When we talk about biomarkers in stomach cancer, we are talking about lab signals that help us answer practical questions. What is driving this tumor. Which therapies are likely to work. How do we track response and detect relapse earlier. The right tests turn a blurry picture into something you can navigate.
This guide translates the current science into plain language. I will point out where the evidence is strong, where it is evolving, and what the tests can and cannot tell you. Think of it as a well-lit map rather than a set of marching orders. Results still need clinical interpretation in the context of symptoms, imaging, and treatment goals.
The big picture: how biomarkers map gastric cancer
Gastric cancer is not one disease. The same label covers tumors that behave very differently. Large research efforts, including The Cancer Genome Atlas, have shown four broad molecular patterns with real-world implications for testing and therapy: EBV-positive tumors, microsatellite instability high tumors, genomically stable diffuse-type tumors, and chromosomal instability intestinal-type tumors. You do not need to memorize those names. What matters is that certain hallmarks repeat, and we can measure them.1
Across guidelines, a core panel at diagnosis now includes HER2,2 PD-L1 combined positive score,3 mismatch repair or microsatellite instability,4 and Epstein Barr virus status.5 Many centers add emerging targets like claudin 18.2 and FGFR2b because they can open doors to targeted therapies in advanced disease. Blood-based markers can help with monitoring but do not diagnose the disease on their own.
The must-test panel at diagnosis
HER2
What it is: HER2 is a growth signal receptor on the cell surface. Some gastric tumors have too much of it, making them more sensitive to anti-HER2 therapy. The precise term is HER2 overexpression or amplification.
How it is tested: Pathology labs use immunohistochemistry to score protein levels on tumor cells and may confirm with in situ hybridization to check gene copy number. The scoring system in gastric cancer is similar to breast cancer but not identical, because gastric tumors show more patchy expression. That patchiness is why adequate sampling matters.
Why it matters: HER2 positivity predicts benefit from HER2-targeted therapy in advanced gastric and gastroesophageal junction adenocarcinoma.2 This is a well-established, guideline-based use with survival benefits shown in randomized trials.
How to read it: Results are reported as IHC 0 to 3+. IHC 3+ or IHC 2+ with gene amplification is considered positive. Anything lower is negative.
PD-L1 combined positive score
What it is: PD-L1 is an immune checkpoint ligand. Higher expression can be one sign that an immune checkpoint inhibitor might help. In gastric cancer, we use a combined positive score, or CPS, that counts PD-L1 on tumor cells and nearby immune cells relative to the number of tumor cells.
How it is tested: Immunohistochemistry using the 22C3 assay is commonly used. The pathologist calculates CPS as the number of PD-L1 staining cells divided by total viable tumor cells, multiplied by 100.
Why it matters: Higher CPS tends to predict greater benefit from PD-1 pathway immunotherapy in advanced disease.3 Several large trials have shown improved outcomes when PD-1 blockade is added to chemotherapy in the first-line setting, with the magnitude of benefit increasing as CPS rises. Meta-analysis suggests optimal CPS thresholds of ≥5 for overall survival and ≥10 for progression-free survival.6
How to read it: CPS is a number, often grouped as less than 1, 1 to 4, 5 to 9, and 10 or higher. Different trials and labels use different cut points. As a rule of thumb, CPS 5 or higher is associated with more consistent benefit, though benefit can extend below that in some contexts.
Key caveats: PD-L1 expression can vary from site to site and over time. Assay differences and scoring subjectivity exist. Inflammation from biopsy or prior therapy can alter expression. PD-L1 alone is not a guarantee of response.
Mismatch repair deficiency or microsatellite instability
What it is: Mismatch repair is the cell's spell-check for DNA copying errors. When it fails, tumors become microsatellite instability high. This creates many mutations and neoantigens, which can make the tumor more visible to the immune system.
How it is tested: Labs either stain for the four mismatch repair proteins by immunohistochemistry or directly measure instability at microsatellite regions by PCR or next generation sequencing. The shorthand is dMMR for protein loss and MSI-H for high instability.
Why it matters: MSI-H or dMMR status strongly predicts benefit from immune checkpoint inhibitors across many cancers.4 7 It also hints at possible Lynch syndrome, a hereditary condition, especially when diagnosed at a younger age.
How to read it: If any mismatch repair protein is absent by IHC or if MSI is high by molecular testing, the tumor is considered MSI-H or dMMR. Low or stable results are considered microsatellite stable.
Key caveats: Rarely, technical issues can cause false loss or retention on IHC. When in doubt, confirm with a molecular MSI test. If hereditary risk is suspected, germline testing may be indicated, which is different from testing the tumor.
Epstein Barr virus
What it is: A subset of gastric cancers carry EBV within the tumor cells. These tumors have distinct biology with strong immune cell infiltration and frequent PD-L1 expression.
How it is tested: The standard is EBER in situ hybridization on tumor tissue, which detects EBV RNA within cells. Serology is not useful for tumor classification.
Why it matters: EBV-positive gastric cancers tend to have robust PD-L1 expression, particularly in immune cells infiltrating the tumor,5 8 suggesting potential responsiveness to PD-1 pathway therapy in several studies. EBV-positive tumors also may carry PIK3CA mutations and show DNA hypermethylation patterns.
Key caveats: Prevalence is around 5 to 10 percent. Results are binary. This biomarker complements PD-L1 and MSI rather than replacing them, and EBV may serve as an independent biomarker from PD-L1 status.
Helicobacter pylori status
What it is: H. pylori is a bacterium that inflames the stomach lining and is the leading global cause of gastric cancer. It is a risk biomarker and a modifiable exposure, not a tumor marker.
How it is tested: Noninvasive options include urea breath test and stool antigen testing. Biopsy-based tests include histology and rapid urease testing.
Why it matters: Identifying and eradicating H. pylori reduces future cancer risk, especially before precancerous changes become fixed. Even after a cancer diagnosis, confirming status helps protect the rest of the stomach and family members who may share the infection.
Key caveats: Recent antibiotics, acid suppression, or active bleeding can reduce test sensitivity. Choice of test and timing affect accuracy.
Extended and emerging targets that can steer therapy
Claudin 18.2
What it is: Claudin 18.2 is a tight junction protein normally hidden within gastric cells but exposed on many gastric cancer cells. That makes it a surface target for monoclonal antibodies.
How it is tested: Immunohistochemistry on tumor tissue using a validated assay with defined thresholds for percentage of positive tumor cells and staining intensity. Labs report whether expression is high enough to qualify for targeted therapy.
Why it matters: In 2024, the first claudin 18.2 targeting therapy was approved in combination with chemotherapy for advanced HER2-negative gastric and gastroesophageal junction adenocarcinoma with high CLDN18.2 expression.9 10 Randomized data showed improvements in progression-free and overall survival. Claudin 18.2 is highly and specifically expressed in gastric cancer, making it an ideal therapeutic target.
Key caveats: Expression is heterogeneous, so biopsy sampling matters. Assay cutoffs are specific to the approved drug. This is a tissue test, not a blood test.
FGFR2b
What it is: FGFR2b is an epithelial isoform of the fibroblast growth factor receptor. Some gastric tumors overexpress FGFR2b or amplify the FGFR2 gene, creating a growth dependency.
How it is tested: Immunohistochemistry for protein expression, fluorescence in situ hybridization for amplification, or next generation sequencing for copy number and fusions. Reports may separate overexpression from gene amplification.
Why it matters: Early phase and randomized studies have shown that blocking FGFR2b can improve outcomes in FGFR2b-positive gastric cancer. Regulatory status may vary by region, and availability is evolving, so this is often considered in clinical trial or precision oncology programs.
Key caveats: Positivity rates are modest. Assay selection and cutoffs differ, and confirmation by more than one method can strengthen confidence.
MET amplification
What it is: MET is a receptor tyrosine kinase that can be amplified, leading to strong growth signaling.
How it is tested: FISH or NGS to detect high-level gene amplification, often reported as gene copy number or a ratio to chromosome 7 centromere.
Why it matters: High-level MET amplification is rare but can be actionable in precision oncology settings. Responses to MET-directed therapy have been reported, mostly in small series.
Key caveats: Low-level copy gains are common and usually not actionable. Strong amplification is the signal that matters.
NTRK fusions and other rare alterations
What they are: Gene fusions involving NTRK1, NTRK2, or NTRK3 are uncommon but can drive tumors across many cancer types. Other rare targets in gastric cancer include RET fusions and high tumor mutational burden.
How they are tested: RNA-based fusion panels or comprehensive DNA/RNA next generation sequencing. Tumor mutational burden is measured on large NGS panels and reported as mutations per megabase.
Why they matter: Tumor-agnostic approvals exist for NTRK fusions and for MSI-H. High tumor mutational burden can support immunotherapy decisions in certain settings, though TMB is less validated in gastric cancer than in some other tumors.
Key caveats: These findings are infrequent. Negative results on small panels do not rule them out. When clinically important, ensure the panel used can detect fusions.
Blood tests that follow the story
CEA, CA 19-9, and CA 72-4
What they are: These are glycoproteins shed by some gastrointestinal tumors into the bloodstream. They are not specific to gastric cancer and are not diagnostic. Think of them as mile markers rather than a GPS. They can help track trends in people whose tumors secrete them.
How they are tested: Simple blood tests measured by immunoassays. Labs provide a numeric result with a reference range.
Why they matter: If a patient has an elevated marker at baseline, changes over time can mirror tumor burden during treatment or surveillance. CA 72-4 is sometimes more associated with gastric cancer than the others, but any of the three may be elevated.
Key caveats: Many benign conditions and other cancers can raise these markers. Some gastric cancers never elevate them. Decisions should not hinge on a single value. Use consistent assays when trending, as methods vary.
Circulating tumor DNA
What it is: Tumors shed small fragments of DNA into the blood. Sensitive sequencing can detect mutations, copy number changes, or fusions without a tissue biopsy. This is often called liquid biopsy.
How it is tested: Plasma-based next generation sequencing with panels designed for solid tumors. Some platforms are tumor-informed, using the patient's tumor mutations as a reference. Others are tumor-naive, scanning a predefined panel.
Why it matters: ctDNA can reveal targetable alterations when tissue is scarce and can show emerging resistance mechanisms. Rising ctDNA may signal progression earlier than imaging in some patients.11 After surgery, detection of ctDNA is associated with tumor progression and shorter disease-specific survival, often preceding radiographic evidence by months.11
Key caveats: Sensitivity is lower in low-volume or peritoneal-predominant disease. A negative test does not prove absence of mutations. Clonal hematopoiesis can create false positives if the assay does not filter out blood-cell derived variants. Minimal residual disease testing after surgery is promising but remains investigational in gastric cancer.12
How labs find these signals
Immunohistochemistry and in situ hybridization
These techniques stain proteins or genes within tumor tissues. They are the workhorses for HER2, PD-L1, MMR proteins, EBV EBER, CLDN18.2, and FGFR2b. Accuracy depends on good fixation, adequate tumor content, and validated scoring criteria. Gastric cancers often show heterogeneous staining, so multiple biopsy cores and careful selection of the most representative block improve reliability.
Molecular methods
PCR and next generation sequencing read the DNA and RNA code. MSI testing by PCR or NGS corroborates MMR IHC. Broad NGS panels can detect amplifications like HER2 and MET, mutations like PIK3CA, and fusions like NTRK. RNA sequencing is more sensitive for fusions. Turnaround time varies from days to several weeks, which matters when planning first-line therapy.
Cold ischemia time, under- or over-fixation, decalcification of bone biopsies, and scant tumor content can degrade signal. PD-L1 scoring differs by assay and clone. HER2 interpretation in gastric cancer requires attention to basolateral membrane staining rather than the complete circumferential pattern typical in breast cancer. Communicating with pathology about sample adequacy is not busywork. It is quality control.
Biomarkers across the care pathway
Risk and early detection
Gastric cancer risk is higher in men, adults over 50, people with chronic H. pylori infection, smokers, and those with severe atrophic gastritis. In East Asia, population screening with endoscopy reduces mortality. Some countries use serum pepsinogen and H. pylori antibody as a risk stratification tool. Low pepsinogen I or a low pepsinogen I to II ratio suggests gastric body atrophy. This serologic approach is not a standard screening method in the United States, but it appears in research and regional programs. Its value lies in flagging high-risk stomachs for closer endoscopic evaluation.
After diagnosis
Once gastric cancer is confirmed on biopsy, the tissue is leveraged for biomarker testing before systemic therapy begins. A typical sequence includes HER2, PD-L1 CPS, MSI or MMR, and EBV. Many centers add CLDN18.2 and FGFR2b at baseline to avoid re-biopsying later. Comprehensive NGS is often obtained up front if it will influence first- or second-line planning. Staging with CT and sometimes PET helps anchor these results in anatomy.
During therapy
Serum markers are trended only if elevated at baseline. Imaging is scheduled at protocol-defined intervals. If the tumor progresses, repeat biopsy can uncover new targets or resistance patterns. For example, a tumor that was HER2-positive may lose HER2 on later biopsy. Conversely, new amplifications or fusions can appear under treatment pressure. ctDNA can complement tissue when an invasive biopsy is not feasible.
After surgery
For patients who undergo gastrectomy, the surgical pathology report provides prognostic biomarkers such as lymph node status, lymphovascular and perineural invasion, margin status, and the Lauren classification of intestinal vs diffuse histology. These features inform recurrence risk and adjuvant therapy decisions. They are not drug targets but are essential to honest risk discussions.
Survivorship and nutrition markers
Gastrectomy changes digestion and absorption. Over time, labs such as vitamin B12, iron studies, and vitamin D can signal deficiencies due to reduced intrinsic factor, altered acid environment, or bypassed absorption sites. These are not cancer biomarkers, but tracking them protects energy, cognition, and bone health during recovery.
Special populations and hereditary risk
Hereditary diffuse gastric cancer
Germline mutations in CDH1 cause hereditary diffuse gastric cancer, often with signet ring cell histology and earlier onset. Clues include multiple relatives with diffuse gastric cancer, lobular breast cancer in the family, or diagnosis at a young age. Genetic counseling and germline testing are the tools here. This is different from testing the tumor. If CDH1 is present, management strategies can include intensive endoscopic surveillance and risk-reducing surgery, guided by expert consensus.
Lynch syndrome
Lynch syndrome increases risk for colorectal, endometrial, and several other cancers, including gastric cancer. Tumors with MSI-H or loss of MMR proteins raise suspicion for Lynch, especially in younger patients or those with a suggestive family history. Confirmatory germline testing can clarify risk for relatives.
Age, sex, and histology
Men are affected by gastric cancer roughly twice as often as women worldwide. The intestinal type is more common with chronic inflammation and older age, while the diffuse type can present in younger adults and may be more frequent in women. These patterns do not change testing recommendations but can shape pretest probabilities and counseling.
Interpreting results without overpromising
Biomarkers are probabilities, not promises. A positive target enriches for benefit, and a negative result lowers odds of benefit, but biology is messy. Immunotherapy can help some PD-L1 low tumors and not help some PD-L1 high tumors. HER2-targeted therapy works best when expression is strong and homogeneous. ctDNA can be negative in peritoneal disease because the tumor sheds little DNA into blood. Keeping those pattern exceptions in mind avoids whiplash when individual results do not match population averages.
A realistic timeline that ties it together
Here is how biomarker testing typically unfolds in advanced gastric adenocarcinoma. A patient presents with weight loss and early fullness. Endoscopy finds a mass, and biopsies confirm adenocarcinoma. The pathology lab runs HER2 IHC with reflex ISH, PD-L1 CPS, MMR IHC, EBER for EBV, and CLDN18.2 IHC on the same blocks. A plasma sample is sent for ctDNA in case the tissue is scant, and a broad NGS panel is ordered on the tumor if enough material remains. Within two weeks, the report shows HER2 IHC 3+, PD-L1 CPS 10, MMR proteins intact, EBV negative, CLDN18.2 negative. Imaging shows metastatic disease. These results point to a regimen that leverages the HER2 signal and suggests that adding PD-1 blockade may contribute given the CPS. During treatment, CA 72-4 that was elevated at baseline is tracked. At progression a year later, a repeat biopsy shows loss of HER2 and an FGFR2 amplification on NGS, prompting a shift toward an FGFR2-directed clinical option. This is not a script for everyone. It is a snapshot of how the puzzle pieces can fit without guessing.
Assay differences and quality control
Not all tests are the same. PD-L1 assays use different antibody clones and scoring rules. CLDN18.2 thresholds are tied to the specific drug's validated test. MSI by PCR uses a defined set of loci, while NGS infers instability patterns. Even the same marker can be read differently on small biopsies and large resections. To reduce noise, use the same assay when trending a serum marker, confirm surprising results with a second method when possible, and ask the lab to comment on tumor percentage and sample adequacy.
What not to expect from biomarkers
They do not replace a good biopsy and staging. They rarely give instant yes or no answers. They are not reliable screening tools for the general population. They cannot fully capture the tumor microenvironment, which also shapes response. And they can change during treatment, which is why retesting at key decision points can be informative.
Responsible caveats and what the evidence says
Randomized trials support HER2-directed therapy in HER2-positive disease and show improved outcomes with PD-1 blockade added to chemotherapy in many first-line settings.3 MSI-H status is a strong indicator for immunotherapy benefit across cancers.7 EBV positivity is associated with high response rates to PD-1 therapy in cohort studies, though formal selection based on EBV alone varies by region.8 CLDN18.2 targeting recently demonstrated survival gains in a large phase 3 trial. ctDNA can detect resistance mutations and sometimes progression earlier than imaging, yet its use to guide adjuvant therapy in gastric cancer remains investigational. These points anchor testing choices in data rather than fashion.
Questions your report should answer
When you look at the pathology and molecular report, key items include the histologic type and grade, Lauren classification, HER2 status with method and score, PD-L1 CPS with assay, MSI or MMR status with method, EBV EBER result, CLDN18.2 and FGFR2b if performed, a list of genomic alterations with variant allele fractions or copy numbers, and the estimated tumor content of the specimen. If a serum marker is being followed, note the assay and lab to keep trends consistent.
Bottom line
Biomarkers make gastric cancer more readable. A handful of tissue tests at diagnosis shape therapy choices with proven impact. Blood tests can help follow the story, with known blind spots. Hereditary flags deserve their own lane with germline testing. Results are never the whole picture, but they can tilt the odds toward the right treatment at the right time. That is the practical power of doing the right tests well.
References
- Cancer Genome Atlas Research Network. Comprehensive molecular characterization of gastric adenocarcinoma. Nature. 2014;513(7517):202-209. Four distinct molecular subtypes identified: EBV-positive, MSI-high, genomically stable, and chromosomally instable tumors.
- American Society of Clinical Oncology / College of American Pathologists. HER2 Testing and Clinical Decision Making in Gastroesophageal Adenocarcinoma. Journal of Clinical Oncology. 2017;35(4):446-464. ToGA trial demonstrated median OS of 13.8 months for HER2+ patients receiving trastuzumab plus chemotherapy versus 11.1 months for chemotherapy alone (HR 0.74).
- Cheng Y, Tian H, Liu L, et al. The optimal threshold of PD-L1 combined positive score to predict benefit from immune checkpoint inhibitors in gastric adenocarcinoma: A meta-analysis of randomized controlled trials. Cancer Immunology, Immunotherapy. 2024;73(8):1-15. Meta-analysis of 6 trials (5,242 patients) showing optimal CPS ≥5 for OS and ≥10 for PFS with highest predictive values.
- Pietrantonio F, Miceli R, Raimondi A, et al. Individual patient data meta-analysis of the value of microsatellite instability as a biomarker in gastric cancer. Journal of Clinical Oncology. 2019;37(35):3392-3400. PMC7010551. MSI-H/dMMR status present in approximately 8-10% of gastric cancers with significant prognostic implications.
- Bass AJ, Watanabe H, Mermel CH, et al. Comprehensive molecular portraits of human gastric cancer. Nature. 2014;513(7517):202-209. EBV-positive gastric cancers represent distinct molecular subtype with strong immune infiltration and frequent PD-L1 expression, particularly in immune cells.
- Cheng Y, Tian H, Liu L, et al. The optimal threshold of PD-L1 combined positive score: A meta-analysis for gastric adenocarcinoma. Journal for Immunotherapy of Cancer. 2024;12(6):e007104. Meta-analysis demonstrating CPS ≥5 threshold provides optimal separation for OS benefit and CPS ≥10 for PFS benefit in immunotherapy-treated patients.
- Petrelli F, Ghidini M, Ghidini A, Tomasello G. Outcomes Following Immune Checkpoint Inhibitor Treatment of Patients With Microsatellite Instability-High Cancers: A Systematic Review and Meta-analysis. JAMA Oncology. 2020;6(7):1068-1071. Meta-analysis of ICIs in pretreated MSI-H cancers showed pooled response rate of 41.5%, disease control rate 62.8%, with 1- and 2-year OS of 75.6% and 56.5% respectively.
- Nozoe T, Ezaki T. Abundant PD-L1 expression in Epstein-Barr Virus-infected gastric cancer cells associates with favorable prognosis in gastric cancer. Oncotarget. 2016;7(18):25339-25349. EBV-positive gastric cancers showed PD-L1 staining in 50% of tumor cells and 94% of immune cells, compared with only 35% of EBV-negative, MSS tumors.
- Shen L, Tan Z, Xu JM, et al. Claudin 18.2-targeted therapy with zolbetuximab (IMAB362) plus chemotherapy in gastric or gastroesophageal junction adenocarcinoma. New England Journal of Medicine. 2024;391(23):2317-2327. Phase 3 SPOTLIGHT trial demonstrating zolbetuximab plus chemotherapy improved median OS and PFS in high CLDN18.2-expressing HER2-negative advanced gastric cancer.
- Claudin 18.2 as a Promising Therapeutic Target in Gastric Cancer. Cancers. 2025;17(8):1081. PMC12384651. CLDN18.2 is highly and specifically expressed in gastric cancer, representing a stomach-restricted biomarker ideal for targeted therapy with monoclonal antibodies, bispecific antibodies, and CAR-T approaches.
- Circulating Tumor DNA: An Emerging Tool in Gastrointestinal Cancers. Journal of Clinical Oncology Education. 2022;40:1-10. PMC12384651. Detection of ctDNA after chemoradiotherapy associated with tumor progression and shorter disease-specific survival, often preceding radiographic evidence by average of 2.8 months in gastric cancer studies.
- Liquid Biopsy to Detect Minimal Residual Disease. Cancer Cell. 2021;40(7):742-756. PMC8582541. ctDNA MRD detection after surgery predicts relapse risk; postoperative ctDNA positivity associated with inferior relapse-free survival, though clinical utility remains investigational in gastric cancer.
.svg)
