Initial systemic therapy for metastatic esophageal and gastric cancer

INTRODUCTION — 

Gastric, gastroesophageal junction (GEJ), and esophageal cancers often present as advanced unresectable or metastatic disease. These advanced unresectable or metastatic cancers are not curable, and the goals of systemic therapy include palliating symptoms, improving quality of life, and prolonging overall survival (OS). Clinical trial enrollment is encouraged, where available.

This topic will present initial systemic therapy for advanced unresectable and metastatic gastric, GEJ, and esophageal cancer. Second- and later-line systemic therapy, palliative therapies, and the epidemiology and risk factors for these cancers are discussed separately.

●(See "Second- and later-line systemic therapy for metastatic gastric and esophageal cancer".)

●(See "Local palliation for advanced gastric cancer".)

●(See "Management of locally advanced unresectable or inoperable esophageal cancer", section on 'Endoscopic interventions'.)

●(See "Epidemiology and risk factors for esophageal cancer" and "Epidemiology of gastric cancer" and "Risk factors for gastric cancer".)

PRETREATMENT EVALUATION

Diagnosis and staging — The diagnosis and staging of esophageal and gastric cancer are discussed separately. (See "Clinical presentation, diagnosis, and staging of gastric cancer" and "Clinical manifestations, diagnosis, and staging of esophageal cancer".)

Biomarker assessment — Patients with advanced unresectable or metastatic gastric and esophageal cancers who are potential candidates for systemic therapy should have their tumors assessed for mismatch repair (MMR) status, human epidermal growth factor receptor 2 (HER2) expression, programmed cell death ligand 1 (PD-L1) expression, and claudin18.2 (CLDN18.2) expression, which are used to guide initial management [1].

dMMR/MSI-H — All patients should be evaluated for mismatch repair deficiency (dMMR) and high levels of microsatellite instability (MSI-H). Diagnostic testing is discussed separately. (See "Lynch syndrome (hereditary nonpolyposis colorectal cancer): Clinical manifestations and diagnosis", section on 'Tumor MSI/IHC testing'.)

PD-L1 expression — All patients should have their tumors evaluated for programmed cell death ligand 1 (PD-L1) expression. We prefer assessing PD-L1 expression using a combined positive score (CPS). A tumor proportion score (TPS) is also an option, but TPS may not be as predictive as CPS for PD-L1 expression in upper gastrointestinal tract adenocarcinomas [2]. (See "Principles of cancer immunotherapy", section on 'Diagnostic tests'.)

PD-L1 expression is a dynamic biomarker that can change in relation to local cytokines and other factors. PD-L1 expression has been studied in clinical trials evaluating immune checkpoint inhibitors. Although PD-L1 as a biomarker has some limitations including interobserver scoring as well as spatial and temporal heterogeneity, most clinical studies evaluating immune checkpoint inhibitors in upper gastrointestinal cancer have used prespecified categorical cutoff values. [3]. In general, the level of PD-L1 expression correlates with treatment responses to immune checkpoint inhibitors, although this varies by tumor type. Further details on using PD-L1 expression to predict tumor response to immune checkpoint inhibitors are discussed separately. (See "Principles of cancer immunotherapy", section on 'Predictors of response to immune checkpoint inhibitors'.)

There are several important questions regarding the use of PD-L1 expression assays in gastric and esophageal cancer:

●Are PD-L1 expression assays interchangeable? – The 22C3 PharmDx immunohistochemistry (IHC) assay is a US Food and Drug Administration (FDA)-approved companion diagnostic assay for assessing the safety and effectiveness of pembrolizumab in a variety of malignancies, including gastric and esophageal cancer [4]. By contrast, the PD-L1 28-8 PharmDx assay is the approved companion diagnostic assay for nivolumab. Data indicate that the agreement between these two assays in gastric cancer at a CPS cutpoint of 1 and 10 was 96 percent [5], suggesting that they are interchangeable. Other studies suggest less concordance between the 22C3 IHC assay and the SP263 IHC assay, another commercially available PD-L1 IHC [6]. Further details on available diagnostic tests for PD-L1 are discussed separately. (See "Principles of cancer immunotherapy", section on 'Diagnostic tests'.)

●What is the optimal diagnostic scoring method for PD-L1 expression? – CPS and TPS are available companion diagnostic scoring methods for PD-L1 expression on formalin-fixed, paraffin-embedded tissue. (See "Principles of cancer immunotherapy", section on 'Diagnostic tests'.)

•CPS – CPS is the preferred approach for gastric and esophageal cancer. CPS is the total number of PD-L1 stained cells (including tumor cells, lymphocytes, macrophages) divided by the total number of viable tumor cells and multiplied by 100. CPS is a robust, reproducible PD-L1 scoring method that predicts a response to pembrolizumab in gastric and esophageal cancer with interpathologist and intrapathologist agreement of 97 percent, and interinstitution agreement of 92 percent [3].

•TPS – In tumors other than gastric cancer (eg, non-small cell lung cancer), the TPS (ie, the number of PD-L1-stained tumor cells divided by the total number of viable tumor cells and multiplied by 100) can identify those patients with PD-L1 expressing tumors who are likely to respond to pembrolizumab. However, immune cell PD-L1 expression is also important as this phenomenon may be specific to the unique tumor microenvironment of gastroesophageal cancer. Nevertheless, some clinical trials with metastatic gastric and esophageal cancer have used TPS.

There are limited data on the correlation between CPS and TPS. In one randomized trial (CheckMate 648) in advanced esophageal squamous cell carcinoma (SCC) that evaluated immunotherapy plus chemotherapy versus chemotherapy alone, a TPS cutpoint of ≥1 percent identified a population whose overall survival (OS) with combined therapy was comparable to the OS of those treated at a CPS cutpoint of ≥10 percent [7].

●Is there spatial heterogeneity of PD-L1 expression between tumor sites? – In multiple cancer types, including upper gastrointestinal tract tumors, PD-L1 expression can have spatial heterogeneity or vary between the primary tumor site (including between the biopsy and surgical resection samples) and metastatic sites [8]. In one observational study of patients with gastroesophageal adenocarcinoma, the concordance rate for PD-L1 expression between paired baseline primary tumors and baseline metastatic sites was 61 percent [9]. Among the primary tumors with CPS ≥1, less than half (42 percent) had corresponding metastases that were also CPS ≥1. By contrast, among primary tumors with CPS <1, the proportion with corresponding CPS <1 metastases was higher (88 percent); similar concordance results were seen for primary tumors with CPS ≥10. These findings argue against the empiric use of immune checkpoint inhibitors in gastric and esophageal cancer.

●Is there temporal heterogeneity of PD-L1 expression and is a repeat biopsy needed following systemic therapy? – Temporal heterogeneity is the variation in PD-L1 expression between tumors before and after systemic therapy (61 to 69 percent concordance) [9]. This finding can also be seen soon after initiating chemotherapy; one study showed significantly upregulated immune cell activity gene programs (including those related to checkpoint inhibition) after only a single dose of fluorouracil [FU] plus platinum chemotherapy [10].

The issue of temporal heterogeneity raises the question as to whether tumors should be rebiopsied to reassess PD-L1 expression status following systemic therapy. It is possible for tumors that are initially PD-L1-negative to become PD-L1-positive following systemic therapy. Nevertheless, repeating a biopsy if CPS is absent or low is not a standard approach as there is no high-quality evidence for improved outcomes. It is also not known whether "systemic therapy-induced" PD-L1 expression has the same biologic implication as "native" PD-L1 expression.

In an observational study evaluating temporal heterogeneity among gastroesophageal adenocarcinoma, among tumors that were CPS ≥1 at baseline, most (64 percent) remained CPS ≥1 on posttreatment biopsy [9]. However, among tumors with CPS <1, the proportion that became CPS ≥1 on posttreatment biopsy was 39 percent. For those with tumors with CPS <10 at baseline, the proportion that became CPS ≥10 on posttreatment biopsy was only 17 percent. Additional studies are needed to determine the optimal location, timing, and measurement of PD-L1 as well as other emerging biomarkers.

HER2 expression — For patients with advanced or metastatic gastric and esophageal adenocarcinoma who are eligible for trastuzumab, we assess for human epidermal growth factor receptor 2 (HER2) overexpression and/or gene amplification (table 1 and algorithm 1). This biomarker determines eligibility for HER2-targeted therapies. Although the landmark clinical trials that confirmed the efficacy of HER2-targeted therapy included only gastric and gastroesophageal junction (GEJ) adenocarcinoma (TOGA [11] and KEYNOTE-811 [12]), we extrapolate from these studies to also evaluate for HER2 expression in esophageal adenocarcinoma. (See 'HER2-positive adenocarcinomas' below.)

In gastric and esophageal adenocarcinoma, the heterogeneity of immunostaining for HER2 is greater than in other tumor types, such as breast cancer. Therefore, tumor-specific criteria for HER2 expression have been developed for these tumors. These criteria incorporate the method of tissue sampling (biopsy versus surgical specimen) and IHC staining, with fluorescence in situ hybridization (FISH) for equivocal IHC results. High levels of HER2 expression are typically confirmed by either 3+ IHC staining or equivocal (2+ IHC staining) in conjunction with positive FISH. Further details on assessing HER2 status in gastric and esophageal adenocarcinoma are discussed separately. (See "Gastric cancer: Pathology and molecular pathogenesis", section on 'HER2 expression'.)

CLDN18.2 — For patients with advanced or metastatic HER2-negative gastric and esophageal adenocarcinoma, we obtain testing for claudin18.2 (CLDN18.2), which determines eligibility for zolbetuximab, a cytolytic antibody directed at CLDN18.2. A companion diagnostic (Ventana CLDN18 [43-14A] RxDx assay) has been approved by the FDA. CLDN18.2 positivity has been studied at a threshold of moderate (2+) to strong (3+) staining in greater than or equal to 75 percent of cells. [4]. (See 'CPS of less than 5, CLDN18.2-positive' below.)

Studies are evaluating the overlap between the expression of CLDN18.2 and other actionable biomarkers associated with gastric and esophageal cancer. Most CLDN18.2-positive gastric and esophageal adenocarcinomas are HER2-negative and mismatch repair proficient (pMMR) [13]. However, the prevalence of the biomarkers HER2, PD-L1, and MMR are generally similar between CLDN18.2-positive and -negative tumors.

Early supportive care — All patients with newly diagnosed advanced or metastatic gastric, GEJ, or esophageal cancer should have a full assessment of symptom burden, nutritional and psychologic status, and social support. This assessment should occur as early as possible, ideally prior to starting systemic therapy [14]. Many patients will benefit from formal palliative care consultation and services. Early referral and initiation of interdisciplinary and palliative care services improve clinical and quality of care outcomes, including OS. (See "Benefits, services, and models of subspecialty palliative care", section on 'Rationale for palliative care'.)

Because of the anatomy, and complications from surgery or local disease progression, patients with advanced esophagogastric cancer have a high incidence of malnutrition [15,16], and psychologic distress [17,18], both of which may impair survival. Several therapeutic options are available to control symptoms of local disease progression (eg, nausea, pain, gastric outlet obstruction, bleeding), including palliative surgical resection, surgical bypass (gastrojejunostomy), radiation therapy (RT), and endoscopic techniques. When making decisions for local palliative therapy, clinicians must account for the patient's prognosis to avoid excessive morbidity, mortality, or lengthy hospital stays in those with a limited life span. (See "Local palliation for advanced gastric cancer".)

GOALS OF THERAPY — 

The goals of chemotherapy in patients with advanced or metastatic esophageal and gastric cancer are to palliate symptoms (including malignant dysphagia), improve quality of life, and prolong overall survival (OS). Randomized trials and meta-analyses provide evidence for the survival benefit of palliative systemic chemotherapy [19-26]. As an example, in one meta-analysis of three randomized trials comparing chemotherapy with best supportive care, chemotherapy improved OS compared with supportive care alone (median 11 versus 4 months, hazard ratio [HR] 0.3, 95% CI 0.24-0.55) [23].

BACKBONE CHEMOTHERAPY REGIMENS — 

Our general approach to initial systemic therapy is to select a backbone chemotherapy regimen and then decide whether to pair it with either immunotherapy or targeted therapy, based on histology and tumor biomarker status.

Selection of therapy — For patients with advanced unresectable or metastatic gastric and esophageal adenocarcinoma and esophageal squamous cell carcinoma (SCC), regardless of mismatch repair (MMR)/microsatellite stability (MSS) status, we suggest a chemotherapy backbone that includes both a fluoropyrimidine and oxaliplatin, rather than other chemotherapy combinations.

●Options include FOLFOX (table 2), CAPOX/XELOX (table 3), or modified FLOT (or TFOX; for gastric and esophageal adenocarcinoma only). FOLFIRI is an acceptable alternative for those unable to tolerate oxaliplatin (eg, peripheral neuropathy). (See 'FOLFOX and CAPOX' below and 'Modified FLOT/TFOX' below and 'FOLFIRI' below.)

●We use modified FLOT in patients with gastric or esophageal adenocarcinoma, good performance status, and minimal comorbidities who require a robust initial treatment response due to malignant obstruction, visceral crisis, or other significant symptoms. (See 'Modified FLOT/TFOX' below.)

●For patients treated initially with modified FLOT or FOLFIRI, we start with chemotherapy alone. Those on modified FLOT may be switched to the appropriate initial systemic therapy (chemotherapy plus immunotherapy and/or targeted therapy, based on biomarker status) once their disease has stabilized. (See 'Modified FLOT/TFOX' below.)

FOLFOX and CAPOX — In advanced esophageal and gastric cancer, FOLFOX and CAPOX are active initial regimens [27-40]. Randomized trials suggest that these fluoropyrimidine plus oxaliplatin doublet regimens are similarly effective and better tolerated than fluoropyrimidine plus cisplatin combination regimens [22,27,40-44]. As examples:

●A network meta-analysis of randomized clinical trials with patients with esophageal and gastric cancer compared 17 initial chemotherapy regimens for overall survival (OS; 50 trials, 10,249 patients) and progression-free survival (PFS; 34 trials, 7795 patients) [22]. Compared with a fluoropyrimidine plus cisplatin, fluoropyrimidine plus oxaliplatin doublets improved OS compared with a fluoropyrimidine plus cisplatin doublets (hazard ratio [HR] 0.83, 95% CI 0.71-0.98). Fluoropyrimidine plus cisplatin doublets were also associated with more grade 3 to 4 toxicity.

●In another meta-analysis of three randomized trials of 1294 patients with advanced unresectable gastric cancer, fluoropyrimidine plus oxaliplatin-based regimens improved PFS (HR 0.88, 95% CI 0.8-0.98) and OS (HR 0.88, 95% CI 0.78-0.99) compared with fluoropyrimidine plus cisplatin-based regimens [44]. Compared with fluoropyrimidine plus cisplatin-based regimens, fluoropyrimidine plus oxaliplatin-based regimens caused less neutropenia, anemia, alopecia, and thromboembolic events but more neurotoxicity and diarrhea.

●In a randomized phase II trial (CALGB 80403) comparing epirubicin, cisplatin, and fluorouracil (FU; ECF) with FOLFOX (table 2), both in combination with cetuximab, both regimens had similar OS but FOLFOX was better tolerated [40].

Modified FLOT/TFOX — FU, leucovorin, oxaliplatin, and docetaxel (modified FLOT or TFOX) is an option for initial chemotherapy in patients with metastatic gastric and esophageal adenocarcinoma, good performance status, and minimal comorbidities who require a robust initial treatment response due to malignant obstruction, visceral crisis, or other significant symptoms. We do not offer FLOT to patients with metastatic esophageal SCC, as it has not been evaluated in this population. We start with modified FLOT alone, and do not combine this regimen with other targeted agents or immunotherapy. Patients can be switched to the appropriate initial systemic therapy (chemotherapy plus immunotherapy and/or targeted therapy, based on tumor biomarker status) once their disease has stabilized on modified FLOT.

Based on data from phase II clinical trials [38,39], modified FLOT was subsequently evaluated in an open-label phase III trial (PRODIGE51-FFCD-GASTFOX). In this study, 507 patients with treatment-naïve human epidermal growth factor receptor 2 (HER2)-negative locally advanced unresectable or metastatic gastric or gastroesophageal junction (GEJ) adenocarcinoma were randomly assigned to initial therapy with either modified FLOT (FOLFOX plus docetaxel at 50 mg/m² every two weeks) or FOLFOX alone [45]. At a median follow-up of 43 months, compared with FOLFOX, modified FLOT improved OS (median 15.1 versus 12.7 months, HR 0.82, 95% CI 0.68-0.99), PFS (median 7.6 versus 6 months), and objective response rate (ORR; 62 versus 53 percent). Grade 3 toxicities that were higher with modified FLOT than FOLFOX were neuropathy (32 versus 20 percent), diarrhea (15 versus 7 percent), neutropenia (26 versus 8 percent), and fatigue (16 versus 8 percent). However, median time to deterioration in quality of life was longer with modified FLOT than FOLFOX (17 versus 13.7 months).

FOLFIRI — FU, leucovorin, and irinotecan (FOLFIRI) (table 4) is an alternative to oxaliplatin-based chemotherapy for the initial treatment of advanced or metastatic gastric cancer. This regimen may be appropriate for patients who are unable to receive oxaliplatin (eg, peripheral neuropathy). We start with FOLFIRI alone and do not combine this regimen with other targeted agents or immunotherapy. Data are as follows:

●A network meta-analysis of patients with esophageal and gastric cancer compared 17 initial chemotherapy regimens for OS (50 trials, 10,249 patients) and PFS (34 trials, 7795 patients) [22]. Compared with a fluoropyrimidine plus cisplatin, a fluoropyrimidine plus irinotecan doublet improved OS (HR 0.85, 95% CI 0.71-0.99). Cisplatin-fluoropyrimidine doublet therapy was also associated with more grade 3 or 4 toxicity.

●In a randomized phase II trial of 136 patients with advanced gastric cancer [46], compared with FU plus leucovorin alone or with cisplatin, FOLFIRI improved ORR (40 versus 13 and 27 percent, respectively), PFS (median 6.9 versus 3.2 and 4.9 months, respectively), and OS (median 11.3 versus 6.8 and 9.5 months, respectively).

●In a phase III trial conducted in France, 416 patients with previously untreated, advanced gastric or GEJ adenocarcinoma were randomly assigned to FOLFIRI or epirubicin, cisplatin, and capecitabine (ECX) [47]. While PFS, OS, and ORR were similar between the two treatment arms, FOLFIRI was better tolerated (grade 3 or 4 toxicity rate of 69 versus 84 percent with ECX).

Although other irinotecan-containing regimens are also active such as irinotecan plus cisplatin [48-53] and irinotecan plus capecitabine [54-56], we do not offer these regimens due to toxicity.

Other regimens — Other options for chemotherapy include FU plus cisplatin and platinum plus paclitaxel [57,58]. However, these regimens are generally less preferred to FU and oxaliplatin-based doublet for initial therapy.

Fluoropyrimidine plus cisplatin

●FU plus cisplatin – FU plus cisplatin was previously one of the most used regimens in metastatic and localized esophageal cancer due to its activity and well-established toxicity profile. In a randomized phase II trial, 88 patients with locally advanced or metastatic esophageal SCC were treated with single-agent cisplatin (100 mg/m² every 21 days) either alone or in combination with FU (1000 mg/m² per day by continuous infusion on days 1 to 5) [59]. Although the addition of FU to cisplatin increased ORR (35 versus 19 percent), OS was similar between the treatment arms (median 33 versus 28 weeks; one-year OS 34 versus 27 percent). Furthermore, the addition of FU to cisplatin increased treatment-related mortality rate by 17 percent (primarily due to sepsis and cerebrovascular episodes).

In a separate noncomparative study, split-dose cisplatin (20 mg/m² per day on days 1 to 5) plus leucovorin and bolus FU (300 mg/m² per day for five days) resulted in similar ORR (27 percent) and median OS (six months) with less toxicity (treatment-related mortality rate 3 percent) [60].

●Cisplatin plus capecitabine – Two randomized trials evaluated the substitution of capecitabine for infusional 5FU (including REAL-2) [41,61]. In a meta-analysis of these two trials, compared with FU combinations, capecitabine combinations improved ORR (odds ratio 1.38, 95% CI 1.1-1.73) and improved OS (HR for death 0.87, 95% CI 0.77-0.98) [62].

Regimens not used — We do not offer the triplet regimens ECF or docetaxel, cisplatin, and FU (DCF) in patients receiving chemotherapy alone. These regimens are more toxic than fluoropyrimidine plus platinum doublets [22], and the contributions of certain agents (eg, anthracyclines for ECF, cisplatin for DCF) are questionable.

●Epirubicin, cisplatin, and fluorouracil – Based on initial data from a phase II trial [63], a subsequent phase III trial compared ECF (table 5) with FU, doxorubicin, and methotrexate (FAMTX) in 274 patients with advanced esophagogastric adenocarcinoma or undifferentiated cancer [64]. Compared with FAMTX, ECF improved ORR (45 versus 21 percent) and OS (median 8.9 versus 5.7 months). ECF caused more alopecia and nausea, while FAMTX caused more hematologic toxicity and infections.

In a noninferiority phase III trial (REAL-2) of 1002 patients with gastric cancer, ECF was compared against three other chemotherapy regimens, ECX (table 6), EOF, and EOX (table 7) [41]. This study demonstrated that capecitabine could be substituted for infusional FU in the ECF regimen without compromising efficacy. Similar results have been seen in other studies [65].

However, the contribution of the anthracycline in ECF is questionable. An individual patient data meta-analysis of the Global Advanced/Adjuvant Stomach Tumor Research International Collaboration group concluded that there was no role for epirubicin in combination with a fluoropyrimidine and a platinum agent [66].

The benefits of ECF over more contemporary regimens such as FOLFOX and CAPOX are also questionable, as studies suggest fluoropyrimidine and oxaliplatin doublet regimens are as effective and better tolerated than fluoropyrimidines and cisplatin alone [27,67-69]. (See 'FOLFOX and CAPOX' above.)

In a randomized phase II trial (CALGB 80403) in which cetuximab was added to ECF, irinotecan plus cisplatin, or FOLFOX, ECF, and FOLFOX demonstrated similar OS [40]. However, this trial included cetuximab in both arms, and it was not designed to compare FOLFOX and ECF.

●Docetaxel, cisplatin, and fluorouracil – We do not offer docetaxel, cisplatin, and fluorouracil (DCF or TCF) (table 8) as initial therapy in advanced or metastatic esophageal or gastric cancer. Although most studies evaluating DCF [70-84], but not all [85], demonstrate clinical efficacy, and it has regulatory approval in the metastatic setting, it is more toxic than platinum-containing doublet regimens. In addition, the cisplatin in DCF may not be necessary as studies have suggested similar outcomes when using docetaxel and FU alone without cisplatin [73].

In a phase III trial (TAX-325), 457 patients with chemotherapy-naïve advanced gastric cancer were randomly assigned to either DCF or cisplatin plus FU [70]. The addition of docetaxel to cisplatin plus FU improved two-year OS (18 versus 9 percent) and ORR (37 versus 25 percent) but caused more grade 3 or 4 diarrhea (20 versus 8 percent) and neutropenia (30 versus 14 percent).

A modified schedule of DCF is equally effective and better tolerated (table 9) [86]. In a randomized trial of 85 patients with previously untreated, metastatic gastric adenocarcinoma, modified DCF (without prophylactic growth factor support) improved OS (median 19 versus 13 months) and was less toxic than standard DCF (with growth factor support) [87].

Docetaxel, in combination with cisplatin and FU, is approved by the US Food and Drug Administration (FDA) and in Europe for the treatment of advanced gastric cancer.

●FOLFIRINOX – We do not use FU, leucovorin, oxaliplatin, and irinotecan (FOLFIRINOX) for advanced esophageal cancer. In a phase II trial, this regimen demonstrated high ORR but was significantly toxic [88].

MISMATCH REPAIR DEFICIENT/MSI-H TUMORS

Treatment approach — For patients with mismatch repair deficiency (dMMR)/microsatellite instability-high (MSI-H) metastatic esophageal and gastric cancer (either squamous cell carcinoma [SCC] or adenocarcinomas), we recommend the combination of chemotherapy to immunotherapy rather than chemotherapy alone, (algorithm 2 and algorithm 3) as this approach improves overall survival (OS) and can induce durable treatment responses. Immunotherapy alone (nivolumab plus ipilimumab, pembrolizumab, or dostarlimab) is an acceptable alternative for patients who are unlikely to tolerate chemotherapy. Although data are lacking to guide the selection of therapy, we suggest FOLFOX plus nivolumab, followed by maintenance therapy for those patients without disease progression after three to four months of treatment. Options for maintenance therapy include nivolumab plus fluorouracil (FU), nivolumab monotherapy, or pembrolizumab monotherapy. In our clinical experience, this approach results in durable treatment responses and is well-tolerated.

For previously untreated patients with advanced or metastatic dMMR/MSI-H gastric and esophageal adenocarcinoma, studies suggest that the addition of an immune checkpoint inhibitor to chemotherapy confers a substantial OS benefit over chemotherapy alone. This approach is also extrapolated to those with MSI-H esophageal SCC, although such tumors are exceedingly rare and most dMMR/MSI-H gastric and esophageal tumors are adenocarcinoma on histology. Sensitivity to immune checkpoint inhibitors is generally seen with all dMMR/MSI-H cancer histologies, and the biologic rationale for this is discussed separately. (See "Overview of advanced unresectable and metastatic solid tumors with DNA mismatch repair deficiency or high tumor mutational burden", section on 'Biologic principles'.)

Nivolumab plus chemotherapy — In a phase III trial (CheckMate 649), the addition of nivolumab to chemotherapy was evaluated in previously untreated advanced or metastatic gastric or gastroesophageal junction (GEJ) adenocarcinoma [89-91]. In a subset analysis of 44 patients with dMMR/MSI-H tumors, the addition of nivolumab to chemotherapy improved OS (median 38 versus 12 months, hazard ratio [HR] 0.34, 95% CI 0.16-0.74) [91]. An OS benefit was also seen among the subset of 34 patients with both combined positive score (CPS) ≥5 and dMMR/MSI-H tumors (median 45 versus 9 months, HR 0.29, 95% CI 0.12-0.71). The addition of nivolumab to chemotherapy increased grade ≥3 toxicity (60 versus 45 percent). Full results of CheckMate 649 are discussed separately. (See 'CPS of 10 or more' below.)

Pembrolizumab plus chemotherapy — Studies in advanced dMMR/MSI-H gastric and GEJ adenocarcinoma suggest that the addition of pembrolizumab to chemotherapy improves OS [92,93]. As an example, in a phase III trial (KEYNOTE-859), the addition of pembrolizumab to chemotherapy was evaluated in previously untreated advanced or metastatic gastric or GEJ adenocarcinoma [92]. In a subset analysis of 39 patients with MSI-H tumors, the addition of pembrolizumab to chemotherapy improved OS (HR 0.34, 95% CI 0.18-0.66) and progression-free survival (PFS; HR 0.27, 95% CI 0.14-0.53). The addition of pembrolizumab to chemotherapy increased grade ≥3 toxicity (60 versus 51 percent). Full results of KEYNOTE-859 are discussed separately. (See 'CPS of 10 or more' below.)

An exploratory analysis of three separate trials evaluating pembrolizumab as initial (KEYNOTE-062) and later-line (KEYNOTE-059 and KEYNOTE-060) therapy for advanced or metastatic gastric or GEJ adenocarcinoma included a subgroup of 50 patients with tumors with dMMR/MSI-H and CPS ≥1. In this subgroup, relative to chemotherapy alone, pembrolizumab plus chemotherapy was associated with improved OS (median not reached versus 8.5 months), improved PFS (median not reached versus 6.6 months), and a higher objective response rate (ORR; 65 versus 37 percent) [93]. Full results from this study are discussed separately. (See 'CPS of 10 or more' below.)

Immunotherapy alone — Immunotherapy alone is also an acceptable alternative, with equally good evidence for OS benefit and tolerability when compared with chemotherapy in randomized trials. Immunotherapy may be preferred for patients who are poor candidates for chemotherapy (ie, due to older age, complex comorbidities, limited performance status and/or other clinical factors). Options include nivolumab plus ipilimumab, pembrolizumab, or dostarlimab.

Nivolumab plus ipilimumab — Nivolumab plus ipilimumab is an option for the initial treatment of patients with metastatic MSI-H gastric and esophageal cancer which improved OS in a phase III trial [90].

The standard dosing is nivolumab 1 mg/kg plus ipilimumab 3 mg/kg every three weeks for four cycles, followed by nivolumab 240 mg every two weeks for a maximum of up to two years or until disease progression or unacceptable toxicity during this period [90]. However, we do not use this dosing, which is associated with significant toxicity. Instead, we offer an alternative dosing, which is better tolerated (nivolumab 240 mg every two weeks plus ipilimumab 1 mg/kg every six weeks for 16 weeks, followed by nivolumab 240 mg intravenous [IV] every two weeks or 480 mg every four weeks up to a maximum of two years), extrapolating from data in the neoadjuvant setting [94,95].

In a phase III trial (CheckMate 649), nivolumab plus ipilimumab was directly compared to investigator's choice of chemotherapy alone (CAPOX or FOLFOX) in 813 patients with previously untreated advanced or metastatic gastric or GEJ adenocarcinoma [90]. At a minimum follow-up of 36 months, among the 22 patients with MSI-H tumors, nivolumab plus ipilimumab improved OS (median OS not reached versus 10 months, unstratified HR 0.28, 95% CI 0.08-0.92) with a higher ORR (70 versus 57 percent) relative to chemotherapy.

However, in the entire study population, relative to chemotherapy, nivolumab plus ipilimumab demonstrated similar OS (median OS 12 months for both; HR 0.91, 95% CI 0.77 to 1.07) and worsened PFS (median PFS three versus seven months, HR 1.66, 95% CI 1.40-1.95) [90]. Grade 3 to 4 toxicity was lower with nivolumab plus ipilimumab than chemotherapy (38 versus 46 percent).

Pembrolizumab — An exploratory analysis of three separate trials evaluating pembrolizumab as initial (KEYNOTE-062) and later-line (KEYNOTE-059 and KEYNOTE-060) therapy for advanced or metastatic gastric or GEJ adenocarcinoma included a subgroup of 84 patients with dMMR/MSI-H tumors [93]. In this subgroup, relative to chemotherapy alone, pembrolizumab monotherapy was associated with improved OS (median not reached versus nine months, one-year OS 79 versus 47 percent), improved PFS (median 11 versus 7 months), higher ORR (57 versus 37 percent), and longer duration of response (21 versus 7 months). Pembrolizumab had less grade 3 to 4 toxicity than chemotherapy (17 versus 69 percent). Full results from this study are discussed separately. (See 'CPS of 10 or more' below.)

Dostarlimab — The use of dostarlimab in advanced or metastatic dMMR/MSI-H tumors is discussed separately. (See "Overview of advanced unresectable and metastatic solid tumors with DNA mismatch repair deficiency or high tumor mutational burden", section on 'Dostarlimab'.)

MISMATCH REPAIR PROFICIENT GASTRIC AND ESOPHAGEAL ADENOCARCINOMA

Selection of chemotherapy backbone — For patients with mismatch repair proficient (pMMR)/microsatellite stable (MSS) advanced unresectable or metastatic gastric and esophageal adenocarcinoma, we suggest a chemotherapy backbone that includes both a fluoropyrimidine and oxaliplatin, rather than other chemotherapy combinations. Options include FOLFOX (table 2) or CAPOX/XELOX (table 3). Alternative initial chemotherapy regimens are discussed separately. (See 'Backbone chemotherapy regimens' above.)

Additional therapy is guided by the presence or absence of biomarkers including tumor human epidermal growth factor receptor 2 (HER2) status, claudin18.2 (CLDN18.2), and programmed cell death ligand 1 (PD-L1) expression (algorithm 2).

HER2-positive adenocarcinomas

Treatment approach — For patients with pMMR/MSS human epidermal growth factor receptor 2 (HER2)-positive gastric and esophageal adenocarcinomas (algorithm 1), we recommend the addition of trastuzumab to chemotherapy. Preferred options include FOLFOX plus trastuzumab (table 10) and CAPOX plus trastuzumab (table 11). (See 'Trastuzumab plus chemotherapy' below.)

Additionally, for HER2-positive cancers that also have combined positive score (CPS) ≥1, we recommend the addition of pembrolizumab to trastuzumab and chemotherapy. Preferred options include pembrolizumab plus FOLFOX and trastuzumab (table 12) and pembrolizumab plus CAPOX and trastuzumab (table 13). (See 'Pembrolizumab plus trastuzumab and chemotherapy' below.)

For patients who are ineligible for trastuzumab (table 14), we offer an initial treatment approach similar to those with HER2-negative adenocarcinoma. (See 'Ineligible for trastuzumab' below.)

Trastuzumab plus chemotherapy — The combination of trastuzumab and various chemotherapy regimens has been investigated in advanced gastric and esophageal adenocarcinoma. Data are as follows:

●Trastuzumab plus fluoropyrimidine and cisplatin – In an open-label phase III trial (ToGA), the addition of trastuzumab to fluoropyrimidine and cisplatin-based chemotherapy improved overall survival (OS) and progression-free survival (PFS) and was well-tolerated [11]. In this study, 594 patients with advanced HER2-positive gastric or gastroesophageal junction (GEJ) adenocarcinoma were randomly assigned to six cycles of platinum-based chemotherapy (capecitabine plus cisplatin or infusional fluorouracil [FU] plus cisplatin) with or without trastuzumab (8 mg/kg loading dose, followed by 6 mg/kg every three weeks until disease progression or unacceptable toxicity).

At a median follow-up of 17 to 19 months, the addition of trastuzumab to chemotherapy improved OS (median 14 versus 11 months, hazard ratio [HR] 0.74, 95% CI 0.6-0.91), PFS (median 6.7 versus 5.5 months, HR 0.71, 95% CI 0.59-0.86), and objective response rates (ORR; 47 versus 35 percent). In an exploratory subgroup analysis, trastuzumab improved OS in the patients with HER2 immunohistochemistry (IHC) 3+ tumors (HR 0.66, 95% CI 0.5-0.87) but was less effective in those with IHC 2+ tumors (HR 0.78, 95% CI 0.55-1.1) and ineffective in those with HER2 gene-amplified (ie, fluorescent in situ hybridization [FISH]-positive) but nonprotein-expressing (IHC 0 or 1+) tumors.

Grade 3 and 4 toxicities were similar between the two treatment arms (68 percent each), but the addition of trastuzumab to chemotherapy had higher rates of grade 3 or 4 diarrhea (9 versus 4 percent) and an asymptomatic decrease in left ventricular ejection fraction (5 versus 1 percent).

Based on the results of the ToGA trial, trastuzumab, in combination with cisplatin and a fluoropyrimidine, is approved by the US Food Drug and Administration (FDA) for the treatment of patients with metastatic HER2-positive gastric or GEJ adenocarcinomas who have not received prior treatment for metastatic disease [96].

●Trastuzumab plus fluoropyrimidine and oxaliplatin – When selecting between chemotherapy regimens, we suggest adding trastuzumab to fluoropyrimidine plus oxaliplatin-based chemotherapy rather than fluoropyrimidine plus cisplatin-based chemotherapy as data suggest better OS and less toxicity with this approach.

In a meta-analysis of 155 prospective and retrospective cohort studies that included 557 patients with advanced HER2-positive gastric and GEJ adenocarcinoma, trastuzumab was evaluated in combination with various chemotherapy regimens for initial therapy [97]. Compared with trastuzumab, cisplatin, and a fluoropyrimidine (the regimen used in the ToGA trial [11]), trastuzumab plus oxaliplatin and either capecitabine or FU was associated with improved OS (median 20.7 versus 16 months, HR 0.75, 95% CI 0.59-0.99) and less toxicity.

●Trastuzumab plus other chemotherapy agents – Alternatively, trastuzumab may be combined with other regimens used for initial therapy in advanced gastric and esophageal adenocarcinoma (eg, FOLFIRI; paclitaxel with or without a platinum; FU; and docetaxel, cisplatin, and fluorouracil [DCF]). However, this is a less preferred approach as there are no randomized trials evaluating the addition of trastuzumab to these regimens.

In a meta-analysis of 15 prospective and retrospective cohort studies that included 557 patients with advanced HER2-positive gastric and GEJ adenocarcinoma, trastuzumab was evaluated in combination with various chemotherapy regimens for initial therapy [97]. Compared with trastuzumab, cisplatin, and a fluoropyrimidine (the regimen used in the ToGA trial), the addition of trastuzumab to various chemotherapy regimens resulted in the following:

•Trastuzumab with a triplet chemotherapy (eg, docetaxel, a platinum, and a fluoropyrimidine) or with bevacizumab plus a doublet chemotherapy – Similar OS and more toxicity.

•Trastuzumab plus cisplatin and S-1 – Similar OS and with a different toxicity profile, including less hand-foot syndrome.

•Trastuzumab plus cisplatin or capecitabine – Worsened OS and more toxicity.

Pembrolizumab plus trastuzumab and chemotherapy — Based on data from phase I and II clinical trials [98,99], the combination of pembrolizumab, trastuzumab, and chemotherapy was evaluated in a double-blind, placebo-controlled phase III trial (KEYNOTE-811) [12,100,101]. In this study, 698 patients with HER2-positive advanced gastric or GEJ adenocarcinoma and no prior systemic therapy for advanced disease were randomly assigned to either pembrolizumab or placebo, in combination with trastuzumab and platinum plus fluoropyrimidine-based chemotherapy (investigator's choice of cisplatin plus FU or CAPOX) (table 12 and table 13) every three weeks for up to 35 cycles or until disease progression or unacceptable toxicity during this period. More patients had tumors with CPS ≥1 than CPS <1 (85 versus 15 percent).

Relative to the addition of placebo, the addition of pembrolizumab to trastuzumab plus chemotherapy demonstrated the following:

●Entire study population – Improved PFS (median 10 versus 8 months, HR 0.73, 95% CI 0.61-0.87) and ORR (73 versus 60 percent) at a median follow-up of 38 months [12], and improved OS (median 20 versus 17 months, HR 0.80, 95% CI 0.67-0.94) at a median follow-up of 50 months [101].

●CPS ≥1 – Improved PFS (median 11 versus 7 months, HR 0.71, 95% CI 0.59-0.86) and ORR (73 versus 58 percent) at a median follow-up of 38 months [12], and improved OS (median 20 versus 16 months, HR 0.79, 95% CI 0.66-0.95) at a median follow-up of 50 months [101]. Results for those with CPS <1 are discussed separately. (See 'Regimens not used' below.)

Grade ≥3 toxicity rates were similar between the two treatment arms (58 versus 51 percent), and no new toxicity profiles were identified [100].

Pembrolizumab in combination with trastuzumab, fluoropyrimidine, and platinum-containing chemotherapy has accelerated approval from the FDA for the first-line treatment of adult patients with locally advanced unresectable or metastatic HER2-positive gastric or GEJ adenocarcinoma whose tumors express PD-L1 (CPS ≥1) using an FDA-approved test [96]. (See "Principles of cancer immunotherapy", section on 'Diagnostic tests'.)

Ineligible for trastuzumab — For patients who are ineligible for trastuzumab, we offer an initial treatment approach similar to those with HER2-negative adenocarcinoma. (See 'HER2-negative adenocarcinoma' below.)

Cardiac criteria that render a patient ineligible for adjuvant trastuzumab in breast cancer are available (table 14), and we extrapolate its use to advanced gastric and esophageal cancer to determine which patients are not candidates for trastuzumab.

We also do not offer initial therapy that integrates an alternative HER2-directed agent, such as lapatinib [102], pertuzumab [103], or trastuzumab emtansine (T-DM1) [104] due to lack of OS benefit for these agents in phase III trials.

Regimens not used — For patients with HER2-positive gastric and esophageal adenocarcinoma and CPS <1, we do not offer pembrolizumab in combination with trastuzumab and chemotherapy as data suggest no PFS benefit and a possible trend towards worsened OS [12].

A phase III trial (KEYNOTE-811) of treatment-naïve advanced HER2-positive gastric and GEJ adenocarcinoma included a subgroup of patients with CPS <1 [12]. Among this subgroup, the addition of pembrolizumab demonstrated similar PFS (10 months each, HR 1.03, 95% CI 0.65-1.64) and similar ORRs (69 percent each). At a median follow-up of 28 months, OS was also numerically lower, but the difference did not meet statistical significance (16 versus 22 months, HR 1.61, 95% CI 0.98-2.64). Follow-up of OS is ongoing. Results for the entire study population in KEYNOTE-811 are discussed separately. (See 'Pembrolizumab plus trastuzumab and chemotherapy' above.)

For patients with HER2-positive gastric and esophageal adenocarcinoma, we also do not offer trastuzumab plus FOLFIRINOX. In a phase II trial, this combination demonstrated higher ORRs but was significantly toxic [88].

HER2-negative adenocarcinoma

Treatment approach — For patients with pMMR/MSS, human epidermal growth factor receptor 2 (HER2)-negative gastric and esophageal adenocarcinoma, selection of therapy is based on tumor expression of PD-L1 and CLDN18.2 (algorithm 2). In advanced gastric and esophageal adenocarcinoma, studies suggest there is a gradient of OS benefit according to PD-L1 expression (ie, CPS), which is the most clinically meaningful in those with the highest CPS [105]. (See 'PD-L1 expression' above.)

●CPS ≥5 – For patients with CPS ≥5, regardless of CDLN18.2 expression, we recommend the addition of immunotherapy to chemotherapy, as this approach improves OS.

•For those with CPS ≥10, we offer either nivolumab or pembrolizumab plus chemotherapy. (See 'CPS of 10 or more' below.)

•For those with CPS ≥5 and <10, we suggest nivolumab plus chemotherapy rather than pembrolizumab plus chemotherapy. (See 'CPS of 5 to less than 10' below.)

Although there are no comparative randomized trials, for patients with CPS ≥5, we prefer chemotherapy plus immunotherapy over chemotherapy plus zolbetuximab regardless of CLDN18.2 expression status. This is due to the more durable OS and treatment responses seen with chemotherapy plus immunotherapy. Additionally, zolbetuximab is associated with symptomatic toxicities such as nausea and vomiting during infusion.

●CPS <5 – For patients with CPS <5 and CLDN18.2-positive disease, we recommend the addition of zolbetuximab to chemotherapy.

For those with CPS <5 and CLDN18.2-negative disease, we do not typically incorporate immunotherapy, given the OS benefits are limited in this population and may not outweigh the risks of therapy (eg, immune-related adverse events, financial costs). We generally treat such patients with chemotherapy alone. However, we acknowledge that there may be variability in clinical practice and that providers outside of UpToDate may prefer to offer chemotherapy plus immunotherapy to patients with CPS in this range. In this situation, we emphasize a discussion about the risks and benefits of this approach and encourage shared decision-making with patients on a case-by-case basis. (See 'CPS of less than 5, CLDN18.2-negative' below.)

●Ineligible for immunotherapy or zolbetuximab – Patients who are ineligible for immunotherapy or zolbetuximab are treated with chemotherapy alone. (See 'Backbone chemotherapy regimens' above.)

A meta-analysis conducted by the FDA Oncology Drugs Advisory Committee(ODAC) combined data from three trials that evaluated the addition of immunotherapy to chemotherapy in advanced gastric and GEJ adenocarcinoma (CheckMate 649 [nivolumab plus chemotherapy] [89-91], KEYNOTE-859 [pembrolizumab plus chemotherapy] [92], and RATIONALE-305 [tislelizumab plus chemotherapy [106]) [105]. Among the patients with pMMR/MSS tumors, adding immunotherapy to chemotherapy conferred a gradient of OS benefit according to PD-L1 expression that was the most clinically evident in patients with CPS ≥10 (median OS 16.5 versus 11.5 months, HR 0.64, 95% CI 0.56-0.73). There was also OS benefit for those with CPS ≥5 (median OS 14.6 versus 12.1 months, HR 0.73, 95% CI 0.66-0.82). By contrast, among those with lower CPS scores, the addition of immunotherapy to chemotherapy demonstrated limited OS benefit for those with CPS <5 (median OS 12.6 versus 11.9 months, HR 0.89, 95% CI 0.8-1) and no benefit for those with CPS <1 (median OS 13.4 versus 12.4 months, HR 0.91, 95% CI 0.75-1.09).

CPS of 10 or more — For patients with HER2-negative adenocarcinoma and CPS ≥10, options include either nivolumab, pembrolizumab, or tislelizumab plus chemotherapy as this approach improved OS in randomized trials such as CheckMate 649 [89-91], KEYNOTE-859 [92], and RATIONALE-305 [106].

Nivolumab plus fluorouracil and oxaliplatin — Nivolumab plus oxaliplatin-based chemotherapy is an option for patients with HER2-negative metastatic gastric and esophageal adenocarcinoma and CPS ≥5. The recommended nivolumab doses in combination with chemotherapy are:

●360 mg every three weeks in combination with every three-week chemotherapy (eg, capecitabine plus oxaliplatin)

●240 mg every two weeks in combination with every two-week chemotherapy (eg, FOLFOX)

In an open-label phase III trial (CheckMate 649), 1581 patients with previously untreated HER2-negative, advanced/unresectable or metastatic gastric, GEJ, or esophageal adenocarcinoma were randomly assigned to either nivolumab (360 mg every three weeks or 240 mg every two weeks) plus chemotherapy or chemotherapy alone [89-91]. Chemotherapy regimens were oxaliplatin plus either leucovorin plus short-term infusional FU (FOLFOX) or capecitabine (XELOX; also known as CAPOX). Although patients were enrolled regardless of PD-L1 expression, the CPS was ≥5 in 955 patients (60 percent). At a median follow-up of approximately 36 months, compared with chemotherapy alone, nivolumab plus chemotherapy demonstrated the following results [91]:

●Entire study population – Improved OS (median 13.7 versus 11.6 months; three-year OS 17 versus 10 percent, HR 0.79, 95% CI 0.71-0.88), improved PFS (median 7.7 versus 6.9 months; three-year PFS 11 versus 7 percent, HR 0.79, 95% CI 0.71-0.89), and higher ORR (58 versus 46 percent).

●CPS ≥5 – Improved OS (median 14.4 versus 11.1 months; three-year OS 21 versus 10 percent, HR 0.7, 95% CI 0.61-0.81), improved PFS (median 8.3 versus 6.1 months; three-year PFS 13 versus 8 percent, HR 0.7, 95% CI 0.6-0.81), and higher ORR (60 versus 45 percent).

●Other CPS subgroups – Similar OS for CPS <10 (median 12.4 versus 12.5 months, HR 0.91, 95% CI 0.79-1.06). Outcomes for other CPS subgroups are discussed separately. (See 'CPS of less than 5, CLDN18.2-negative' below.)

●Toxicity – More frequent grade ≥3 treatment-related adverse events (60 versus 45 percent), treatment discontinuation for toxicity (19 versus 10 percent) [91], and treatment-related deaths (17 versus 9 patients, 2 versus <1 percent) [89]. However, patients receiving nivolumab plus chemotherapy reported stable or improved health-related quality of life while on therapy and were at reduced risk of deterioration in health-related quality of life [107].

Based on the results of CheckMate 649, nivolumab was approved by the FDA, in combination with a fluoropyrimidine- and platinum-containing chemotherapy, for advanced or metastatic gastric and GEJ cancer and esophageal adenocarcinoma irrespective of PD-L1 expression (ie, CPS) [96]. However, in our view, the benefits of immunotherapy for gastroesophageal adenocarcinoma with CPS <5 are not established. Notably, the European Medicines Agency has restricted approval of nivolumab to those with CPS ≥5 [108]. Further details are discussed separately. (See 'CPS of less than 5, CLDN18.2-negative' below.)

Pembrolizumab plus fluorouracil and platinum — Pembrolizumab plus oxaliplatin-based chemotherapy is another initial treatment option for patients with HER2-negative metastatic gastric and esophageal adenocarcinoma and CPS ≥10 (algorithm 2).

In a double-blind, placebo-controlled phase III trial (KEYNOTE-859), the addition of pembrolizumab to chemotherapy improved OS and PFS [92]. In this study, 1579 patients with previously untreated HER2-negative, locally advanced, or metastatic gastric (79 percent) or GEJ (21 percent) adenocarcinoma were randomly assigned to the addition of either pembrolizumab (200 mg intravenously every three weeks for up to 35 cycles) or placebo to investigator's choice of fluoropyrimidine- and platinum-based chemotherapy (either FU plus cisplatin or CAPOX). At a median follow-up of 31 months, relative to placebo plus chemotherapy, pembrolizumab plus chemotherapy demonstrated the following results:

●Entire study population – Improved OS (median 12.9 versus 11.5 months, HR 0.78, 95% CI 0.7-0.87) and PFS (median 6.9 versus 5.6 months, HR 0.76, 95% CI 0.67-0.85) and higher ORR (51 versus 42 percent).

●CPS ≥1 – Improved OS (median 13 versus 11.4 months, HR 0.74, 95% CI 0.65-0.84) and PFS (median 6.9 versus 5.6 months, HR 0.72, 95% 0.63-0.82) and higher ORR (52 versus 43 percent).

●CPS ≥10 – Improved OS (median 15.7 versus 11.8 months, HR 0.65, 95% CI 0.53-0.79) and PFS (median 8.1 versus 5.6 months, HR 0.62, 95% CI 0.51-0.76), and higher ORR (61 versus 43 percent).

●Other CPS subgroups – Outcomes for other CPS subgroups (CPS ≥5 and <10; CPS <5) are discussed separately. (See 'CPS of 5 to less than 10' below and 'CPS of less than 5, CLDN18.2-negative' below.)

●Toxicity – Grade ≥3 toxicity rates for pembrolizumab plus chemotherapy versus placebo plus chemotherapy were 60 and 51 percent, respectively. Patients receiving pembrolizumab plus chemotherapy were more likely to have long-term immune-mediated endocrine toxicities, such as hypothyroidism. (See "Overview of toxicities associated with immune checkpoint inhibitors", section on 'Autoimmune thyroid disease'.)

By contrast, in a separate phase III trial (KEYNOTE-062) of patients with HER2-negative gastric or GEJ adenocarcinoma, the addition of pembrolizumab to fluoropyrimidine and cisplatin-based chemotherapy failed to improve OS or PFS for those with CPS ≥1, OS for those with CPS between 1 and 9 [109], or OS for those with CPS ≥10 [110]. Unlike CheckMate 649 and KEYNOTE-859, patients in KEYNOTE-062 did not receive oxaliplatin-based chemotherapy. However, pembrolizumab plus chemotherapy and pembrolizumab monotherapy did benefit the subset of patients with both positive PD-L1 expression and dMMR/MSI-H. These data are discussed separately. (See 'Is there a role for pembrolizumab monotherapy?' below and 'Mismatch repair deficient/MSI-H tumors' above.)

Based on the results of KEYNOTE-859, pembrolizumab, in combination with fluoropyrimidine- and platinum-based chemotherapy is approved by the FDA for the first-line treatment of adult patients with locally advanced unresectable or metastatic HER2-negative gastric or GEJ adenocarcinoma, regardless of PD-L1 expression or CPS [96]. However, for pMMR gastric and GEJ adenocarcinoma, we restrict the use of pembrolizumab plus chemotherapy to patients with tumor CPS ≥10. This patient population derives an OS benefit from this combination whereas the absolute OS benefit is limited for those with pMMR adenocarcinoma and CPS <10. (See 'CPS of 5 to less than 10' below and 'CPS of less than 5, CLDN18.2-negative' below.)

Tislelizumab plus fluorouracil and platinum — Tislelizumab plus oxaliplatin-based chemotherapy is another initial treatment option for patients with HER2-negative metastatic gastric and esophageal adenocarcinoma and CPS ≥10.

In a double-blind, placebo-controlled phase III trial (RATIONALE-305), the addition of tislelizumab to chemotherapy improved OS and PFS [106]. In this study, 997 patients with previously untreated HER2-negative, advanced unresectable or metastatic gastric (80 percent) or GEJ (20 percent) adenocarcinoma were randomly assigned to the addition of either tislelizumab (200 mg intravenously every three weeks) or placebo to investigator's choice of fluoropyrimidine- and platinum-based chemotherapy (either FU plus cisplatin or CAPOX for up to six cycles; maintenance capecitabine was allowed for the latter group) until disease progression or unacceptable toxicity. After two years of treatment, tislelizumab or placebo could be stopped if a complete response, partial response, or stable disease was achieved. At a median follow-up of 25 months, relative to placebo plus chemotherapy, tislelizumab plus chemotherapy demonstrated the following results:

●Entire study population – Improved OS (median 15 versus 13 months; two-year OS 33 versus 23 percent; HR 0.80, 95% CI 0.70-0.92) and PFS (median 6.9 versus 6.2 months; two-year OS 18 versus 9 percent, HR 0.78, 95% CI 0.67-0.90).

●Tumor area positivity (TAP) ≥5 percent – Improved OS (median 17 versus 13 months; two-year OS 38 versus 25 percent; HR 0.74, 95% CI 0.59-0.94) and PFS (median 7.2 versus 5.9 months; two-year OS 22 versus 9 percent; HR 0.67, 95% CI 0.55-0.93).

●CPS ≥5 – In an exploratory subgroup analysis, improved OS (median 17.8 versus 13.2 months, unstratified HR 0.73, 95% CI 0.60-0.89).

●Other CPS subgroups – Outcomes for other PD-L1 expression subgroups (CPS <5; TAP <5 percent) are discussed separately. (See 'CPS of less than 5, CLDN18.2-negative' below.)

●Toxicity – Grade ≥3 toxicity rates for tislelizumab plus chemotherapy versus placebo plus chemotherapy were similar (54 versus 50 percent).

Based on these data, tislelizumab, in combination with platinum- and fluoropyrimidine-based chemotherapy, is approved by the FDA for the first-line treatment of adults with unresectable or metastatic HER2-negative gastric or GEJ adenocarcinoma whose tumors express PD-L1 (≥1). However, in our view, the benefits of adding immunotherapy to chemotherapy are limited for those with gastroesophageal adenocarcinoma and CPS <5 [96]. (See 'CPS of less than 5, CLDN18.2-positive' below.)

Is there a role for pembrolizumab monotherapy? — We do not typically offer initial therapy with single-agent pembrolizumab outside of a clinical trial, including to patients whose tumors have CPS ≥10. In an exploratory analysis of a randomized trial (KEYNOTE-062), pembrolizumab alone prolonged OS relative to chemotherapy alone in patients whose tumors had a CPS ≥10, but this difference was not statistically tested. In addition, these and other data also suggest that patients who receive immunotherapy alone as initial treatment for metastatic disease (either as monotherapy or combination therapy [ie, nivolumab plus ipilimumab]) are at risk for worse PFS (ie, early disease progression or death) [90,110].

In a phase III trial (KEYNOTE-062), 763 patients with previously untreated, advanced, gastric or GEJ adenocarcinoma with a CPS ≥1 (281 with a CPS ≥10) were randomly assigned to pembrolizumab alone, chemotherapy alone (cisplatin plus a fluoropyrimidine), or pembrolizumab plus chemotherapy [110]. At a median follow-up of approximately 29 months, data were as follows:

●CPS ≥1 – For patients with CPS ≥1, pembrolizumab monotherapy compared with chemotherapy demonstrated noninferior, but not superior, OS (median 10.6 versus 11.1 months, HR 0.91, 99% CI 0.69-1.18), worsened PFS (median 2 versus 6.4 months, HR 1.66, 95% CI 1.37-2.01), and lower ORR (15 versus 37 percent). Pembrolizumab had fewer any-grade (54 versus 92 percent) and grade 3 or 4 (17 versus 69 percent) adverse effects than chemotherapy.

●CPS ≥10 – In an exploratory analysis of patients with CPS ≥10, compared with chemotherapy alone, pembrolizumab monotherapy prolonged OS, although this was not statistically tested (median 17.4 versus 10.8 months, HR 0.69, 95% CI 0.49-0.97) and demonstrated a nonstatistically significant trend towards lower PFS (median 2.9 versus 6.1 months, HR 1.1, 95% CI 0.79-1.51).

CPS of 5 to less than 10 — For patients with HER2-negative gastric and esophageal adenocarcinoma and CPS of 5 to less than 10, we suggest either nivolumab or tislelizumab plus chemotherapy rather than pembrolizumab plus chemotherapy.

Nivolumab plus chemotherapy — In a phase III trial (CheckMate 649) of patients with metastatic HER2-negative gastric and GEJ adenocarcinoma, for the subgroup with CPS ≥5, nivolumab plus chemotherapy improved OS versus chemotherapy alone [89-91]. (See 'CPS of 10 or more' above.)

Tislelizumab plus chemotherapy — In a phase III trial (RATIONALE-305) of patients with metastatic HER2-negative gastric and GEJ adenocarcinoma, for the subgroup with CPS ≥5, tislelizumab plus chemotherapy improved OS versus chemotherapy alone [106]. (See 'CPS of 10 or more' above.)

What is the role of pembrolizumab plus chemotherapy? — We do not offer the combination of pembrolizumab plus chemotherapy to patients with a CPS of 5 to less than 10. Although this combination is approved by the FDA for the initial treatment of HER2-negative gastric or GEJ adenocarcinoma regardless of PD-L1 expression (ie, CPS), data suggest limited absolute OS benefit for this combination in patients with CPS of 5 to less than 10. In our view, this limited OS benefit is outweighed by the increased toxicity and cost of this combination.

A phase III trial (KEYNOTE-859) demonstrated an OS benefit for adding pembrolizumab to fluoropyrimidine plus platinum-based chemotherapy among those with CPS ≥1, but it is unlikely that this OS benefit is meaningfully driven by those with a CPS of 5 to less than 10. In an exploratory subgroup analysis of KEYNOTE-859, among patients with CPS of 5 to less than 10, the addition of pembrolizumab to chemotherapy demonstrated a statistically significant improvement in OS (HR 0.83, 95% CI 0.7-0.98) [92], but the absolute OS benefit is likely marginal (on the order of weeks at best). Further results of KEYNOTE-859 for the entire study population and other CPS subgroups are discussed separately. (See 'CPS of 10 or more' above and 'CPS of less than 5, CLDN18.2-negative' below.)

CPS of less than 5, CLDN18.2-positive

Zolbetuximab plus chemotherapy — For patients with metastatic HER2-negative gastric and esophageal adenocarcinoma that is combined positive score (CPS) <5 and claudin18.2 (CLDN18.2)-positive, we suggest the addition of zolbetuximab to chemotherapy, as this approach improved OS with an acceptable toxicity profile in randomized trials.

Zolbetuximab is a monoclonal antibody that targets CLDN18.2, a tight junction protein exclusively expressed in normal gastric/GEJ mucosal cells that are differentially exposed by gastric and GEJ cancers after malignant transformation [111,112]. Zolbetuximab is administered in conjunction with 5FU and platinum-based chemotherapy at an initial loading dose of 800 mg/m² intravenous (IV) followed by a maintenance dose of either 600 mg/m² IV every three weeks or 400 mg/m² IV every two weeks until disease progression or unacceptable toxicity during this period.

Zolbetuximab is highly emetogenic due to the gastric/GEJ mucosal cell specificity of CLDN18.2. Patients can be safely and adequately managed with dose interruptions, slowing of the infusion rate, and an aggressive prophylactic and supportive antiemetic regimen. Nausea and vomiting appear to occur most commonly during the first and second cycles and the rate of subsequent severe (grade ≥3) nausea and vomiting are reduced to less than 2 percent in later cycles [112,113]. (See "Prevention of chemotherapy-induced nausea and vomiting in adults" and "Management of poorly controlled or breakthrough chemotherapy-induced nausea and vomiting in adults".)

Data supporting the addition of zolbetuximab to chemotherapy are as follows:

●Zolbetuximab plus FOLFOX – In a placebo-controlled, double-blind phase III trial (SPOTLIGHT), 565 patients with previously untreated locally advanced or unresectable, HER2-negative, CLDN18.2-positive gastric or GEJ adenocarcinoma were randomly assigned to FOLFOX plus either zolbetuximab or placebo [113]. The addition of zolbetuximab to FOLFOX improved PFS at a median follow-up of 13 months (median 10.6 versus 8.7 months, HR 0.75, 95% CI 0.60-0.94) and improved OS at a median follow-up of 22 months (median 18.2 versus 15.5 months; three-year OS 21 versus 9 percent; HR 0.75, 95% CI 0.60-0.94). Grade ≥3 toxicity rates for the zolbetuximab versus placebo groups were 87 and 78 percent, respectively. The most common grade ≥3 toxicities in the zolbetuximab group were neutropenia (28 percent), nausea and vomiting (16 percent each), and anemia (9 percent).

●Zolbetuximab plus CAPOX – In a placebo-controlled, double-blind phase III trial (GLOW), 507 patients with previously untreated locally advanced or unresectable, HER2-negative, CLDN18.2-positive gastric or GEJ adenocarcinoma were randomly assigned to CAPOX plus either zolbetuximab or placebo [112]. At a median follow-up of 18 months, the addition of zolbetuximab to CAPOX improved PFS (median 8.2 versus 6.8 months, HR 0.67, 95% CI 0.54-0.86) and OS (median 14.4 versus 12.2 months; two-year OS 29 versus 17 percent; HR 0.77, 95% CI 0.62-0.97). Grade ≥3 toxicity rates for the zolbetuximab versus placebo groups were 73 and 70 percent, respectively. The most common grade ≥3 toxicities in the zolbetuximab group were vomiting (12 percent), anemia (11 percent), decreased neutrophil count (10 percent), and nausea (8 percent).

Combined analyses of both the SPOTLIGHT and GLOW trials also confirmed that the addition of zolbetuximab to chemotherapy conferred an OS advantage at final analysis (HR 0.77) [114] and maintained health-related quality of life relative to baseline [115].

Based on these data, zolbetuximab, in combination with fluoropyrimidine- and platinum-containing chemotherapy was approved by the FDA for the first-line treatment of adults with locally advanced unresectable or metastatic HER2-negative gastric or GEJ adenocarcinoma whose tumors are CLDN18.2-positive, as determined by an FDA-approved test [96]. This agent is also approved in Japan, the United Kingdom, the European Union, and South Korea.

CPS of less than 5, CLDN18.2-negative — For patients with pMMR/MSS HER2-negative adenocarcinoma that is combined positive score (CPS) <5 and claudin18.2 (CLDN18.2)-negative, we typically do not incorporate immunotherapy, given that the OS benefits are limited in this population and may not outweigh the risks of therapy (ie, immune-related adverse events). Such patients are treated with chemotherapy alone (see 'Backbone chemotherapy regimens' above). However, we acknowledge that there may be variability in clinical practice and that providers outside of UpToDate may prefer to offer chemotherapy plus immunotherapy to patients with CPS between 1 and 4 [116]. In this situation, we emphasize a discussion about the risks and benefits of this approach and encourage shared decision-making with patients on a case-by-case basis.

Although clinical practice is variable and specific CPS cutpoints that define low PD-L1 expression vary between trials, data from randomized trials converge to generally demonstrate limited OS benefit for the addition of immunotherapy to chemotherapy in those with CPS <5. Data are as follows:

●Nivolumab plus fluorouracil and oxaliplatin – A phase III trial (CheckMate 649) evaluated the addition of nivolumab to FU and oxaliplatin-based chemotherapy in patients with previously untreated HER2-negative, advanced/unresectable or metastatic gastric, GEJ, or esophageal adenocarcinoma [89-91]. In an exploratory subgroup analysis, the addition of nivolumab to chemotherapy failed to improve OS in those with CPS <1 (median 13.1 versus 12.5 months, unstratified HR 0.95, 95% CI 0.74-1.24), CPS <5 (median 12.4 versus 12.3 months, unstratified HR 0.95, 95% CI 0.80-1.12), or CPS <10 (median 12.4 versus 12.5 months, HR 0.91, 95% CI 0.79-1.06) [90,91]. Interaction analysis of OS by PD-L1 CPS cutoffs was not provided. Further results of CheckMate 649 are discussed separately. (See 'CPS of 10 or more' above.)

Another study reconstructed unreported Kaplan-Meier plots of CPS subgroups of various phase III trials evaluating the addition of immunotherapy with chemotherapy in the initial treatment of gastroesophageal adenocarcinoma. For those patients enrolled in CheckMate 649 and CPS of 1 to 4, there were no statistically significant differences in OS (HR 0.95) and PFS (HR 0.96) for immunotherapy plus chemotherapy compared with chemotherapy alone [109].

In another phase III trial (ATTRACTION-4) of patients from Asia with HER2-negative advanced or recurrent gastric or GEJ adenocarcinoma, the addition of nivolumab to chemotherapy (oxaliplatin plus either S-1 or capecitabine) improved PFS but failed to improve OS, regardless of tumor proportion score (TPS) score [117]. Of note, TPS may not be as predictive as CPS for PD-L1 expression in upper gastrointestinal tract adenocarcinomas [2]. (See 'PD-L1 expression' above.)

●Pembrolizumab plus fluorouracil and platinum – A phase III trial (KEYNOTE-859) evaluated the addition of pembrolizumab to FU and platinum-based chemotherapy in patients with previously untreated HER2-negative, locally advanced or metastatic gastric or GEJ adenocarcinoma [92]. In an exploratory subgroup analysis, the addition of pembrolizumab to chemotherapy failed to improve OS in those with CPS <1 (HR 0.92, 95% CI 0.73-1.17) and demonstrated limited OS benefit for those with CPS <10 (HR 0.86, 95% CI 0.75-0.98) [92]. Full results of this study are discussed separately. (See 'CPS of 10 or more' above.)

Similarly, for another phase III trial (KEYNOTE-062), in an exploratory subgroup analysis of 318 patients with previously untreated HER2-negative gastric or GEJ adenocarcinoma, the addition of pembrolizumab to fluoropyrimidine- and cisplatin-based chemotherapy failed to improve OS for those with CPS between 1 and 9 (median OS 12.5 versus 11 months, HR 0.84, 0.66-1.06) [109]. Of note, these patients did not receive oxaliplatin-based therapy.

Another phase III trial (KEYNOTE-590) evaluated pembrolizumab plus fluoropyrimidine- and cisplatin-based chemotherapy in patients with adenocarcinoma or squamous cell carcinoma (SCC) of the esophagus/GEJ [109,118,119]. In an exploratory subgroup analysis of 100 patients with gastric/GEJ adenocarcinoma and CPS <10, the addition of pembrolizumab to chemotherapy demonstrated a numerically higher OS, but the analysis was underpowered to detect a statistically significant difference (median OS 12.7 versus 8.4 months, HR 0.66, 95% 0.42-1.04) [109,118]. In extended follow-up, preliminary results from this trial suggest no significant benefit for the addition of pembrolizumab to chemotherapy for subgroups with CPS <10, regardless of histology (HR 0.84, 95% CI 0.67-1.06), but a test for interaction was not provided [119]. Of note, these patients did not receive oxaliplatin-based therapy. Full results of the KEYNOTE-590 trial are discussed separately. (See 'Pembrolizumab plus chemotherapy' below.)

●Tislelizumab plus fluorouracil plus platinum chemotherapy – A phase III trial (RATIONALE-305) evaluated the addition of tislelizumab to FU and oxaliplatin-based chemotherapy in patients with previously untreated HER2-negative, advanced/unresectable or metastatic gastric or GEJ adenocarcinoma [106]. In an exploratory subgroup analysis, the addition of tislelizumab to chemotherapy failed to improve OS in those with CPS <5 (median 13.6 versus 13 months, unstratified HR 0.89, 95% CI 0.72-1.09) and TAP <5 percent (median 14.1 versus 129 months, unstratified HR 0.91, 95% CI 0.74-1.12). Interaction analysis of OS by PD-L1 CPS or TAP cutoffs was not provided. Further results of RATIONALE-305 are discussed separately. (See 'CPS of 10 or more' above.)

●Meta-analysis – A meta-analysis conducted by the FDA ODAC combined data from three trials that evaluated the addition of immunotherapy to chemotherapy in advanced gastric and GEJ adenocarcinoma (CheckMate 649 [nivolumab plus chemotherapy] [89-91], KEYNOTE-859 [pembrolizumab plus chemotherapy] [92], and RATIONALE-305 [tislelizumab plus chemotherapy [106]) [105]. Among the patients with pMMR/MSS tumors, adding immunotherapy to chemotherapy demonstrated limited OS benefit among those with CPS <5 (median OS 12.6 versus 11.9 months, HR 0.89, 95% CI 0.8-1) and no benefit for those with CPS <1 (median OS 13.4 versus 12.4 months, HR 0.91, 95% CI 0.75-1.09).

However, we acknowledge that there may be variability in clinical practice and that providers outside of UpToDate may offer immunotherapy plus chemotherapy to patients with CPS between 1 and 4. In this situation, we emphasize a discussion about the risks and benefits of this approach and encourage shared decision-making with patients on a case-by-case basis. One meta-analysis extracted individual patient data from phase III clinical trials (KEYNOTE-859, CheckMate 649, and RATIONALE-305) that evaluated the addition of various immunotherapy agents to chemotherapy in patients with systemic-therapy naïve advanced gastroesophageal adenocarcinoma [116]. Among those with CPS between 1 and 4, the addition of immunotherapy to chemotherapy conferred a modest OS benefit (HR 0.87, 95% CI 0.78-0.98), although the median OS was the same for both treatment arms (12 months for each). However, a minority of patients appeared to have durable benefit after 18 months (eg, two-year OS 24 versus 18 percent; three-year OS 13 versus 7 percent) corresponding to a number needed to treat 17 patients; in other words, 17 patients whose tumors express PD-L1 CPS 1 to 4 must be treated with immunotherapy plus chemotherapy for one patient to be alive two or three years later, who otherwise would not be with chemotherapy alone.

Other regimens — In patients with previously untreated advanced gastric and esophageal adenocarcinoma, other immunotherapy agents that have demonstrated an OS benefit when combined with chemotherapy (fluoropyrimidine plus oxaliplatin) in randomized trials include sintilimab (ORIENT-16) [120] and sugemalimab (GEMSTONE-303) [121].

Targeted agents not used — Comprehensive molecular profiling of tumor samples using next-generation sequencing is frequently used to select targeted therapy for many cancers. (See "Next-generation DNA sequencing (NGS): Principles and clinical applications".)

However, in gastric and esophageal adenocarcinoma, various targeted agents, either alone or in combination with chemotherapy, are not used as part of standard initial therapy due to the lack of OS benefit in randomized trials. Examples of targeted agents that are not part of standard initial therapy include epidermal growth factor receptor (EGFR) inhibitors such as gefitinib [122] (although a clinical benefit was identified in a subset of EGFR-amplified tumors [123]), cetuximab [124], and panitumumab [125]; mesenchymal-epithelial transition (MET) inhibitors such as rilotumumab [126] and onartuzumab [127], everolimus [128,129] (a mechanistic target of rapamycin [mTOR] inhibitor), and antiangiogenic agents such as bevacizumab [130,131] and ramucirumab [132]. The efficacy of ramucirumab as second- and later-line therapy for advanced gastric and esophageal cancer is discussed separately. (See "Second- and later-line systemic therapy for metastatic gastric and esophageal cancer", section on 'Treatments targeting VEGF'.)

Investigational agents — Some targeted agents remain under investigation, such as crizotinib for MET-amplified tumors [133], amivantamab [134] (a bispecific EGFR/MET antibody), and cadonilimab (a programmed cell death-1 [PD-1]/cytotoxic T-lymphocyte antigen-4 [CTLA-4] bispecific antibody) [135]. However, further data are necessary prior to incorporating their use in standard treatment. Patients should be encouraged to enroll in clinical trials, where available.

MISMATCH REPAIR PROFICIENT ESOPHAGEAL SQUAMOUS CELL CARCINOMA

Selection of chemotherapy backbone — For patients with mismatch repair proficient (pMMR)/microsatellite stable (MSS) advanced unresectable or metastatic esophageal squamous cell carcinoma (SCC), we suggest a chemotherapy backbone that includes both a fluoropyrimidine and oxaliplatin, rather than other chemotherapy combinations. Options include FOLFOX (table 2) or CAPOX/XELOX (table 3). Further details are discussed separately. (See 'Backbone chemotherapy regimens' above.)

Treatment approach — For patients with pMMR/MSS advanced or metastatic esophageal SCC, selection of therapy is based on tumor programmed cell death ligand 1 (PD-L1) expression (algorithm 3). PD-L1 expression is more common with esophageal SCC than gastric and esophageal adenocarcinoma [136]. Data also suggest there is a gradient of overall survival (OS) benefit according to PD-L1 expression (ie, combined positive score [CPS]), which is the most clinically meaningful for those with higher CPS.

●CPS ≥1 – For those with CPS ≥1, we recommend the addition of immunotherapy to chemotherapy. Although nivolumab plus ipilimumab is an alternative option, there is a risk of early disease progression and death with immunotherapy alone [7].

For patients treated with chemotherapy plus immunotherapy, regimen selection is as follows:

•For patients with CPS ≥10, we offer either nivolumab, pembrolizumab, or tislelizumab plus chemotherapy. (See 'Nivolumab plus chemotherapy' below and 'Pembrolizumab plus chemotherapy' below.)

•For those with CPS of 1 to less than 10, we suggest either nivolumab or tislelizumab plus chemotherapy rather than pembrolizumab plus chemotherapy. (See 'Nivolumab plus chemotherapy' below.)

●CPS <1 – For patients with CPS <1, we do not incorporate immunotherapy as the OS benefit is limited in this population. We treat such patients with chemotherapy alone. (See 'Backbone chemotherapy regimens' above.)

●Ineligible for immunotherapy – Patients who are ineligible for immunotherapy receive initial treatment with chemotherapy alone. (See 'Backbone chemotherapy regimens' above.)

A meta-analysis conducted by the US Food and Drug Administration (FDA) Oncology Drugs Advisory Committee (ODAC) combined data from three trials that evaluated the addition of immunotherapy to chemotherapy in advanced esophageal SCC (CheckMate 648 [nivolumab plus chemotherapy] [7], KEYNOTE-590 [pembrolizumab plus chemotherapy] [118], and RATIONALE-306 [tislelizumab plus chemotherapy] [137]) [138]. Among the patients with esophageal SCC, adding immunotherapy to chemotherapy conferred an OS benefit among patients with CPS ≥1 (median OS 14.6 versus 9.8 months, hazard ratio [HR] 0.68, 95% CI 0.6-0.77) but not among those with CPS <1 (median OS 11.6 versus 12.8 months, HR 1.1, 95% CI 0.76-1.58).

CPS of 1 or more

Nivolumab plus chemotherapy — In patients with pMMR/MSS advanced esophageal SCC and CPS ≥1, nivolumab plus chemotherapy is one preferred option for initial therapy. In a randomized trial (CheckMate 648), the addition of nivolumab to chemotherapy improved OS [7].

In an open-label phase III trial (CheckMate 648), 970 patients with previously untreated, advanced unresectable, recurrent, or metastatic esophageal SCC regardless of PD-L1 expression (49 percent with TPS ≥1 percent) were randomly assigned 1:1:1 to treatment with either nivolumab plus chemotherapy (fluorouracil [FU] plus cisplatin); nivolumab plus ipilimumab; or chemotherapy (FU plus cisplatin) alone [7].

At minimum follow-up of 13 months, the addition of nivolumab to chemotherapy resulted in the following, for the overall population and in prespecified subgroups:

●All patients – Among 645 patients, improved OS (median OS 13.2 versus 10.7 months, HR 0.74, 99.1% CI 0.58-0.96), similar progression-free survival (PFS; median PFS 5.8 versus 5.6 months, HR 0.81, 98.5% CI 0.64-1.04) and improved objective response rates (ORRs; 47 versus 27 percent).

●TPS ≥1 percent – Among the 315 patients with tumor proportion score (TPS) of 1 percent or more, improved OS (median OS 15.4 versus 9.1 months, HR 0.54, 95% CI 0.37-0.8), improved PFS (median PFS 6.9 versus 4.4 months, HR 0.65, 98.5% CI 0.46-0.92), and higher ORRs (53 versus 20 percent).

●CPS ≥1 – Improved OS among the 558 patients with CPS ≥1 (median OS 13.8 versus 9.8 months, HR 0.69, 95% CI 0.56-0.84) as well as the 280 patients with CPS ≥10 (median OS 16.1 versus 11.6 months, HR 0.63, 95% CI 0.47-0.84).

●Other CPS and TPS subgroups – The outcomes for patients with CPS <1 and TPS ≤1 percent are discussed separately. (See 'CPS less than 1' below.)

●Toxicity – Toxicity was higher with nivolumab plus chemotherapy (47 percent, with 34 percent discontinuing treatment) than chemotherapy alone (36 percent, respectively, with fewer than 20 percent discontinuing treatment). This finding is possibly because of the longer duration of treatment with nivolumab plus chemotherapy (5.7 versus 3.4 months).

Outcomes for nivolumab plus ipilimumab in CheckMate 648 are discussed separately. (See 'Nivolumab plus ipilimumab' below.)

Based on these data, nivolumab, in combination with platinum plus fluoropyrimidine-based chemotherapy, was approved by the FDA for the first-line treatment of patients with advanced or metastatic esophageal SCC, regardless of PD-L1 expression (ie, CPS) [96]. However, in our view, the addition of nivolumab to chemotherapy has limited benefit in advanced or metastatic SCC with low PD-L1 expression (CPS <1). Notably, the European Medicines Agency has restricted to approval of nivolumab to advanced esophageal SCCs with TPS ≥1 percent [139]. Further details are discussed separately. (See 'CPS less than 1' below.)

Pembrolizumab plus chemotherapy — Pembrolizumab plus chemotherapy is a preferred option for initial therapy in patients with pMMR/MSS advanced esophageal SCC and CPS ≥10, which improved OS in a randomized trial.

In a double-blind, placebo-controlled phase III trial (KEYNOTE-590), 749 patients with previously untreated advanced unresectable or metastatic esophageal SCC (73 percent) and esophageal adenocarcinoma (27 percent, which included gastroesophageal junction [GEJ] Siewert type 1 adenocarcinoma (figure 1)) regardless of PD-L1 expression were randomly assigned to either pembrolizumab (200 mg intravenously [IV] every three weeks for up to 35 cycles) or placebo, plus chemotherapy (FU 800 mg/m² IV days 1 through 5 every three weeks for up to 35 cycles), and cisplatin (80 mg/m² IV every three weeks for up to 6 cycles) [118]. At a median follow-up of 23 months, relative to placebo plus chemotherapy, pembrolizumab plus chemotherapy resulted in the following, for the overall population and in prespecified subgroups:

●All patients – Improved OS (median OS 12.4 versus 9.8 months, HR 0.73, 95% CI 0.62-0.86), which was driven more by those with esophageal SCC (median OS 12.6 versus 9.8 months, HR 0.72, 95% CI 0.60-0.88) than those with adenocarcinoma (median 11.6 versus 9.9 months, HR 0.74, 95% CI 0.54-1.02); improved PFS (median PFS 6.3 versus 5.8 months, HR 0.65, 95% CI 0.55-0.76), including those with esophageal SCC (median PFS 6.3 versus 5.8 months, HR 0.65, 95% CI 0.54-0.78) and those with adenocarcinoma (median PFS 6.3 versus 4.7 months, HR 0.63, 95% CI 0.46-0.87); and higher ORRs (45 versus 29 percent).

●CPS ≥10 – Improved OS (median OS 13.5 versus 9.4 months, HR 0.62, 95% CI 0.49-0.78), which was driven more by those with SCC (median OS 13.9 versus 8.8 months, HR 0.57, 95% CI 0.43-0.75) than those with adenocarcinomas (median OS 12.1 versus 10.7 months, HR 0.83, 95% CI 0.52-1.34); improved PFS (median 7.5 versus 5.5 months, HR 0.51, 95% CI 0.41-0.65), including for those with SCC (HR 0.53) and adenocarcinoma (HR 0.49).

●CPS <10 – Similar OS (median 10.5 versus 10.6 months, HR 0.86, 95% CI 0.68-1.1), including for those with SCC (median 10.5 versus 11.1 months, HR 0.99, 95% CI 0.74-1.32) and those with adenocarcinoma (median 12.7 versus 8.4 months, HR 0.66, 95% CI 0.42-1.04); similar PFS (median 6.2 versus 6 months, HR 0.80, 95% CI 0.45-1.01), including for those with SCC (HR 0.83) and adenocarcinoma (HR 0.76).

Grade ≥3 toxicity rates for pembrolizumab plus chemotherapy versus placebo plus chemotherapy alone were 86 and 83 percent, respectively. Pembrolizumab plus chemotherapy also maintained health-related quality of life [140].

Based on the initial data from KEYNOTE-590, pembrolizumab, in combination with platinum- and fluoropyrimidine-based chemotherapy, is approved by the FDA for the treatment of patients with metastatic or locally advanced esophageal or GEJ carcinoma (SCC and adenocarcinoma) with an epicenter 1 to 5 cm above the GEJ and who are not eligible for resection or chemoradiation, regardless of PD-L1 expression [96]. However, in our view, the addition of pembrolizumab to chemotherapy has limited benefit in advanced or unresectable SCC with CPS <10. Notably, the European Medicines Agency has restricted approval of pembrolizumab to esophageal cancers with CPS ≥10 [141].

Tislelizumab plus chemotherapy — Tislelizumab plus chemotherapy is another preferred option for initial therapy in patients with pMMR/MSS advanced esophageal SCC and CPS ≥1, which improved OS in a randomized trial.

In a double-blind, placebo-controlled phase III trial (RATIONALE-306), 649 patients with advanced unresectable or metastatic esophageal SCC were randomly assigned to either tislelizumab (200 mg IV every three weeks) or placebo, plus chemotherapy (investigator's choice of platinum [cisplatin or oxaliplatin] in combination with either a fluoropyrimidine [FU or capecitabine] or paclitaxel) [137,142]. The platinum could be stopped after six cycles, and the fluoropyrimidine or paclitaxel could be continued, if appropriate. Tislelizumab or placebo could be stopped after two years if the patient demonstrated a complete response, partial response, or stable disease to therapy. Relative to placebo plus chemotherapy, tislelizumab plus chemotherapy resulting in the following, for the overall population and in a post-hoc analysis of specific subgroups:

●Entire study population – At a median follow-up of 16 months, improved OS (median OS 17 versus 11 months, HR 0.66, 95% CI 0.54-0.80), improved PFS (median 7.3 versus 5.6 months, HR 0.62, 95% CI 0.52-0.75), and higher ORRs (63 versus 42 percent) [137].

●CPS ≥1 – At a median follow-up of 43 months, among the 480 patients with CPS ≥1, improved OS (median 17 versus 10 months, HR 0.66, 95% CI 0.54-0.81). At a median follow-up of 23 months, improved PFS (median 7 versus 6 months, HR 0.58, 95% CI 0.46-0.72) and higher ORRs (58 versus 36 percent) [142].

●Tumor area positivity (TAP) ≥1 percent – At a median follow-up of 43 months, among the 481 patients with TAP ≥1 percent, improved OS (median OS 17 versus 10 months, HR 0.66, 95% CI 0.54-0.81). At a median follow-up of 23 months, improved PFS (median 7 versus 6 months, HR 0.56, 95% CI 0.45-0.70) and higher ORRs (58 versus 36 percent) [142].

●Other CPS and TAP subgroups – The outcomes for patients with CPS <1 and TAP ≤1 percent are discussed separately. (See 'CPS less than 1' below.)

●Toxicity – Grade ≥3 toxicity rates for tislelizumab plus chemotherapy versus placebo plus chemotherapy were similar (67 versus 65 percent) [142].

Based on the data from RATIONALE-306, tislelizumab, in combination with platinum-containing chemotherapy, is approved by the FDA for first-line treatment of adults with unresectable or metastatic esophageal SCC whose tumors express PD-L1 (≥1) [96].

CPS less than 1 — For patients with advanced or metastatic esophageal SCC and CPS <1, we do not incorporate immunotherapy as the OS benefit is limited in this population. We treat such patients with chemotherapy alone. (See 'Backbone chemotherapy regimens' above.)

Most studies [7,118,143], but not all [144], suggest limited OS benefit for the addition of immunotherapy to chemotherapy in advanced or metastatic SCC and low PD-L1 expression (CPS <1). Data are as follows:

●Nivolumab plus chemotherapy – In an open-label phase III trial (CheckMate 648), 970 patients with previously untreated, advanced unresectable, recurrent, or metastatic esophageal SCC were randomly assigned 1:1:1 to treatment with either nivolumab plus chemotherapy (FU plus cisplatin); nivolumab plus ipilimumab; or chemotherapy (FU plus cisplatin) alone [7]. At minimum follow-up of 13 months, the addition of nivolumab to chemotherapy demonstrated the following among those with low PD-L1 expression:

•TPS <1 percent – Among the 329 patients with TPS <1 percent, similar OS (median 12 versus 12.2 months, HR 0.98, 95% CI 0.76-1.28).

•CPS <1 – Among the 51 patients with CPS <1, median OS of 9.9 and 12.1 months respectively, but the difference was not statistically significant (HR 0.98, 95% CI 0.50-1.95).

•Further results from CheckMate 648 are discussed separately. (See 'Nivolumab plus chemotherapy' above and 'Nivolumab plus ipilimumab' below.)

●Pembrolizumab plus chemotherapy – In a placebo-controlled phase III trial (KEYNOTE-590), the addition of pembrolizumab to chemotherapy demonstrated similar OS among patients with esophageal SCC and CPS <10 [118]. (See 'Pembrolizumab plus chemotherapy' above.)

●Tislelizumab plus chemotherapy – In a placebo-controlled phase III trial (RATIONALE-306), the addition of tislelizumab to chemotherapy failed to confer an OS benefit for patients with CPS <1 (HR 1.52, 95% CI 0.81-2.84) and those with TAP <1 percent (HR 1.34, 95% CI 0.73-2.46) [96]. Further results from RATIONALE-306 are discussed separately. (See 'Tislelizumab plus chemotherapy' above.)

●Meta-analysis – A meta-analysis conducted by the FDA ODAC combined data from three trials that evaluated the addition of immunotherapy to chemotherapy in advanced esophageal SCC (CheckMate 648 [nivolumab plus chemotherapy] [7], KEYNOTE-590 [pembrolizumab plus chemotherapy] [118], and RATIONALE-306 [tislelizumab plus chemotherapy] [137]) [138]. Among the patients with esophageal SCC with CPS <1, adding immunotherapy to chemotherapy fails to confer an OS benefit (median OS 11.6 versus 12.8 months, HR 1.1, 95% CI 0.76-1.58).

Other regimens

Nivolumab plus ipilimumab — Nivolumab plus ipilimumab is an alternative option for initial therapy in patients with advanced or metastatic esophageal SCC with CPS ≥1. However, patients should be cautioned that combination immunotherapy alone is associated with a risk of early disease progression and death [7]. In a randomized trial (CheckMate 648), nivolumab plus ipilimumab improved OS over chemotherapy in the entire study population, but this OS benefit was mainly driven by those with higher PD-L1 expression.

We do not offer nivolumab plus ipilimumab to patients with SCC and low PD-L1 expression (CPS <1) due to concerns about early progression and death compared with chemotherapy alone [7,90,110,145].

In an open-label phase III trial (CheckMate 648), 970 patients with previously untreated, advanced unresectable, recurrent, or metastatic esophageal SCC regardless of PD-L1 expression (49 percent with PD-L1 ≥1 percent) were randomly assigned 1:1:1 to treatment with either nivolumab plus chemotherapy (FU plus cisplatin), nivolumab plus ipilimumab, or chemotherapy (FU plus cisplatin) alone [7].

●Nivolumab plus ipilimumab versus chemotherapy – At minimum follow-up of 13 months, compared with chemotherapy, nivolumab plus ipilimumab demonstrated the following:

•All patients – Improved OS (median OS 12.7 versus 10.7 months, HR 0.78, 98.2% CI 0.62-0.98), lower PFS (2.9 versus 5.6 months, HR 1.26, 95% CI 1.04-1.52), and ORR (28 versus 27 percent).

•TPS ≥1 percent – Among the 315 patients with TPS ≥1 percent, improved OS (median OS 13.7 versus 9.1 months, HR 0.64, 98.6% CI 0.46-0.90), similar PFS (median 4.0 versus 4.4 months, HR 1.02, 98.5% CI 0.73-1.43), and higher ORR (35 versus 20 percent).

•CPS ≥1 – Among the 546 patients with CPS ≥1, improved OS (median OS 12.7 versus 9.8 months, HR 0.76, 95% CI 0.62-0.93).

•TPS <1 percent – Among the 330 patients with TPS <1 percent, similar OS (median 12 versus 12.2 months, HR 0.96, 95% CI 0.74-1.25).

•CPS <1 – Among the 55 patients with CPS <1, similar OS (11.5 versus 12.1 months, HR 1, 95% CI 0.52-1.94).

●Nivolumab plus ipilimumab versus nivolumab plus chemotherapy – At minimum follow-up of 13 months, nivolumab plus ipilimumab versus nivolumab plus chemotherapy demonstrated the following results. Although these results were not formally compared, PFS was numerically shorter and the OS curves were initially lower for nivolumab plus ipilimumab compared with nivolumab plus chemotherapy. ORR was also higher with nivolumab plus chemotherapy. These data suggest a risk of early disease progression and death with nivolumab plus ipilimumab alone.

•All patients – Median OS of 12.7 versus 13.2 months and ORRs of 28 versus 47 percent, respectively.

•TPS ≥1 percent – Median OS of 13.7 versus 15.4 months and ORRs of 56 versus 84 percent, respectively.

•CPS >1 – Median OS of 12.7 versus 13.8 months, respectively.

Further results from CheckMate 648 are discussed separately. (See 'Nivolumab plus chemotherapy' above.)

Based on these data, nivolumab in combination with ipilimumab was approved by the FDA for the first-line treatment of adult patients with advanced unresectable or metastatic esophageal SCC, regardless of PD-L1 expression (ie, CPS) [96]. However, in our view, this immunotherapy combination has limited benefit in advanced or metastatic SCC with low PD-L1 expression (CPS <1).

Other immunotherapy combinations — In patients with previously untreated advanced esophageal SCC, other immunotherapy agents that have demonstrated an OS benefit when combined with chemotherapy (cisplatin or oxaliplatin plus a fluoropyrimidine or paclitaxel) in randomized trials include camrelizumab (ESCORT-1^st) [146], sintilimab (ORIENT-15) [147], toripalimab (JUPITER-06) [148], and sugemalimab (GEMSTONE-304) [149].

SPECIAL CONSIDERATIONS

Poor performance status or significant comorbidities — For older patients or those with a poor performance status or significant comorbidity, options include infusional fluorouracil (FU) plus leucovorin, single-agent capecitabine, oral S-1 (where available), single-agent irinotecan, low-dose weekly taxanes (paclitaxel or docetaxel), or dose-reduced CAPOX.

Single-agent chemotherapy — Single-agent chemotherapy is an appropriate option for older or frail patients. Although these agents offer modest clinical benefit, they are typically better tolerated than combination therapy. Options include fluoropyrimidines (infusional FU plus leucovorin [150-152], single-agent capecitabine, and oral S-1, where available), which demonstrate objective response rates (ORRs) of up to 41 percent [48,153-159], single-agent irinotecan (ORRs of 14 to 20 percent) [160,161], and low-dose weekly taxanes (paclitaxel or docetaxel; ORRs of 15 to 24 percent) [26,162-169].

Dose-reduced CAPOX — In older or frail (ie, those with poor performance status and/or significant comorbidities) adults, dose-reduced CAPOX (at 60 percent of the usual dose) is a treatment option that does not compromise efficacy and is well-tolerated.

In a phase III trial (GO2), 514 older and frail patients with advanced esophagogastric cancer, including squamous cell carcinoma (SCC) histology, with an adequate baseline comprehensive geriatric assessment were randomly assigned to standard 21-day cycles of oxaliplatin (130 mg/m² on day 1) plus capecitabine (625 mg/m² twice daily on days 1 to 21), the same regimen with 80 percent of the usual doses, or the same regimen with 60 percent of the usual doses [170]. Patients with a glomerular filtration rate (GFR) of 30 to 50 mL/minute or a total bilirubin of 1.5 to 2 times the upper limit of normal received 75 percent of the allocated capecitabine doses. Compared with standard doses, noninferiority was confirmed for both the 80 percent dose regimen (hazard ratio [HR] for progression-free survival [PFS] 1.09, 95% CI 0.89-1.32) and the 60 percent dose regimen (HR 1.1, 95% CI 0.9-1.33). Overall survival (OS) was similar in all three groups (median OS 7.5, 6.7, and 7.6 months, respectively). The lowest doses provided noninferior cancer control and the best patient experience (as assessed by overall treatment utility, toxicity, and quality of life). No subgroup (based on age, performance status, extent of frailty, or baseline geriatric assessment) clearly benefited from higher-dose therapy. The study did not address whether clinicians should subsequently attempt dose escalation if initial doses of CAPOX are well-tolerated.

ASSESSING TREATMENT RESPONSE — 

Treatment response is assessed using a combination of interval radiographic studies such as contrast-enhanced computed tomography (CT) of the chest, abdomen, and pelvis (typically every two to three cycles); serum tumor markers (if elevated at baseline); and the clinical status of the patient. Radiographic tumor response is usually quantified using Response Evaluation Criteria in Solid Tumors (table 15). Immunotherapy response criteria are also available. (See "Principles of cancer immunotherapy", section on 'Immunotherapy response criteria'.)

If tumor markers are elevated at initial evaluation, such as carcinoembryonic antigen (CEA) or CA 19-9, they can be serially assessed during treatment as a surrogate measure of response to therapy. While persistently rising levels of a serum tumor marker suggest disease progression, this should be confirmed with radiologic studies prior to a change in therapy. (See "Clinical presentation, diagnosis, and staging of gastric cancer", section on 'Serologic markers'.)

TREATMENT DURATION — 

In general, initial systemic therapy is given until the patient has progressive disease or cannot tolerate further therapy. Such patients who are eligible for further systemic therapy should be promptly switched to the appropriate second-line treatment regimen. (See "Second- and later-line systemic therapy for metastatic gastric and esophageal cancer".)

There are limited data on alternative approaches such as "stop and go" strategies (ie, stopping therapy upon disease control and restarting therapy upon disease progression) [171] and modified maintenance regimens [172]. Many of these strategies also did not include the use of immunotherapy, so their clinical relevance is not clear. Each patient's treatment plan should be individualized depending on treatment response, tolerance of therapy, and the patient's wishes for treatment breaks or dose modifications.

SOCIETY GUIDELINE LINKS — 

Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Gastric cancer" and "Society guideline links: Esophageal cancer".)

INFORMATION FOR PATIENTS — 

UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Esophageal cancer (The Basics)" and "Patient education: Stomach cancer (The Basics)")

SUMMARY AND RECOMMENDATIONS

●Goals of therapy – In patients with advanced unresectable and metastatic gastric, gastroesophageal junction (GEJ), and esophageal cancer, the goals of systemic therapy include palliating symptoms, improving quality of life, and prolonging overall survival (OS). (See 'Goals of therapy' above.)

●Treatment selection – Our general approach to initial systemic therapy is to select a backbone chemotherapy regimen, and then decide whether to pair it with immunotherapy and/or targeted therapy, based on histology and tumor biomarker status (algorithm 2 and algorithm 3).

●Selection of backbone chemotherapy regimens – For patients with advanced unresectable or metastatic gastric and esophageal adenocarcinoma and esophageal squamous cell carcinoma (SCC), regardless of mismatch repair (MMR)/microsatellite stability (MSS) status, we suggest a chemotherapy backbone that includes both a fluoropyrimidine and oxaliplatin, rather than other chemotherapy combinations (Grade 2C). (See 'Backbone chemotherapy regimens' above.)

•Options include FOLFOX (table 2), CAPOX/XELOX (table 3), or modified FLOT (or TFOX; for gastric and esophageal adenocarcinoma only). FOLFIRI is an acceptable alternative for those unable to tolerate oxaliplatin. (See 'FOLFOX and CAPOX' above and 'Modified FLOT/TFOX' above and 'FOLFIRI' above.)

•We use modified FLOT in patients with gastric or esophageal adenocarcinoma, good performance status, and minimal comorbidities who require a robust initial treatment response due to malignant obstruction, visceral crisis, or other significant symptoms. (See 'Modified FLOT/TFOX' above.)

•For patients treated initially with modified FLOT or FOLFIRI, we start with chemotherapy alone. Those on modified FLOT may be switched to the appropriate initial systemic therapy (chemotherapy plus immunotherapy and/or targeted therapy, based on biomarker status) once their disease has stabilized. (See 'Modified FLOT/TFOX' above.)

●dMMR/MSI-H tumors – For patients with mismatch-repair deficient (dMMR)/microsatellite instability-high (MSI-H) metastatic esophageal and gastric cancer (either SCC or adenocarcinoma), we recommend the combination of chemotherapy and immunotherapy rather than chemotherapy alone (Grade 1B), as this approach improves OS and can induce durable treatment responses. However, immunotherapy alone (nivolumab plus ipilimumab, pembrolizumab, or dostarlimab) is an acceptable alternative for patients who are unlikely to tolerate chemotherapy.

In selecting a specific regimen, we suggest initial therapy with FOLFOX plus nivolumab (Grade 2C), followed by maintenance therapy for patients without disease progression, although comparative data between regimens are lacking. (See 'Mismatch repair deficient/MSI-H tumors' above.)

●pMMR/MSS gastric and esophageal adenocarcinoma – For patients with mismatch repair proficient (pMMR)/MSS metastatic gastric and esophageal adenocarcinoma, additional therapy is guided by tumor biomarker expression (algorithm 2) (see 'Mismatch repair proficient gastric and esophageal adenocarcinoma' above):

•HER2-positive adenocarcinoma – For patients with human epidermal growth factor receptor 2 (HER2)-positive adenocarcinomas, we recommend the addition of trastuzumab (Grade 1B) to chemotherapy. Preferred options include FOLFOX plus trastuzumab (table 10) and CAPOX plus trastuzumab (table 11). (See 'Trastuzumab plus chemotherapy' above.)

Additionally, for HER2-positive cancers that also have combined positive score (CPS) ≥1, we recommend the addition of pembrolizumab to trastuzumab and chemotherapy (Grade 1B). Preferred options include pembrolizumab plus FOLFOX and trastuzumab (table 12) and pembrolizumab plus CAPOX and trastuzumab (table 13). (See 'Pembrolizumab plus trastuzumab and chemotherapy' above.)

For patients who are ineligible for trastuzumab (table 14), we offer an initial treatment approach similar to those with HER2-negative adenocarcinoma. (See 'Ineligible for trastuzumab' above.)

•HER2-negative adenocarcinoma – For patients with HER2-negative adenocarcinomas, selection of therapy is based on tumor expression of programmed cell death ligand 1 (PD-L1) and claudin18.2 (CLDN18.2). (See 'Treatment approach' above.)

-CPS ≥5 – For patients with CPS ≥5, regardless of CDLN18.2 expression, we recommend the addition of immunotherapy to chemotherapy (Grade 1B), as this approach improves OS. For those with CPS ≥10, options include either nivolumab, pembrolizumab, or tislelizumab plus chemotherapy (see 'CPS of 10 or more' above); and for those with CPS ≥5 and <10, we suggest either nivolumab or tislelizumab plus chemotherapy rather than pembrolizumab plus chemotherapy (Grade 2C). (See 'CPS of 5 to less than 10' above.)

Although there are no comparative randomized trials, for patients with CPS ≥5, we prefer chemotherapy plus immunotherapy rather than chemotherapy plus zolbetuximab, regardless of CLDN18.2 expression status. Chemotherapy plus immunotherapy offers the opportunity for more durable treatment responses and OS, and zolbetuximab is associated with symptomatic toxicities (eg, nausea and vomiting with infusion).

-CPS <5 and CLDN18.2-positive – For those with CPS <5 and CLDN18.2-positive disease, we recommend the addition of zolbetuximab to chemotherapy (Grade 1B), which improves OS. (See 'CPS of less than 5, CLDN18.2-positive' above.)

-CPS <5 and CLDN18.2-negative – For those with CPS <5 and CLDN18.2-negative disease, we do not typically incorporate immunotherapy, given the OS benefits are limited in this population and may not outweigh the risks of therapy; we generally treat such patients with chemotherapy alone. However, we acknowledge that there may be variability in clinical practice and that providers outside of UpToDate may prefer to offer chemotherapy plus immunotherapy to patients with CPS in this range. (See 'CPS of less than 5, CLDN18.2-negative' above.)

●pMRR/MSS esophageal squamous cell carcinoma – For patients with pMMR/MSS metastatic esophageal SCC, selection of therapy is based on tumor PD-L1 expression (algorithm 3) (see 'Treatment approach' above):

•CPS ≥1 – For patients with CPS ≥1, we recommend the addition of immunotherapy to chemotherapy (Grade 1B), which improves OS. Although nivolumab plus ipilimumab is an alternative option, there is a risk of early disease progression and death with immunotherapy alone. (See 'CPS of 1 or more' above.)

-CPS ≥10 – For patients with CPS ≥10, we offer either nivolumab, pembrolizumab, or tislelizumab plus chemotherapy. (See 'CPS of 1 or more' above.)

-CPS of 1 to 10 – For those with CPS of 1 to less than 10, we suggest either nivolumab or tislelizumab plus chemotherapy rather than pembrolizumab plus chemotherapy (Grade 2C). (See 'CPS of 1 or more' above.)

•CPS <1 – For patients with CPS <1, we do not incorporate immunotherapy as the OS benefit is limited in this population. We treat such patients with chemotherapy alone. (See 'CPS less than 1' above.)

ACKNOWLEDGMENTS — 

The UpToDate editorial staff acknowledges Panos Fidias, MD, and Johanna Bendell, MD, who contributed to earlier versions of this topic review.