Clinical Guide to Biomarkers in Oesophageal Cancer

If you or someone you care about is navigating oesophageal (esophageal) cancer, biomarkers are the part of the report that looks like the "spec sheet." They don't replace the basics, like stage and tumor type, but they help answer key questions: What treatments might fit? How likely is benefit from immunotherapy? Is there a targetable signal? Below is a practical tour of which biomarkers matter, how they are measured, what the results mean, and where the science is solid versus still evolving.

The two cancers under one name

Oesophageal cancer is really two diseases with different biology. Adenocarcinoma typically arises near the gastroesophageal junction and is linked to reflux and Barrett's esophagus. Squamous cell carcinoma (ESCC) tends to occur higher in the esophagus and is tied to tobacco, alcohol, and, in some populations, nutritional exposures. This split matters because some biomarkers are far more informative in one subtype than the other.

What biomarkers can and cannot do

Biomarkers are measurements from tumor tissue or blood that carry diagnostic, prognostic, or predictive information. In oesophageal cancer today, the clinically useful ones are mostly predictive. They tilt the odds toward or away from benefit with specific classes of therapy, especially immunotherapy and HER2-directed treatments in adenocarcinoma. There are no reliable blood tests to screen for oesophageal cancer, and common "tumor markers" like CEA or CA 19-9 are not recommended for early detection.

The core biomarker panel most people will encounter

For advanced or metastatic disease, most guidelines support testing the tumor for the following, ideally early in the treatment course so results can guide first-line decisions:

• PD-L1 by combined positive score (CPS)¹
• HER2 by immunohistochemistry (IHC) with reflex in situ hybridization (ISH) if equivocal, primarily in adenocarcinoma and gastroesophageal junction tumors²
• Mismatch repair status (dMMR) or microsatellite instability (MSI)³
• Broad next-generation sequencing (NGS) panel if available to identify rare but actionable alterations and to report tumor mutational burden (TMB)

For earlier-stage disease that is being treated with surgery and chemoradiation, pathology and stage remain the main drivers, though select biomarkers still matter for future options or trials.

PD-L1 CPS: the immunotherapy signal

What it is: PD-L1 is a protein tumors and immune cells can express to dampen immune attack. The combined positive score (CPS) is a way of quantifying PD-L1 staining that counts both tumor cells and tumor-associated immune cells and divides by the number of viable tumor cells, then multiplies by 100. A CPS of 10 means roughly 10 PD-L1 positive cells per 100 tumor cells when both compartments are considered.

Why it matters: Across several large randomized trials in oesophageal cancer, higher PD-L1 CPS correlates with greater benefit from adding a PD-1 inhibitor to chemotherapy, especially in ESCC and in adenocarcinoma at the gastroesophageal junction.^{1 4 5} Trials like KEYNOTE-590 and CheckMate 648 used CPS thresholds such as 10 to enrich for responders, while showing some benefit across broader groups. Regulators in different countries have anchored approvals to PD-L1 in slightly different ways, but the clinical principle is consistent: a higher CPS increases the likelihood of meaningful benefit from PD-1 based therapy.

How it's measured: Your report will list a CPS number generated with a specific PD-L1 antibody clone and assay (for example, 22C3). Different assays are not fully interchangeable, which is one reason results must be interpreted in context.

Limitations to know:

• Heterogeneity is real. A primary tumor and a metastasis can have different CPS values.
• Timing matters. PD-L1 can change after chemotherapy or radiation.
• Borderline results can wobble. A CPS of 8 in one sample may be 12 in another. When a treatment decision hinges on a cutoff, repeating PD-L1 on another block or site can be reasonable.

HER2 in adenocarcinoma: the classic target

What it is: HER2 (ERBB2) is a receptor that, when overexpressed or amplified, can drive tumor growth. In oesophagogastric adenocarcinomas, HER2 behaves similarly to how it does in breast cancer, though expression can be more patchy.

Why it matters: In HER2-positive metastatic or unresectable adenocarcinoma of the esophagus or gastroesophageal junction, adding HER2-directed therapy to chemotherapy improves outcomes.^{2 6} The landmark ToGA study established the principle more than a decade ago, and newer antibody-drug conjugates have shown activity after initial therapy, including in tumors with lower levels of HER2 expression in some settings, though definitions and best use are still being refined.

How it's measured: Testing starts with IHC. IHC 3+ is positive. IHC 2+ is equivocal and should reflex to ISH (also called FISH) to check for ERBB2 gene amplification. IHC 0–1+ is negative. Reports may note heterogeneity, which is common in gastric and junctional tumors.

Who should be tested: All adenocarcinomas of the esophagus and gastroesophageal junction with advanced or metastatic disease should be tested. In early-stage disease, HER2 status does not currently change standard surgical or chemoradiation plans, but it can still inform trial options and future systemic therapy if recurrence occurs.

Pitfalls:

• Sample choice matters. Tiny biopsies can miss focal HER2-positive areas.
• HER2 can "convert" over time. Re-biopsy at progression can uncover new positivity or loss of expression.
• Bone metastases processed with acid decalcification often yield uninterpretable HER2 results.

MSI and mismatch repair: uncommon but high-impact

What it is: Tumors with deficient mismatch repair (dMMR) accumulate DNA errors and can exhibit microsatellite instability (MSI). These cancers tend to have many mutations, which can make them more visible to the immune system.

Why it matters: MSI-high or dMMR status predicts strong responsiveness to immune checkpoint blockade across many cancers.^{3 7} In oesophageal and junctional adenocarcinoma, MSI-high is uncommon, but when present it is clinically meaningful.

How it's measured: Labs either stain for mismatch repair proteins (MLH1, MSH2, MSH6, PMS2) by IHC or directly measure MSI by PCR or NGS.⁷ The two approaches are complementary.

Context points:

• Prevalence is low in oesophageal primaries but higher in true gastric cancers. Junctional tumors sit in the middle.
• Because of the therapeutic implications, testing is reasonable in most advanced cases even when chances are low.
• MSI and TMB overlap but are not identical. A tumor can be TMB-high without being MSI-high and vice versa.

TMB and broad genomic profiling: casting a wider net

Tumor mutational burden (TMB) is a count of mutations per megabase of DNA tested.⁸ Higher TMB sometimes correlates with better response to immunotherapy. In oesophageal cancer, TMB can add nuance when PD-L1 is borderline, but it is not a stand-alone decision-maker. Cutoffs like 10 mutations per megabase are used by some assays, yet results vary by platform and panel size.

Broad next-generation sequencing (NGS) can also detect gene amplifications, point mutations, and fusions. Actionable findings beyond HER2 and MSI are uncommon in oesophageal cancer but not zero. NTRK fusions are rare and tumor-agnostic targeted options exist for them. MET amplification, KRAS mutations, and EGFR alterations show up in some cases, but targeted strategies have not consistently improved outcomes in unselected oesophageal populations. These results most often inform clinical trial eligibility rather than standard therapy.

Squamous cell carcinoma: what's different

PD-L1 is particularly informative in ESCC, where high CPS is common and correlates with benefit from PD-1 blockade combined with chemotherapy in first-line treatment.¹ HER2 is typically not relevant in ESCC. EGFR overexpression occurs, but EGFR inhibitors have not become standard of care in this disease. As for blood-based markers, squamous cell carcinoma antigen (SCCA) and CYFRA 21-1 can be elevated but are neither specific nor sensitive enough for diagnosis or screening. Some clinicians track them during treatment in ESCC as a rough trend marker, yet changes need to be interpreted alongside imaging and symptoms, not in isolation.

Barrett's esophagus and early adenocarcinoma: where biomarkers fit

Barrett's esophagus is a change in the lining of the lower esophagus that raises the risk of adenocarcinoma over time. Surveillance relies on endoscopy and biopsies. p53 immunohistochemistry is commonly used by pathologists to support a diagnosis of dysplasia in Barrett's, but p53 status does not guide systemic treatment in invasive cancer. When early adenocarcinoma is removed endoscopically, that tissue also provides staging clues like depth of invasion and lymphovascular involvement, which often matter more than molecular markers at that stage.

Circulating tumor DNA (ctDNA): promising, not a crystal ball

ctDNA looks for fragments of tumor DNA in the bloodstream. In oesophageal cancer, ctDNA is useful in a few ways:

• Genomic profiling when tissue is scarce
• Monitoring treatment response by tracking specific mutations over time
• Research on minimal residual disease after surgery or chemoradiation^{9 10}

The minimal residual disease use case is particularly compelling. If ctDNA is detectable after curative-intent therapy, relapse risk is higher; if it's undetectable, risk is lower.^{9 10} Several studies in upper GI cancers support this risk stratification signal, though exact thresholds and best actions based on results are still being defined in trials. Sensitivity depends on the assay, tumor shedding, and timing, so a negative ctDNA test does not guarantee absence of disease.

Reading a real-world report: what common results mean

Consider a typical advanced adenocarcinoma biopsy from the distal esophagus or gastroesophageal junction. You might see:

• PD-L1 CPS 12
• HER2 IHC 2+, ISH amplified
• MMR proteins intact, MSI-stable
• TMB 6 mut/Mb
• NGS: TP53 mutation, ERBB2 amplification, no fusions detected

Translation: The PD-L1 CPS suggests a meaningful chance of benefit from PD-1 based therapy. HER2 is positive by amplification, which opens HER2-directed approaches with chemotherapy in the metastatic setting. MSI-stable and modest TMB don't add immunotherapy leverage beyond PD-L1. TP53 mutations are common and not directly targetable today. If this same profile were found in a limited biopsy, repeating HER2 on another block or site could be reasonable to confirm heterogeneity, especially when the IHC is 2+.

For a squamous cell carcinoma of the mid-esophagus you might see:

• PD-L1 CPS 25
• HER2 negative
• MMR proteins intact
• NGS: No immediately actionable target; CDKN2A loss, TP53 mutation

Translation: PD-L1 supports adding PD-1 based therapy to chemotherapy in advanced disease, consistent with trial data in ESCC. Targeted therapy options based on NGS are uncommon here outside clinical trials.

Stage and setting still lead

Biomarkers fine-tune but do not override stage. Endoscopic ultrasound, CT and PET inform whether disease is localized, locally advanced, or metastatic. In resectable disease, the backbone is surgery integrated with chemotherapy and radiation, with adjuvant immunotherapy after neoadjuvant chemoradiation and resection improving disease-free survival in a large trial (CheckMate 577).¹¹ PD-L1 subgroup signals exist, yet the clinical decision in this adjuvant setting has been made broadly, not exclusively by biomarker.

Assay caveats that change results

• Cold ischemia time and fixation: Delays before formalin fixation can degrade proteins and nucleic acids.
• Decalcification: Acid decalcification of bone biopsies can ruin IHC and DNA quality.
• Tumor content: Very low tumor cellularity can yield false negatives, especially for NGS and ISH.
• Clone and platform: PD-L1 assays use different antibody clones and scoring rules. Reports typically name the assay used.
• Temporal drift: PD-L1 and HER2 can change after therapy. Fresh tissue can be more reflective of current biology than archival blocks.

Who should get what, practically

For advanced or metastatic oesophageal cancer, a practical, guideline-consistent approach includes PD-L1 CPS for all, HER2 testing for adenocarcinoma and junctional tumors, and MMR/MSI for all because even a small chance of a high-yield finding is worth the test. Broad NGS is useful when available to uncover uncommon targets and report TMB, with the understanding that most findings beyond HER2 and MSI will influence trials more than standard care. For ESCC, focus on PD-L1 and MSI; HER2 is rarely relevant.

Geography and histology matter

Prevalence of ESCC versus adenocarcinoma varies by region, as do exposures and background genetics. For example, variants in the ALDH2 gene common in East Asian populations increase acetaldehyde exposure when drinking alcohol and raise ESCC risk. That's an important public health signal, but it is not used to select therapy once cancer is present. Within adenocarcinoma, true gastric tumors have a higher chance of MSI-high and distinct targets like CLDN18.2, while distal esophageal and junctional tumors sit in a biologic gray zone. This is why pathology reports carefully define tumor location and type, not just "esophageal cancer."

Blood tests that sound helpful but usually aren't

CEA, CA 19-9, and CA 72-4 are sometimes ordered. None are recommended for screening. Levels can be normal despite advanced disease or elevated because of benign conditions. In selected cases, if a marker is clearly elevated at baseline and then falls with treatment, it can be used as a rough personal yardstick along with scans. Changes should not drive big decisions on their own because false swings are common.

How biomarkers connect to everyday choices

Think of immunotherapy like hiring a security team for your immune system. PD-L1 CPS estimates how many doors are unlocked for that team to get into the tumor. It's not perfect access control, but higher CPS generally means more open doors. HER2 is more like a loudspeaker the tumor uses to broadcast growth signals. If the loudspeaker is blaring, HER2-directed therapy can turn the volume down. MSI-high tumors carry many genetic "typos," which ironically make them easier for the immune system to spot when you remove the brakes.

What biomarkers cannot tell you

• They cannot provide reliable early detection in the general population.
• They cannot predict surgical outcomes better than careful staging.
• They cannot replace how you feel and function day to day. Energy levels, swallowing comfort, weight trends, and recovery from treatment still matter more for assessing wellbeing.

Lifestyle choices are central to overall health and treatment tolerance, but they rarely move core tumor biomarkers. For example, regular physical activity can improve insulin sensitivity and help preserve muscle during treatment by upregulating glucose transport into muscle cells independent of insulin, yet it won't change a HER2 amplification or an MSI result. Nutrition that maintains lean mass supports recovery from chemotherapy through better protein turnover and mitochondrial function, but it doesn't alter PD-L1 staining on a tumor block.

Key questions to ask when you read a report

• What is the histologic type and exact anatomic location?
• Is PD-L1 reported as CPS, and what is the number and assay used?
• For adenocarcinoma, is HER2 IHC 3+, or if 2+, is ISH amplified?
• Is the tumor MSI-high or dMMR?
• Was broad NGS performed, and are there rare fusions or amplifications that are potentially targetable or trial-relevant?
• If a past biopsy was negative for HER2 or low for PD-L1, was the sample adequate and is there another site to test?

Evidence anchors

The backbone facts in this guide come from large randomized trials and consensus guidelines. For PD-1 therapy, trials such as KEYNOTE-590 in oesophageal and gastroesophageal adenocarcinoma and CheckMate 648 in ESCC showed survival gains tied to PD-L1 expression levels.^{1 4} For post-operative settings, CheckMate 577 demonstrated improved disease-free survival with adjuvant immunotherapy after neoadjuvant chemoradiation and surgery in oesophageal and junctional cancers.¹¹ For HER2, the ToGA trial established benefit of adding HER2-targeted therapy to chemotherapy in HER2-positive metastatic adenocarcinoma of the stomach and gastroesophageal junction,^{2 6} with subsequent studies refining use in the oesophagogastric spectrum. MSI-high and dMMR as predictors of immunotherapy response are supported by tumor-agnostic approvals based on durable responses across multiple cancer types.⁷ ctDNA for minimal residual disease is a fast-moving field with encouraging observational and interventional data, though standardized thresholds and action plans are still maturing.

Putting it together

A clear biomarker plan for oesophageal cancer looks like this: define the histology and location accurately, stage the disease thoroughly, and obtain a high-quality tissue sample early for PD-L1, HER2 (if adenocarcinoma), and MSI/MMR. Add broad NGS if available to capture rare targets and TMB. Consider ctDNA when tissue is limited or when monitoring is valuable during systemic therapy. Expect heterogeneity and be open to repeat testing at key decision points, especially if the clinical picture changes. Above all, interpret results in context. Biomarkers tilt probabilities; they do not make promises. Science continues to sharpen these tools, and while the details evolve, the core strategy of personalizing therapy based on a tumor's signals has been a durable step forward.

That's the essence of biomarker-guided care in oesophageal cancer today. It's modern medicine at the intersection of pathology and therapeutics, grounded in evidence and careful interpretation rather than hype. As with any fast-moving area, specifics may change as new trials read out, but the principles here will help you read the next report with more confidence and the right questions in hand.