Zolbetuximab for CLDN18.2-Positive Gastric Cancer Therapy

Zolbetuximab is the first targeted therapy approved specifically for gastric and gastroesophageal junction cancers that express a protein called claudin 18.2 (CLDN18.2). In two large phase III trials, adding zolbetuximab to standard chemotherapy reduced the risk of disease progression and death by roughly 25% compared with chemotherapy alone. That makes it a meaningful new option in a disease where treatment advances have been hard to come by, but it comes with distinctive side effects, requires a specific tissue test to confirm eligibility, and works for only a subset of patients.

What CLDN18.2 Is and Why It Matters

Claudin 18.2 is a structural protein normally found in the lining of the stomach. In healthy tissue, it sits tucked between cells, helping form the tight junctions that hold the stomach’s mucosal barrier together. When stomach cells become cancerous, that orderly arrangement breaks down and claudin 18.2 becomes exposed on the outer surface of the tumor cell, where therapeutic antibodies can reach it. Because CLDN18.2 has very limited expression in other healthy tissues, it offers an unusually clean target for cancer therapy: you can aim at it without causing widespread collateral damage elsewhere in the body.¹

How Zolbetuximab Attacks Tumor Cells

Zolbetuximab is a monoclonal antibody, meaning it is an engineered protein designed to latch onto one specific target on a cell’s surface. Once zolbetuximab binds to CLDN18.2 on a cancer cell, it triggers two distinct immune-mediated killing pathways. The first is antibody-dependent cellular cytotoxicity, in which the body’s natural killer cells recognize the antibody-coated tumor cell and destroy it. The second is complement-dependent cytotoxicity, where a cascade of blood proteins punches holes in the cancer cell’s membrane. Together, these mechanisms recruit the patient’s own immune system to do much of the heavy lifting.²

Preclinical research in mouse models has shown that combining zolbetuximab with chemotherapy also increases the number of cancer-killing immune T cells that infiltrate the tumor. In those same models, pairing zolbetuximab with a checkpoint inhibitor (an anti-PD-1 antibody) slowed tumor growth more effectively than either agent used on its own.³ Those findings have fueled interest in combining zolbetuximab with immunotherapy in humans, a question that clinical trials are now actively exploring.

The Phase III Evidence

Zolbetuximab’s approval rests on two randomized, placebo-controlled phase III trials, SPOTLIGHT and GLOW, each pairing the antibody with a different standard chemotherapy backbone in patients with CLDN18.2-positive, HER2-negative, locally advanced or metastatic gastric and gastroesophageal junction cancer.

In SPOTLIGHT, zolbetuximab was combined with mFOLFOX6 (a regimen built around oxaliplatin, leucovorin, and fluorouracil). Patients who received zolbetuximab had a median progression-free survival of about 10.6 months compared with 8.7 months for those on placebo plus chemotherapy. The risk of progression or death dropped by 25%, and the risk of death alone also fell by 25%.⁴

GLOW used a capecitabine-and-oxaliplatin backbone (CAPOX) instead. The results tracked in the same direction: median progression-free survival was about 8.2 months with zolbetuximab versus 6.8 months without it, and overall survival improved from roughly 12.2 months to 14.4 months. The reductions in risk were consistent with SPOTLIGHT.⁵ Having two trials with two different chemotherapy partners converge on the same benefit size strengthened the case that zolbetuximab itself was the driver of the improvement.

An analysis of patient-reported outcomes from both trials found that adding zolbetuximab to chemotherapy did not erode quality of life relative to chemotherapy alone. Patients maintained their measured health-related quality of life compared with baseline despite receiving an extra drug.⁶

Who Qualifies and How Testing Works

Zolbetuximab is only used in tumors confirmed to express high levels of CLDN18.2. That determination is made through an immunohistochemistry (IHC) test on a tissue sample, typically from a biopsy or surgical specimen. The clinical cutoff used in the registration trials requires at least 75% of tumor cells to show moderate-to-strong membrane staining for CLDN18.⁷ That is a fairly high bar, and it means many patients whose tumors show some CLDN18.2 expression still won’t meet the threshold for treatment.

Getting the test right is not trivial. A concordance study found that different laboratory assays for CLDN18.2 can produce variable results, though they generally agree when using the same scoring criteria that defined positivity in the phase III trials.⁸ Standardizing the assay across pathology labs around the world remains an active area of work.

Across the more than 4,500 patients screened in SPOTLIGHT and GLOW combined, about 38% met the CLDN18.2-positivity threshold. The proportion varied by region: roughly 44% in Europe and the Middle East, about 38% in North America, and around 35% in mainland China.⁹ So while zolbetuximab meaningfully helps the patients who qualify, it does not apply to the majority of gastric cancer cases.

Where CLDN18.2 Fits Among Other Biomarkers

Gastric cancer treatment now relies on multiple biomarkers to guide therapy: HER2 status determines eligibility for trastuzumab, PD-L1 scores help select candidates for immune checkpoint inhibitors, and CLDN18.2 expression directs zolbetuximab use. A natural question is how much these categories overlap. If a patient’s tumor is already HER2-positive, for instance, are they likely to also express CLDN18.2?

In a study of gastric cancer patients with peritoneal spread, CLDN18.2 expression showed no significant correlation with either HER2 status or PD-L1 scores in the primary tumors. Among HER2-negative tumors, about 28% were CLDN18.2-positive at the strict 75% cutoff. Among tumors that were both HER2-negative and PD-L1-low, roughly 27% still expressed enough CLDN18.2 to qualify.¹⁰ That independence is clinically important: zolbetuximab fills a gap for patients whose tumors lack the markers that would make them candidates for other targeted therapies.

Nausea and Vomiting During Infusion

The most distinctive side effect of zolbetuximab is nausea and vomiting, which occurs prominently during the first infusion and then tapers significantly in later cycles. In pooled data from SPOTLIGHT and GLOW, nearly half of patients experienced nausea on the first infusion day and about 41% had vomiting. By the second infusion day, both symptoms roughly halved, and they continued declining with each subsequent dose.¹¹ About 10% of patients on zolbetuximab had nausea or vomiting severe enough to lead to inadequate dosing or early discontinuation, compared with less than 1% on placebo.

Real-world experience has confirmed this pattern. In one early clinical-practice cohort, nausea during the first cycle hit about 62% of patients, infusion interruptions were needed for a similar proportion, and the median first infusion took over four hours to complete. But all patients finished their first dose, and by the second cycle, only about a fifth experienced any nausea and none had vomiting.¹²

Managing the Side Effects

Because nausea and vomiting are so predictable with zolbetuximab and tend to concentrate in the first infusions, oncologists have developed specific protocols to keep patients on treatment. An international expert panel reached consensus on several strategies:

Premedication: Patients receive a combination of anti-nausea drugs before each infusion, typically a serotonin receptor blocker, a neurokinin-1 receptor blocker, and dexamethasone.
Infusion-rate adjustment: If nausea begins during the infusion, the rate is slowed. If vomiting occurs, the infusion is paused for 30 to 60 minutes and restarted at half the rate when symptoms improve.
Rescue medications: Additional anti-nausea drugs not already used for prevention are given during the infusion as needed.

The expert panel specifically recommended against permanently stopping zolbetuximab due to nausea and vomiting without first trying these adjustments.¹³

A Japanese retrospective study added some useful nuance. Patients who had not undergone prior gastrectomy were far more likely to develop nausea and vomiting during the first infusion (about 39%) than those who had (about 8%), and younger patients faced higher risk. Olanzapine, a drug more commonly associated with psychiatry, reduced moderate-or-worse nausea and vomiting when added to the anti-nausea regimen. The study also identified a practical threshold: symptoms tended to emerge when infusion rates reached around 200 to 250 mg per hour, suggesting that staying below about 150 mg per hour could lower the risk.¹⁴

One Japanese initiative, the Project VYLOY protocol, formalized a stepwise infusion-rate ladder. Patients start at a set rate, and if moderate-or-worse symptoms occur, the infusion pauses and restarts one level lower. Subsequent cycles begin at whatever rate the patient tolerated last time. This structured approach, along with agents like olanzapine for patients with stubborn symptoms, has helped clinics standardize care and keep more patients on treatment.¹⁵

Tumor Heterogeneity and Resistance

One persistent challenge with CLDN18.2-directed therapy is that the target itself is not always uniformly expressed. A study comparing primary stomach tumors with their metastatic deposits found that about 20% of cases showed discordance: the primary tumor and its metastasis disagreed on CLDN18.2 positivity.¹⁶ A separate study of metastatic gastric cancer patients found that about 47% had CLDN18.2-positive tumors overall, but the degree of staining varied within and between sites.¹⁷ This spatial heterogeneity means a single biopsy might not capture the full picture, and some tumor clones could escape zolbetuximab from the start.

There is also the question of what happens to the target over time. Early data suggest that CLDN18.2 expression can decrease during zolbetuximab treatment, possibly through a selection process in which the antibody kills off high-expressing cells while lower-expressing clones survive and expand. Researchers have drawn a parallel to the loss of HER2 expression that sometimes occurs during anti-HER2 therapy. However, there are hints that the loss of CLDN18.2 expression may be reversible, which could have different implications for subsequent treatment choices compared with truly permanent target loss.¹⁸ Beyond target loss, a broader review noted several other hurdles: incomplete initial sensitivity in some patients, gastrointestinal toxicity, and the lack of an established treatment sequence after progression on zolbetuximab.¹⁹

Next-Generation CLDN18.2 Therapies

Zolbetuximab is only the first entrant in what is becoming a crowded pipeline of CLDN18.2-directed treatments. Three broad categories of newer agents are in clinical development, each using a fundamentally different strategy to exploit the same target.

Antibody-Drug Conjugates

Antibody-drug conjugates (ADCs) work by linking a CLDN18.2-targeting antibody to a potent cell-killing chemical payload. The antibody delivers the toxic cargo directly to tumor cells, sparing most healthy tissue. CMG901, a first-in-class CLDN18.2 ADC carrying a microtubule-disrupting agent, showed a manageable safety profile and promising antitumor activity in patients with advanced gastric cancer in a phase I trial.²⁰ Another ADC, IBI343, uses a different payload chemistry and is also in early-phase testing.²¹ ADCs could be especially valuable if they prove effective in tumors with moderate CLDN18.2 expression that falls below the strict threshold required for zolbetuximab, because even a modest amount of target protein may be enough to deliver a lethal payload inside the cell.

CAR T-Cell Therapy

A more radical approach involves engineering a patient’s own T cells to recognize CLDN18.2. Satricabtagene autoleucel (satri-cel) is the most advanced CLDN18.2-directed CAR T-cell product. In a phase I trial involving 98 patients with CLDN18.2-positive gastrointestinal cancers, the overall response rate was about 39% and disease was controlled in over 90% of patients. The median progression-free survival was 4.4 months and overall survival reached about 8.8 months. All cytokine release syndrome events were mild (grade 1 or 2), with no neurotoxicity or treatment-related deaths reported.²²

A randomized phase II trial compared satri-cel with physician’s choice of chemotherapy in patients who had already failed at least two prior treatment lines. Progression-free survival nearly doubled, going from about 1.8 months with chemotherapy to 3.3 months with satri-cel. Grade 3 or higher adverse events occurred in nearly all CAR T recipients, but these were predominantly expected blood-count drops from the lymphodepletion conditioning, and cytokine release syndrome occurred in 95% of patients.²³

Perhaps the most striking signal has come from a small series of patients with peritoneal metastases, one of the hardest-to-treat patterns in gastric cancer. Three patients with high CLDN18.2 expression received satri-cel and achieved durable peritoneal disease control. One patient responded well enough to undergo surgical removal of residual disease eight months after infusion and remained without further progression. The other two patients survived 44 and 35 months respectively. CAR T cells were found in resected tumor tissue months after infusion, even though they had become undetectable in the blood by about day 30.²⁴

Bispecific Antibodies

Bispecific antibodies are engineered to grab two things at once: one arm binds CLDN18.2 on the tumor cell, while the other grabs CD3 on a T cell, physically dragging the immune cell into contact with its target. IBI389 is one such molecule now in early clinical testing and has shown preliminary activity in patients with gastric and gastroesophageal tumors.²⁵ AZD5863, another CLDN18.2×CD3 bispecific antibody, is in a phase I/II dose-escalation study in patients with advanced solid tumors.²⁶ Bispecifics offer a potential off-the-shelf alternative to CAR T cells, since they don’t require engineering each patient’s own cells, which makes manufacturing faster and treatment more accessible.

The Possibility of Combination With Immunotherapy

One of the open questions in the field is whether combining CLDN18.2-targeted therapy with checkpoint inhibitors could produce better outcomes than either approach alone. The biological rationale is there: CLDN18.2 appears to play a role in shaping immune dysfunction within gastric tumors, so blocking it might make the tumor microenvironment more responsive to checkpoint blockade.²⁷ Clinicians already assess both PD-L1 and CLDN18.2 when deciding on frontline treatment, and the fact that the two markers appear to be independent of each other means many patients could in principle qualify for both strategies. Trials testing these combinations are underway, and they could reshape the first-line treatment landscape if the results are positive.

Cost-Effectiveness Concerns

Even when a therapy works, its price tag matters for real-world adoption. A Japanese cost-effectiveness analysis compared several first-line regimens for HER2-negative advanced gastric cancer, including zolbetuximab plus CAPOX. On a strict cost-per-quality-adjusted-life-year basis, zolbetuximab came with a very high incremental cost over pembrolizumab-based regimens, exceeding $1.6 million per additional quality-adjusted life year gained. Standard chemotherapy regimens like SOX had the highest probability of being cost-effective at conventional willingness-to-pay thresholds.²⁸ These analyses depend heavily on local drug pricing and healthcare system structures, so the numbers will look different in other countries. But they flag a real tension: zolbetuximab offers a genuine survival benefit for a molecularly defined group, yet payers and health systems will have to weigh that benefit against considerable cost. How aggressively manufacturers negotiate pricing, and whether biomarker-based selection can be refined to identify the patients who benefit most, will shape access for years to come.