic laparoscopy ^[14]D5. CT of the chest, abdomen, and pelvis is the primary tool for detecting distant metastases and assessing local tumor extent. EUS provides detailed assessment of T category and allows biopsy of suspicious lymph nodes. PET can identify occult metastatic disease, particularly in patients with proximal tumors. Diagnostic laparoscopy is critical for detecting peritoneal metastases that are not visible on cross-sectional imaging; the value of peritoneal lavage cytology to detect occult peritoneal disease remains under investigation ^[14]D5.

Pathologic Staging

After gastrectomy with adequate lymphadenectomy (D1+ or D2), pathologic staging is performed. The number of lymph nodes examined is a key quality metric; guidelines recommend examining at least 16 lymph nodes for accurate N categorization. Pathologic stage drives adjuvant treatment decisions. For example, the NCCN guidelines recommend postoperative chemoradiation for patients with T3-T4 tumors and node-positive T1-T2 tumors after complete resection ^[84]A1c. The JACCRO GC-07 trial demonstrated that adding docetaxel to S-1 improved 3-year relapse-free survival in pathologic stage III disease (66% vs 50%; HR 0.632, 99.99% CI 0.400-0.998) ^[126]A1b.

Staging for Advanced Disease

For patients with metastatic disease (stage IV), staging confirms the presence of distant metastases and identifies the specific sites involved. The REGATTA trial defined a single non-curable factor as liver metastases (H1), peritoneal metastases (P1), or para-aortic lymph node metastases (16a1/b2) ^[18]A1b. In that trial, gastrectomy followed by chemotherapy did not improve survival compared with chemotherapy alone (median OS 14.3 vs 16.6 months; HR 1.09, 95% CI 0.78-1.52) ^[18]A1b, underscoring that stage IV disease is managed with systemic therapy, not surgery.

Nodal Response to Preoperative Therapy

Pathologic nodal response after preoperative therapy carries prognostic significance. In a National Cancer Database analysis of clinically node-positive patients, the nodal response rate was 38.9% and was higher with preoperative radiation than with chemotherapy alone (cardia: 46.0% vs 29.1%; P<.001) ^[131]B3b. A nodal response with residual primary disease was significantly associated with improved survival (HR 0.54, 95% CI 0.44-0.65) ^[131]B3b, while ypN+ disease predicted worse survival regardless of postoperative therapy use.

Table: Staging Modalities and Their Roles

Pearl: Diagnostic laparoscopy with peritoneal lavage should be performed before initiating neoadjuvant therapy in all patients with locally advanced gastric cancer, as up to 20% of patients will have occult peritoneal metastases that upstage them to stage IV and change the treatment plan from curative-intent to palliative systemic therapy.

Staging determines the treatment approach for gastric cancer, with management stratified by disease extent, biomarker profile, and patient performance status. Multidisciplinary team management is essential for all patients with localized disease ^[136]A1c. The following sections outline the general principles; detailed modality-specific guidance is provided on dedicated pages for surgery, radiation, systemic therapy, palliative care, follow-up, and recurrent disease.

Risk Stratification and Treatment Goals

Treatment intent is defined by stage. For early-stage disease (T1N0), endoscopic or surgical resection is curative. For locally advanced disease (stage II/III), the goal is cure through multimodality therapy. For metastatic disease (stage IV), the aim is palliation, prolongation of survival, and maintenance of quality of life. Biomarker testing for HER2, PD-L1, MSI, and CLDN18.2 is mandatory before initiating first-line systemic therapy for advanced disease ^{[114]A1c[136]A1c[142]B2b}.

Early-Stage Disease (Stage I)

Endoscopic mucosal resection or endoscopic submucosal dissection is appropriate for selected T1a tumors without lymph node metastasis. For T1b or node-positive early disease, gastrectomy with D2 lymphadenectomy is standard. Laparoscopic gastrectomy is an oncologically safe alternative to open surgery in experienced centers, with comparable postoperative complications (44% vs 42%), R0 resection rates (95% vs 95%), and 1-year overall survival (76% vs 78%) ^[7]A1b.

Locally Advanced Disease (Stage II/III)

For resectable locally advanced gastric cancer, perioperative chemotherapy is the standard of care in Western countries. The PRODIGY trial demonstrated that neoadjuvant docetaxel , oxaliplatin , and S-1 (DOS) followed by surgery and adjuvant S-1 improved progression-free survival (adjusted HR 0.70, 95% CI 0.52-0.95) and overall survival (adjusted HR 0.72, 95%; 8-year OS 63.0% vs 55.1%; NNT = 13) compared with surgery and adjuvant S-1 alone in Asian patients with clinical T2-3N+ or T4Nany disease ^{[103]A1b[139]A1b}. The DRAGON IV/CAP 05 trial showed that adding camrelizumab and low-dose rivoceranib to perioperative SOX significantly improved pathologic complete response (18.3% vs 5.0%; OR 4.5, 95% CI 2.1-9.9), though survival data are pending ^[140]A1b.

In Asia, adjuvant chemotherapy after D2 gastrectomy is standard. The ARTIST 2 trial reported that adjuvant SOX (S-1 plus oxaliplatin) improved 3-year disease-free survival compared with S-1 alone (74.3% vs 64.8%; HR 0.692, P = 0.042) in node-positive stage II/III disease ^[124]A1b. The JACCRO GC-07 trial showed that adding docetaxel to S-1 improved 3-year relapse-free survival (66% vs 50%; HR 0.632, 99.99% CI 0.400-0.998; NNT = 6.25) in stage III disease ^[126]A1b.

Postoperative chemoradiation (45 Gy with fluorouracil and leucovorin) remains an option for patients with T3 or node-positive disease after suboptimal lymphadenectomy, based on the INT-0116 trial (HR for OS 1.32, 95% CI 1.10-1.60; NNT = 12 to prevent one death) ^[55]A1b. However, the ARTIST 2 trial found no additional benefit of radiotherapy when added to SOX after D2 resection (HR 0.971, P = 0.879) ^[124]A1b.

Metastatic Disease (Stage IV)

First-line systemic therapy is based on biomarker status. For HER2-negative, PD-L1 CPS ≥5 disease, nivolumab plus chemotherapy is standard. The 5-year follow-up of CheckMate 649 confirmed sustained overall survival benefit (HR 0.71, 95% CI 0.61-0.81) with 5-year OS rates of 16% vs 6% (NNT = 10) and objective response rate 58% vs 46% ^[137]A1b. For patients with CLDN18.2-positive tumors (moderate-to-strong expression in ≥40% of cells), zolbetuximab plus chemotherapy (EOX) improved progression-free survival (HR 0.44, 95% CI 0.29-0.67) and overall survival (HR 0.55, 95% CI 0.39-0.77) ^[142]B2b. For HER2-positive disease, trastuzumab plus chemotherapy is recommended ^[136]A1c.

Second-line and later-line options include ramucirumab plus paclitaxel (standard after progression on first-line therapy) ^[136]A1c. The RINDBeRG trial showed that continuing ramucirumab beyond progression with irinotecan did not improve overall survival (HR 0.91, 95% CI 0.74-1.12) compared with irinotecan alone, though progression-free survival was improved (HR 0.72, 95% CI 0.59-0.89) ^[141]A1b.

Special Considerations: Peritoneal Metastasis

For patients with peritoneal metastasis, systemic chemotherapy is the mainstay. The GASTRIPEC-I trial found no overall survival benefit from adding hyperthermic intraperitoneal chemotherapy (HIPEC) to cytoreductive surgery (median OS 14.9 months in both arms), though progression-free survival was improved (7.1 vs 3.5 months; HR not reported) ^[6]A1b. The PHOENIX-GC trial showed a trend toward improved overall survival with intraperitoneal paclitaxel plus S-1 versus cisplatin plus S-1 (median OS 17.7 vs 15.2 months; HR 0.72, 95% CI 0.49-1.04), with a significant benefit after adjusting for baseline ascites (HR 0.59, 95% CI 0.39-0.87) ^[138]A1b.

Palliative and Supportive Care

Palliative management, including systemic therapy, chemoradiation, and best supportive care, is recommended for all patients with unresectable or metastatic disease ^[136]A1c. Symptom control, nutritional support, and psychosocial care are integral components.

Controversies and Guideline Disagreement

Pearl: For metastatic gastric cancer, biomarker testing (HER2, PD-L1, MSI, CLDN18.2) is mandatory before initiating first-line therapy; the choice of regimen, nivolumab plus chemotherapy for PD-L1 CPS ≥5, trastuzumab for HER2-positive, or zolbetuximab for CLDN18.2-positive, can improve survival by 30-50% compared with chemotherapy alone.

The current standard of care for gastric cancer has been shaped by a series of landmark trials over the past two decades, each addressing a specific clinical question in the perioperative, adjuvant, or palliative setting. This section traces the evidentiary origin of today's management algorithms, highlighting the studies that established, and in some cases overturned, prior approaches.

The Adjuvant Era: From Observation to Chemoradiation and Chemotherapy

Before the turn of the century, surgery alone was the standard for resectable gastric cancer, with 5-year survival rates below 40% for node-positive disease. The Intergroup 0116 (INT-0116) trial changed this paradigm. In 559 patients with T3 or greater and/or node-positive gastric cancer after R0 resection, postoperative chemoradiation with fluorouracil, leucovorin, and 45 Gy of radiotherapy reduced the risk of death (HR 1.32, 95% CI 1.10-1.60) and relapse (HR 1.51, 95% CI 1.25-1.83) compared with observation alone ^[55]A1b. This regimen became the standard in the United States, though its applicability was limited by toxicity and the lack of a D2 lymphadenectomy requirement.

In Asia, where D2 gastrectomy is routine, adjuvant S-1 monotherapy was established as standard. The ARTIST 2 trial (n = 546) compared S-1 alone, S-1 plus oxaliplatin (SOX), and SOX plus chemoradiotherapy (SOXRT) in patients with D2-resected, stage II/III, node-positive gastric cancer. Estimated 3-year disease-free survival (DFS) rates were 64.8% with S-1, 74.3% with SOX (HR 0.692; P = 0.042), and 72.8% with SOXRT (HR 0.724 vs S-1; P = 0.074). The addition of radiotherapy to SOX did not further reduce recurrence (HR 0.971; P = 0.879) ^[124]A1b. NNT = 11 to prevent one recurrence at 3 years with SOX versus S-1.

The JACCRO GC-07 trial subsequently demonstrated that intensifying adjuvant chemotherapy improves outcomes. In 915 patients with pathologic stage III gastric cancer, S-1 plus docetaxel yielded a 3-year relapse-free survival of 66% versus 50% with S-1 alone (HR 0.632; 99.99% CI 0.400-0.998) ^[126]A1b. NNT = 6 to prevent one relapse at 3 years. Grade ≥3 neutropenia was higher with the combination, but events were manageable.

The Neoadjuvant Shift

Perioperative chemotherapy, already standard in Europe, was tested in Asian populations by the PRODIGY trial. Korean patients with clinical T2-3N+ or T4Nany gastric cancer were randomly assigned to D2 surgery plus adjuvant S-1 (SC) or neoadjuvant docetaxel, oxaliplatin, and S-1 (DOS) followed by surgery and adjuvant S-1 (CSC). The CSC group had improved progression-free survival (adjusted HR 0.70; 95% CI 0.52-0.95) ^[103]A1b. Two grade 5 adverse events (febrile neutropenia and dyspnea) occurred during neoadjuvant treatment.

The CRITICS trial compared postoperative chemotherapy versus chemoradiotherapy after neoadjuvant chemotherapy and D2 surgery. At completion of postoperative treatment, the chemotherapy group had significantly better physical functioning and less dysphagia compared with the chemoradiotherapy group ^[145]A1b. Baseline health-related quality-of-life scores were independently prognostic for event-free and overall survival.

Targeted Therapy and Immunotherapy

The addition of trastuzumab to chemotherapy for HER2-positive advanced gastric cancer, established by the ToGA trial, was incorporated into NCCN guidelines by 2013 ^[84]A1c. After progression on trastuzumab, the KC-WISE trial evaluated anbenitamab, a novel HER2-targeted agent, plus chemotherapy versus chemotherapy alone. Anbenitamab significantly improved progression-free survival (median 7.1 vs 2.7 months; HR 0.25; 95% CI 0.17-0.39; P < 0.0001) and overall survival (median 19.6 vs 11.5 months; HR 0.29; 95% CI 0.17-0.50; P < 0.0001) ^[85]A1b. Grade ≥3 treatment-related adverse events occurred in 60% versus 45%.

Immunotherapy entered the first-line landscape with CheckMate 649. At 5-year follow-up, nivolumab plus chemotherapy sustained overall survival benefit in patients with PD-L1 combined positive score (CPS) ≥5 (HR 0.71; 95% CI 0.61-0.81) and progression-free survival benefit (HR 0.71; 95% CI 0.61-0.82). Five-year overall survival rates were 16% versus 6% ^[137]A1b. NNT = 10 to prevent one death at 5 years. Objective response rate was 58% versus 46%, and median duration of response was 8.5 versus 6.9 months. No new safety concerns emerged.

Peritoneal Metastasis: An Unmet Need

Peritoneal metastasis carries a particularly poor prognosis. The PHOENIX-GC trial compared intraperitoneal and intravenous paclitaxel plus S-1 (IP) versus S-1 plus cisplatin (SP). Median overall survival was 17.7 versus 15.2 months (HR 0.72; 95% CI 0.49-1.04), failing to reach statistical significance. However, after adjustment for baseline ascites , the HR was 0.59 (95% CI 0.39-0.87) ^[138]A1b. The 3-year overall survival rate was 21.9% in the IP arm versus 6.0% in the SP arm.

The GASTRIPEC-I trial explored cytoreductive surgery (CRS) with or without hyperthermic intraperitoneal chemotherapy (HIPEC) using mitomycin C and cisplatin. Median overall survival was identical at 14.9 months in both groups. Progression-free survival was longer with HIPEC (7.1 vs 3.5 months), and distant metastasis-free survival also favored HIPEC (10.2 vs 9.2 months) ^[6]A1b. Grade ≥3 adverse events were similar (43.6% vs 38.1%).

Evolution of Biomarker-Driven Therapy

NCCN guidelines have progressively incorporated biomarker testing, HER2, microsatellite instability (MSI), and PD-L1 expression, into routine practice ^{[114]A1c[136]A1c[100]A1c}. The Pan-Asian adapted ESMO guidelines (2019, 2024) further tailored recommendations to ethnic differences in screening, molecular profiling, and drug access across Asia ^{[69]A1c[22]A1c}. These adaptations reflect the growing recognition that gastric cancer is a molecularly heterogeneous disease requiring personalized treatment strategies.

Pearl: The evolution from one-size-fits-all adjuvant chemoradiation to biomarker-driven perioperative and palliative regimens, exemplified by the INT-0116, ARTIST 2, CheckMate 649, and KC-WISE trials, underscores that treatment selection must now integrate histologic subtype, HER2 status, PD-L1 CPS, and MSI status at every decision point.

The evolution of multimodal therapy has improved outcomes, but prognosis in gastric cancer remains heterogeneous and is determined by a constellation of clinical, pathologic, and molecular factors. Survival estimates vary widely by stage at diagnosis, treatment received, and tumor biology.

Survival Outcomes by Disease Stage

For patients with resectable disease, 5-year overall survival (OS) after surgery alone is approximately 43-44% for stage II/III, with no benefit from adjuvant FAMTX or FEMTX ^[146]A1a. In the ARTIST 2 trial, patients with D2-resected, node-positive stage II/III gastric cancer achieved 3-year disease-free survival (DFS) of 64.8% with S-1 monotherapy, 74.3% with SOX (S-1 plus oxaliplatin ), and 72.8% with SOX plus chemoradiotherapy (SOXRT) ^[124]A1b. The addition of docetaxel to S-1 in stage III disease improved 3-year relapse-free survival from 50% to 66% (HR 0.632) ^[126]A1b. Neoadjuvant docetaxel, oxaliplatin, and S-1 (DOS) followed by surgery and adjuvant S-1 improved progression-free survival (PFS) compared with surgery plus adjuvant S-1 alone (HR 0.70) ^[103]A1b.

In the advanced/metastatic setting, median OS with first-line chemotherapy alone is approximately 8.7-9.5 months ^[147]A1a. The addition of nivolumab in patients with PD-L1 combined positive score (CPS) ≥5 yields a 5-year OS rate of 16% versus 6% with chemotherapy alone (HR 0.71) ^[137]A1b. For HER2-positive disease, second-line anbenitamab plus chemotherapy after trastuzumab failure produced a median OS of 19.6 months versus 11.5 months (HR 0.29) ^[85]A1b. Third-line apatinib improved median OS from 4.7 to 6.5 months (HR 0.709) ^[73]A1b. Second-line chemotherapy overall reduces the risk of death by 36% compared with best supportive care (HR 0.64) ^[132]B2a.

Prognostic Factors

Multiple factors independently predict survival. The CRITICS trial identified that baseline worse social functioning (HR 2.20), nausea (HR 1.89), worse WHO performance status (HR 1.55), and diffuse or mixed histologic subtype (vs intestinal: HR 1.94 and 2.35, respectively) were significantly associated with worse event-free survival and OS ^[145]A1b. Nodal response to preoperative therapy is a powerful prognosticator: patients who achieve nodal clearance (ypN0) have significantly better survival, whereas those with residual nodal disease (ypN+) fare poorly regardless of postoperative therapy ^[131]B3b. A nodal response with residual primary disease confers a survival benefit (HR 0.54) ^[131]B3b.

Molecular and biomarker factors further stratify prognosis. PD-L1 CPS ≥5 identifies patients most likely to benefit from nivolumab plus chemotherapy ^[137]A1b. Deficient mismatch repair (dMMR)/microsatellite instability-high (MSI-H) status and Epstein-Barr virus (EBV) positivity are associated with higher pathologic complete response rates to chemoimmunotherapy (pCR 43.8% and 25%, respectively, in the DRAGON IV trial) ^[140]A1b. HER2 positivity enables targeted therapy with trastuzumab and, after progression, anbenitamab ^[85]A1b.

Controversies and Guideline Disagreement

Pearl: Nodal response to preoperative therapy is the single most powerful modifiable prognostic factor, achieving ypN0 confers a survival advantage regardless of primary tumor response, and the absence of nodal response should prompt consideration of alternative postoperative strategies ^[131]B3b.

Prognostic models derived from clinical trial populations may not fully capture outcomes in older adults, who are underrepresented in pivotal studies. The ARMANI phase III trial provides the most direct evidence on age-related efficacy and safety in advanced gastric cancer. In a subgroup analysis of patients aged ≥70 years versus <70 years, no significant differences were observed in progression-free survival (P=0.757), overall survival (P=0.588), or overall response rates (P=0.238) ^[152]A1b. The effect of treatment arm ( paclitaxel plus ramucirumab switch maintenance versus continuation of first-line oxaliplatin -based chemotherapy) on survival was similar across age groups (P=0.094) ^[152]A1b. Toxicity profiles differed: peripheral neuropathy occurred more frequently in younger patients (P=0.020), while anemia was more common in older patients (P=0.044) ^[152]A1b. Hand-foot syndrome was more frequent in women (P=0.021) ^[152]A1b. Older patients were less likely to receive post-discontinuation treatment (P=0.010), especially in the continuation arm (P=0.026) ^[152]A1b. Quality of life impact favored the switch maintenance strategy irrespective of age and sex ^[152]A1b.

Elderly Patients

• Efficacy: Survival benefit from ramucirumab-based switch maintenance is preserved in patients ≥70 years, with no significant interaction by age ^[152]A1b.
• Toxicity: Anemia is more common in older patients; peripheral neuropathy is less frequent. Close monitoring of hemoglobin and proactive management of anemia are warranted ^[152]A1b.
• Post-discontinuation therapy: Older patients are less likely to receive further lines of treatment, which may contribute to shorter overall survival in this subgroup ^[152]A1b.
• Trial limitations: Many landmark trials exclude patients >75 years (e.g., DRAGON IV/CAP 05 ^[140]A1b), limiting evidence in the very elderly. Clinicians should extrapolate cautiously and consider performance status rather than chronological age alone.

Other Special Populations

No data from the provided evidence address pregnancy, fertility preservation, or immunocompromised patients with gastric cancer. Management in these groups should follow general oncologic principles with multidisciplinary input, but specific recommendations cannot be made from the available literature.

Pearl: In older adults with advanced gastric cancer, ramucirumab-based switch maintenance offers similar efficacy to younger patients, but anemia is more frequent and post-discontinuation therapy is less common, both factors that should inform treatment planning and follow-up.