INTRODUCTION — Cancers of the upper gastrointestinal tract are highly lethal malignancies. Locally advanced, unresectable, and metastatic esophagogastric cancers are not curable, and the goals of therapy include symptom palliation and prolongation of survival. Palliative modalities for advanced esophageal or gastric cancer can be either local or systemic. While systemic therapy is the most effective treatment modality for patients with metastatic disease and it may adequately palliate dysphagia and other symptoms, such as nausea, pain, obstruction, perforation, or bleeding, from a locally advanced or locally recurrent primary tumor, they often require multidisciplinary management using endoscopic, surgical, radiotherapeutic, or other approaches. (See "Local palliation for advanced gastric cancer" and "Endoscopic palliation of esophageal cancer".)

This topic review will cover the approach to later lines of systemic therapy in patients who fail initial systemic therapy for advanced unresectable, and metastatic esophageal and gastric cancer. The choice of initial systemic treatment for locally advanced and metastatic disease, local methods for palliation in patients with advanced disease, and hepatic metastasectomy for the rare patient with isolated gastric cancer liver metastases are discussed separately. (See "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer" and "Local palliation for advanced gastric cancer" and "Endoscopic palliation of esophageal cancer" and "Surgical management of invasive gastric cancer", section on 'Metastasectomy'.)

APPROACH TO THE PATIENT

Goals of therapy — As with first-line therapy, the goals of second-line therapy for metastatic esophageal and gastric cancer are to palliate symptoms and improve survival. A survival benefit for second-line cytotoxic chemotherapy has been shown in at least five randomized trials of ramucirumab (both alone and with paclitaxel), single-agent docetaxel, and irinotecan [1-5]. A Cochrane review of salvage systemic therapy for advanced gastric and esophagogastric junction adenocarcinoma after failure of a first-line platinum/fluoropyrimidine-containing regimen concluded that both chemotherapy and immunotherapy using immune checkpoint inhibitors improve survival over placebo in the second-line setting [6].

Patient selection — The criteria to select patients for second-line systemic therapy have not been established. In one study, five factors were identified that were independently associated with poor survival: a performance status of 2, hemoglobin ≤11.5 g/dL, serum carcinoembryonic antigen (CEA) level >50 ng/mL, three or more metastatic sites, and a time to progression of six or fewer months after the first-line regimen [7]. A prognostic index was developed, which divided patients into low- (no risk factors), intermediate- (one or two risk factors), or high-risk (three or more risk factors) groups with different median survival durations (12.7, 7.1, and 3.3 months, respectively).

None of these factors can be used to predict which patients might benefit from a second-line regimen, and in clinical practice, we do not use this specific algorithm to select patients for second-line treatment. Instead, we prefer to make this decision on a case-by-case basis, after consulting with the patient and assessing his or her overall medical condition and disease status.

Regimen choice — Performance status, comorbidity, patient preference, molecular analysis, quality of life, histologic type, and the regimen that was used for initial therapy are key considerations when choosing the therapeutic approach. For patients with a poor performance status, poorly controlled comorbidity, or a preference for no additional therapy, we suggest supportive care alone. For patients who retain an adequate performance status, there is no standard approach to second-line therapy and beyond after failure of the first-line regimen. We prefer enrollment in a clinical trial if available. If trials are unavailable or patients are ineligible, our general approach is outlined in the algorithm (algorithm 1).

Particularly in view of the lack of data showing benefit of multiagent cytotoxic chemotherapy versus monotherapy [8,9], quality of life and minimization of side effects are key considerations when choosing the regimen.

In general:

●Adenocarcinoma

•For most patients with adenocarcinoma who retain an excellent performance status and are not eligible for immunotherapy or who received programmed cell death receptor 1 (PD-1) targeted immunotherapy for initial therapy, we suggest ramucirumab plus paclitaxel after progression on or prior treatment with fluoropyrimidine- or platinum-containing chemotherapy. (See "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Frontline immunotherapy'.)

•The optimal second-line treatment strategy for patients with human epidermal growth factor receptor 2 (HER2)-overexpressing, advanced gastroesophageal adenocarcinoma is unclear. Potential benefit has been shown for the antibody-drug conjugate fam-trastuzumab deruxtecan after progression on first-line trastuzumab, and this approach is reasonable, although myelosuppression and interstitial lung disease are notable toxicities. There has been no strong evidence to support continued targeting of HER2 beyond first-line trastuzumab, and this approach cannot be generally recommended. However, a weakness in most of the studies is that HER2 status was not reconfirmed prior to initiating the second-line HER2-directed therapy. If use of second-line trastuzumab is being considered, it is prudent to reconfirm HER2 positivity with repeat biopsy. (See 'HER2-positive disease and continued targeting of HER2 after progression' below.)

•Trifluridine-tipiracil is a reasonable choice as a third-line option for patients with an adenocarcinoma that lacks overexpression of programmed cell death ligand 1 (PD-L1) and has proficient mismatch repair, and as a fourth-line option for patients with a PD-L1-positive/deficient mismatch repair (dMMR) adenocarcinoma. (See 'Trifluridine-tipiracil' below.)

●Squamous cell carcinoma

•For patients with squamous cell carcinoma (SCC) who retain an excellent performance status and who are not eligible for immunotherapy or who received PD-1 targeted immunotherapy for first-line treatment, utilization of other active cytotoxic chemotherapy agents not used in the first-line regimen is reasonable, either in combination or as serial single agents. For example, patients who received FOLFOX (oxaliplatin plus leucovorin and short-term infusional fluorouracil [FU]) initially could be offered single-agent irinotecan, a taxane, or FOLFIRI (irinotecan plus leucovorin and short-term infusional FU (table 1)). (See "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Squamous cell cancers' and "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Single-agent chemotherapy'.)

•Another potential option for treatment at progression for SCCs, irrespective of microsatellite instability (MSI), tumor mutational burden (TMB), or PD-L1 overexpression is single-agent nivolumab, or, where available, camrelizumab, if PD-1 targeted immunotherapy was not administered for first-line therapy. Nivolumab is approved for second-line therapy for esophageal SCC in the United States. (See 'Esophageal squamous cell cancer' below.)

●Either histology

•For patients with either adenocarcinoma or SCC and only an adequate performance status, or those for whom limiting treatment-related toxicity is an important goal, we suggest monotherapy rather than combination chemotherapy. Options include irinotecan, weekly paclitaxel, weekly nanoparticle albumin-bound paclitaxel (nabpaclitaxel), apatinib (where available), or ramucirumab monotherapy (for patients with adenocarcinoma). (See 'Cytotoxic chemotherapy' below and 'Treatments targeting VEGF' below.)

•Another option for treatment at progression for patients who have advanced tumors with dMMR, high levels of MSI (MSI-H), high levels of TMB, or PD-L1 overexpression, and who did not receive PD-1 targeted immunotherapy for first-line treatment, is pembrolizumab monotherapy. (See "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Frontline immunotherapy'.)

However, in this setting, there is disagreement about the timing of pembrolizumab. In the United States, pembrolizumab is approved for third-line treatment in patients with PD-L1-expressing tumors (combined positive score [CPS] 1 or higher) after failure of two separate chemotherapy regimens. However, support for choosing pembrolizumab, rather than second-line chemotherapy, for patients with esophageal SCC or Siewert type I (figure 1) esophagogastric junction (EGJ) adenocarcinomas that overexpress PD-L1 (CPS 10 or higher) is provided by a preliminary report of the KEYNOTE-181 trial, which suggests a survival advantage and better tolerability for pembrolizumab in this setting.

Other clinicians might pursue second-line pembrolizumab off label for a patient with a dMMR tumor of the stomach or EGJ, given the subgroup analysis of the KEYNOTE-061 trial, which showed a significant survival benefit for pembrolizumab over a taxane in this setting. Another group that might be considered for second-line pembrolizumab is the subset of patients with TMB-H tumors. In June 2020, pembrolizumab was approved across a variety of tumor types for the treatment of adult and pediatric patients with unresectable or metastatic TMB-H solid tumors, as determined by the FoundationOne CDx assay, that have progressed following prior treatment and who have no satisfactory treatment alternatives. (See 'Timing of immunotherapy' below.)

In Japan, nivolumab is approved for advanced gastric cancer that has progressed after conventional therapy, without regard to the number of prior regimens or specific biomarker expression.

●Immunotherapy rechallenge

•Some patients will have received PD-1 targeted immunotherapy as initial therapy for advanced disease, either as monotherapy, or in conjunction with cytotoxic chemotherapy. (See "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Frontline immunotherapy'.)

For these patients, we would generally not pursue anti-PD-1 immunotherapy if disease progression occurred during prior PD-1 targeted therapy. However, rechallenge may be considered in the context of a clinical trial, if prior anti-PD-1 immunotherapy was discontinued due to a specific toxicity that has since resolved, and possibly for patients with progression at few sites (oligoprogression) after a prior objective response or prolonged period of stable disease during initial PD-1 targeted immunotherapy. (See "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Frontline immunotherapy'.)

EFFICACY OF INDIVIDUAL TREATMENTS

Cytotoxic chemotherapy — In general, clinical trials assessing the efficacy of a variety of second-line therapy regimens after failure of the first-line regimen in advanced esophagogastric cancer have shown that response rates are lower than they are in previously untreated patients, and toxicity rates tend to be higher [10-29].

The choice of regimen is empiric. No single regimen has emerged as a clear "winner," and few trials have compared different regimens [30,31]:

●A Japanese trial comparing second-line irinotecan monotherapy (150 mg/m² on days 1 and 15 every four weeks) with weekly paclitaxel (80 mg/m² on days 1, 8, and 15 every four weeks) concluded that neither regimen was superior in terms of efficacy or tolerability [32]. A similar conclusion was reached in a second similar trial from South Korea [33].

●No differences in outcomes could be shown in two trials comparing irinotecan plus cisplatin versus irinotecan alone, and in one trial comparing single-agent nanoparticle albumin-bound paclitaxel (nabpaclitaxel) versus solvent-based paclitaxel (although hypersensitivity reactions were less common with nabpaclitaxel) [9,34,35].

●On the other hand, a randomized phase II trial of every three week docetaxel versus weekly paclitaxel in 78 patients with esophageal SCC refractory to fluoropyrimidine and platinum-based chemotherapy concluded that paclitaxel was more efficacious (response rate 26 versus 6 percent, median progression-free survival (PFS) 4.4 versus 2.1 months) and less toxic [36].

Combination regimens containing agents not used in the first-line setting are also reasonable in the second line setting (eg, FOLFIRI [irinotecan plus leucovorin and short-term infusional fluorouracil (FU)] or taxane-based therapy after FOLFOX [oxaliplatin plus leucovorin and short-term infusional FU] or cisplatin plus FU). (See "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Irinotecan-containing regimens' and "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Docetaxel-containing'.)

Any of these approaches could be considered appropriate for second-line treatment of adenocarcinomas and squamous cell cancers. For adenocarcinomas, there are no trials comparing any of these regimens with trifluridine/tipiracil or ramucirumab-containing therapy. (See 'Treatments restricted to adenocarcinomas' below.)

Immunotherapy — Immunotherapeutic approaches to cancer therapy are based on the premise that the immune system plays a key role in surveillance and eradication of malignancy and that tumors evolve ways to elude the immune system. We would only pursue immunotherapy if it was not administered in the frontline setting. There is no evidence to support benefit from a switch to a different immune checkpoint inhibitor for patients whose disease has progressed on one immune checkpoint inhibitor, regardless of histology. (See "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Frontline immunotherapy'.)

PD-1 and PD-L1 inhibitors — One option for treatment at progression for patients who have advanced squamous cell carcinoma (SCC) or adenocarcinoma of the esophagus or stomach with either DNA mismatch repair deficiency (dMMR; as determined by immunohistochemistry [IHC] or the finding of high levels of microsatellite instability [MSI-H]) or overexpression of programmed cell death ligand 1 (PD-L1) is immunotherapy with an immune checkpoint inhibitor that targets programmed cell death receptor 1 (PD-1; ie, pembrolizumab) if it was not administered for frontline therapy. (See "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Frontline immunotherapy'.)

For patients whose tumors are dMMR or overexpress PD-L1 and who did not receive first-line immune checkpoint inhibitor immunotherapy, there is disagreement about the optimal timing of immune checkpoint inhibitors. In the United States, pembrolizumab is approved for third-line treatment in patients with PD-L1-expressing gastric or esophagogastric junction (EGJ) adenocarcinoma (combined positive score [CPS] 1 or higher) after failure of two separate chemotherapy regimens. It is also approved for recurrent locally advanced or metastatic SCC of the esophagus that expresses high levels of PD-L1 (CPS ≥10) and has progressed after one or more prior lines of systemic therapy.

However, support for choosing pembrolizumab off label, rather than as second-line chemotherapy, for patients with highly PD-L1-overexpressing cancers of the esophagus or EGJ, Siewert type I (figure 1), is provided by a preliminary report of the KEYNOTE-181 trial, which suggests a survival advantage and better tolerability for pembrolizumab in this setting. Other clinicians might pursue second-line pembrolizumab off label for a patient with a dMMR tumor in the stomach or EGJ, given the subgroup analysis of the KEYNOTE-061 trial, which showed a significant benefit for pembrolizumab over a taxane in this setting. (See 'Timing of immunotherapy' below.)

Notably, in Japan, nivolumab is approved for the treatment of unresectable, advanced, or recurrent gastric cancer that has progressed after "conventional chemotherapy," without regard to specific biomarker expression.

PD-1 is a key immune checkpoint receptor that is expressed by activated T cells [37,38]. Tumors use the PD-1 pathway to evade immune surveillance. PD-1 binds to its ligands PD-L1 (B7 homolog 1 [B7-H1]) and PD-L2 (B7 homolog 2 [B7-H2]), which are expressed on tumor cells, thereby causing immunosuppression and preventing the immune system from rejecting the tumor. (See "Principles of cancer immunotherapy".)

Gastric and esophagogastric junction adenocarcinoma — Pembrolizumab is an anti-PD-1 monoclonal antibody that is approved for use in advanced melanoma. Activity in advanced refractory PD-L1-expressing esophagogastric cancer has been suggested in the following trials:

●A report of the gastric cancer cohort of the phase Ib KEYNOTE-012 trial evaluating the benefit of pembrolizumab (10 mg/kg every two weeks) in PD-L1-expressing, advanced, solid tumors included 39 patients with gastric or EGJ tumors, 68 percent of whom had ≥2 prior therapies for metastatic disease [39]. Of the 36 patients available for response by central assessment, eight (22 percent) had an objective response (none complete), and an additional 8 percent had stable disease. The median duration of response was 40 weeks. Only five patients had a total of six grade 3 or 4 treatment-related adverse events, which included grade 3 fatigue, pemphigoid, hypothyroidism, and peripheral sensory neuropathy; there was one case of grade 4 pneumonitis.

●A report of the advanced esophageal cancer cohort of the phase Ib KEYNOTE-028 trial evaluating the benefit of pembrolizumab (10 mg/kg every two weeks) in PD-L1-expressing, advanced, solid tumors included 23 patients, 87 percent of whom had ≥2 prior therapies for metastatic disease [40]. Both adenocarcinomas and SCC were eligible. There were seven confirmed partial responses (objective response rate 30 percent), and two additional patients had stable disease. By histologic subtype, the objective response rate was higher for adenocarcinoma (2 of 5 versus 5 of 18 [40 versus 28 percent]). The median duration of response was 15 months. Only four patients had grade 3 treatment-related adverse events, which included lymphopenia, anorexia, liver disorder, and generalized rash.

●The superiority of pembrolizumab over systemic chemotherapy (investigators' choice of paclitaxel, docetaxel, or irinotecan) for second-line therapy in patients with an esophageal or EGJ, Siewert type I (figure 1), cancer whose tumors express high levels of PD-L1 (CPS 10 or more) was shown in the phase III KEYNOTE-181 trial.

●Responses seem to be less frequent (but not completely absent) in patients with PD-L1-nonexpressing tumors. This was shown in two separate analyses:

•In the KEYNOTE-059 cohort of 259 patients with previously treated gastric or EGJ cancer, the objective response rate in the entire cohort was 11.6 percent (2.3 percent complete) [41]. Among those with PD-L1-positive tumors (expression in ≥1 percent of tumor or stromal cells), the objective response rate was 15.5 percent, with 13.5 percent partial and 2 percent complete responses. By contrast, the objective response rate in those with PD-L1-negative tumors was only 6.4 percent, with 2.8 percent experiencing a complete response. Grade 3 to 5 treatment-related adverse events developed in 18 percent, and two were fatal (acute kidney injury, pleural effusion).

•Similar results were noted in the phase II KEYNOTE-180 study of 121 patients with advanced, metastatic SCC or adenocarcinoma of the esophagus or EGJ that had progressed after two or more lines of systemic therapy [42]. The objective response was twofold higher among those with PD-L1-overexpressing tumors (14 versus 6 percent).

Early reports also suggest meaningful activity for nivolumab [43,44], another anti-PD-1 antibody, but not for avelumab [45], an inhibitor of PD-L1, in patients with advanced, treatment-refractory gastroesophageal adenocarcinoma, not previously treated with PD-1 targeted immunotherapy:

●The benefit of nivolumab was explored in a phase 3 trial (ONO-4538, the ATTRACTION-2 trial) in which 493 patients with advanced gastric or EGJ cancer who had failed two or more standard chemotherapy regimens were randomly assigned to nivolumab (3 mg/kg) or placebo every two weeks until unacceptable toxicity or disease progression [46]. The objective response rate with nivolumab was 11 percent, and median overall survival was improved compared with placebo (5.3 versus 4.1 months [hazard ratio (HR) 0.63, 95% CI 0.51-0.78], 12-month survival 27 versus 11 percent); median PFS was also improved (1.61 versus 1.45 months [HR 0.60, 95% CI 0.49-0.75]). Severe (grade ≥3) adverse reactions were infrequent with nivolumab (11.5 versus 5.5 percent with placebo).

Notably, the trial did not restrict therapy to any subset of patients expressing a biomarker such as PD-L1. Furthermore, both of these trials were conducted entirely in Asian populations.

●A comparable degree of benefit in Western populations was suggested in the CheckMate-032 study, in which 160 patients with disease progression on or intolerance of at least one systemic chemotherapy regimen for advanced gastric, esophageal, or EGJ cancer were randomly assigned to nivolumab alone (3 mg/kg) every two weeks, nivolumab 1 mg/kg plus ipilimumab 3 mg/kg (nivo 1 plus ipi 3) every three weeks, or nivolumab 3 mg/kg plus ipilimumab 1 mg/kg (nivo 3 plus ipi 1) every three weeks [47]. Seventy-nine percent had received two or more prior therapies. Nivolumab, with or without ipilimumab, led to durable responses and long-term overall survival, and responses were observed regardless of tumor PD-L1 status. Twelve-month PFS rates for nivolumab alone, nivo 1 plus ipi 3, and nivo 3 plus ipi 1 were 8, 17, and 10 percent; the corresponding rates of 12-month survival were 39, 37, and 24 percent; and objective responses were seen in 12, 24, and 8 percent, respectively. Grade 3 or worse treatment-related adverse events were reported in 17, 47, and 27 percent of the three groups, respectively.

●On the other hand, benefit for avelumab could not be shown in the phase III JAVELIN Gastric 300 trial, in which 371 patients failing two prior lines of systemic therapy were randomly assigned to avelumab or to systemic therapy of the clinician's choice (either weekly paclitaxel or irinotecan monotherapy; patients ineligible for systemic therapy received best supportive care only) [45]. Avelumab did not improve overall survival or PFS.

Esophageal squamous cell cancer — The benefits of second-line therapies targeting the PD-1 pathway in SCC arising in the esophagus have been shown in the following trials:

●The superiority of pembrolizumab (an anti-PD-1 monoclonal antibody [MoAb]) over systemic chemotherapy (investigators' choice of paclitaxel, docetaxel, or irinotecan) for second-line therapy was shown in the phase III KEYNOTE-181 trial, which enrolled 628 patients with advanced/metastatic SCC (64 percent) or adenocarcinoma of the esophagus or EGJ, Siewert type I [48]. Among the 222 patients who had highly PD-L1-positive tumors (PD-L1 CPS ≥10, defined as the number of PD-L1-positive cells divided by the total number of tumor cells and then multiplied by 100), median overall survival was superior with pembrolizumab (9.3 versus 6.7 months), twice as many individuals remained alive at 12 months (43 versus 20 percent), and there were fewer grade 3 to 5 drug-related adverse events (18 versus 41 percent). There was no improvement in overall survival with pembrolizumab in the entire intent-to-treat population, which included the 496 patients without high levels of PD-L1 overexpression (n = 628, median 7.1 months in both groups). In subset analysis, among those with CPS ≥10, the survival benefit was significant for SCC (HR 0.64, 95% CI 0.46-0.90) but not adenocarcinoma (HR 0.93, 95% CI 0.52-1.65). When the analysis was limited to SCC, the survival benefit for pembrolizumab was restricted to those with CPS ≥10 and not lower levels.

Largely based on these data, in July 2019, the US Food and Drug Administration (FDA) approved pembrolizumab for patients with recurrent locally advanced or metastatic SCC of the esophagus whose tumors express high levels of PD-L1 (CPS ≥10) and who have disease progression after one or more prior lines of systemic therapy (ie, second-line therapy).

●Nivolumab, another anti-PD-1 MoAb, is also active, although it is not clear whether PD-L1 expression status is important to tumor response:

•In an early, uncontrolled trial of nivolumab, objective responses were seen in 11 of 65 patients (17 percent) with refractory metastatic esophageal SCC, but the median time to progression was only 2.8 months; the safety profile was expected and manageable [43]. This trial did not restrict enrollment to any subset of patients expressing a biomarker such as PD-L1.

•Support for second-line nivolumab was also provided by the phase 3 ATTRACTION-3 trial, in which 419 patients with previously treated (at least one fluoropyrimidine- and platinum-based regimen), advanced esophageal SCC were randomly assigned to nivolumab (240 mg every two weeks) or a single-agent taxane (investigators' choice of paclitaxel or docetaxel) [49]. Patients were unselected for biomarker expression; approximately one-half had PD-L1 expression levels in tumor cells <1 percent (using a different assay than that used to determine PD-L1 overexpression in the KEYNOTE-181 trial). Overall survival was significantly better with nivolumab (median 10.9 versus 8.4 months, HR for death 0.77, 95% CI 0.62-0.96), and fewer patients had grade 3 or 4 treatment-related adverse effects (18 versus 63 percent). The survival benefit was independent of PD-L1 expression status in the tumor; survival benefit according to PD-L1 expression in immune and tumor cells (ie, the CPS) has not been reported, making it more difficult to compare directly with pembrolizumab.

Largely based on this study, in June 2020, the US FDA approved nivolumab for the treatment of patients with unresectable advanced, recurrent or metastatic esophageal squamous cell carcinoma (ESCC) after prior fluoropyrimidine- and platinum-based chemotherapy [50]. We would generally not pursue anti-PD-1 monotherapy for second-line therapy or beyond if disease progression occurred during prior PD-1 targeted therapy. (See "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Frontline immunotherapy'.)

●Camrelizumab, a third anti-PD-1 MoAb, which is approved in China for treatment of relapsed or refractory classical Hodgkin lymphoma [51], also appears to be active in advanced esophageal SCC [52]. In the randomized phase III ESCORT trial, in which 457 patients previously treated with first-line platinum-based chemotherapy were randomly assigned to camrelizumab or chemotherapy (docetaxel or irinotecan), second-line camrelizumab significantly improved overall survival (8.3 versus 6.2 months, HR 0.71, 95% CI 0.57-0.87), and benefit was irrespective of PD-L1 overexpression.

Biomarkers and benefit from PD-1 inhibitors — The issue of using biomarkers to predict benefit from programmed cell death receptor 1 (PD-1)-blocking antibodies in esophagogastric cancer is in evolution [53].

PD-L1 overexpression — As noted above, responses to PD-1 inhibitors seem to be less frequent (but not completely absent) in patients with programmed cell death ligand 1 (PD-L1)-nonexpressing tumors. In the United States, pembrolizumab is approved for third-line treatment in patients with PD-L1-expressing esophagogastric tumors (CPS 1 or higher), and as a second-line treatment for esophageal squamous cell cancers that express high levels of PD-L1 (CPS 10 or higher), while in Japan, nivolumab is approved for the treatment of unresectable, advanced, or recurrent gastric cancer that has progressed after "conventional chemotherapy," without regard to specific biomarker expression. We would generally not pursue anti-PD-1 monotherapy for second-line therapy or beyond if disease progression occurred during prior PD-1 targeted therapy. (See 'PD-1 and PD-L1 inhibitors' above and "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Frontline immunotherapy'.)

Defective mismatch repair — It has been hypothesized that tumors that lack the mismatch repair mechanism (ie, dMMR) harbor many more mutations (ie, they are hypermutated) than do tumors of the same type without such mismatch repair defects, and that the neoantigens generated by mutations such as these are more immunogenic than those generated by other mutations. The biologic footprint of dMMR tumors is MSI-H.

Several steps are required for the immune system to effectively attack tumor cells. Several immune checkpoints exist to dampen the immune response in order to protect against detrimental inflammation and autoimmunity. In the setting of malignancy, such immune checkpoints can result in immune tolerance and tumor progression. Inhibition of these checkpoints might be expected to halt/reverse disease progression. One well-characterized checkpoint being targeted in several tumor types is PD-1. PD-1 is upregulated on activated T cells, and upon recognition of tumor via the T cell receptor, PD-1 engagement by PD-L1 expressed by the tumor or other immune cells infiltrating the tumor tissue can lead to T cell inactivation and a "brake" on immune-mediated tumor eradication. (See "Principles of cancer immunotherapy".)

Proof of principle that cancers with dMMR might be particularly susceptible to inhibition of the PD-L1/PD-1 interaction was initially provided by a study of pembrolizumab in dMMR colorectal cancer. Subsequently, it has been established that immune checkpoint inhibitor immunotherapy appears to benefit a subset of patients with dMMR tumors, regardless of anatomic site of origin or tissue histology. As a result, in May 2017, the FDA approved pembrolizumab for treatment of a variety of advanced solid tumors, including gastric cancers, that had MSI-H or dMMR, that had progressed following prior treatment, and for which there were no satisfactory alternative treatment options, the first such approval of a tissue-agnostic anticancer treatment. (See "Tissue-agnostic cancer therapy: DNA mismatch repair deficiency, tumor mutational burden, and response to immune checkpoint blockade in solid tumors", section on 'Other tumors with MSI-H/dMMR'.)

Approximately 3 to 7 percent of esophagogastric adenocarcinomas are dMMR/MSI-H, most of which also overexpress PD-L1 [54-63]. Efficacy of pembrolizumab in dMMR esophagogastric cancer has been shown in the following studies:

●In the phase II KEYNOTE-158 study, which enrolled 24 patients with gastric cancer [64], there were 11 objective responses (46 percent), four of which were complete, and median PFS was 11 months.

●A subsequent post-hoc analysis of data from KEYNOTE-059 (third-line pembrolizumab), KEYNOTE-061 (second-line pembrolizumab versus chemotherapy) and KEYNOTE-062 (first-line pembrolizumab, versus pembrolizumab plus chemotherapy versus chemotherapy alone) also demonstrated that compared with the overall study population, patients with dMMR/MSI-H tumors experienced greater benefit from pembrolizumab alone or in combination with chemotherapy regardless of the line of treatment [63].

First-line immunotherapy is becoming an increasingly accepted approach for PD-L1-overexpressing adenocarcinomas, especially if dMMR/MSI-H. We would generally not pursue anti-PD-1 monotherapy for second-line therapy or beyond if disease progression occurred during prior PD-1-targeted therapy. (See "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Frontline immunotherapy'.)

The approach to testing for dMMR is addressed in detail elsewhere. (See "Tissue-agnostic cancer therapy: DNA mismatch repair deficiency, tumor mutational burden, and response to immune checkpoint blockade in solid tumors", section on 'Assessing mismatch repair' and "Tissue-agnostic cancer therapy: DNA mismatch repair deficiency, tumor mutational burden, and response to immune checkpoint blockade in solid tumors", section on 'Approach to testing dMMR as a predictive marker'.)

●Genetic implications – MSI-H or dMMR may indicate the presence of Lynch syndrome, an inherited condition that predisposes to several cancers, including gastric cancer. All patients with an MSI-H/dMMR solid tumor should undergo germline genetic assessment for Lynch syndrome, regardless of tumor type or family history [65]. (See "Lynch syndrome (hereditary nonpolyposis colorectal cancer): Clinical manifestations and diagnosis", section on 'Microsatellite instability testing'.)

High tumor mutational burden — Between 5 and 19 percent of gastric adenocarcinomas and approximately 2.4 percent of esophageal cancers have high levels of tumor mutational burden (TMB) [66,67], although this has been variably quantified. High TMB tumors have lower mutational levels than do those with dMMR, but TMB appears to be an independent biomarker of benefit for immune checkpoint inhibitor immunotherapy. Based upon an early report of the KEYNOTE-158 trial (which did not include patients with advanced esophageal or gastric cancer), pembrolizumab is now approved for patients with any solid tumor, including esophagogastric cancer, that has a TMB ≥10 mutations per megabase (mut/Mb), after progression on standard regimens. We would generally not pursue anti-PD-1 monotherapy for second-line therapy or beyond if disease progression occurred during prior PD-1 targeted therapy. (See "Tissue-agnostic cancer therapy: DNA mismatch repair deficiency, tumor mutational burden, and response to immune checkpoint blockade in solid tumors", section on 'Tumors with high mutational burden' and "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Frontline immunotherapy'.)

Although it remains uncertain whether this is the appropriate threshold to define high TMB in esophagogastric cancer, particularly in patients without microsatellite instability [68], as validation studies were conducted mainly in lung and urothelial cancers, and thresholds for TMB are likely to vary across tumor types, we suggest using a next-generation sequencing (NGS) platform to assess TMB, and the use of a threshold TMB of ≥10 mut/Mb to select patients for immune checkpoint inhibitor immunotherapy. Importantly, the results from KEYNOTE-158 were based on the FoundationOne CDx assay; cutoffs for TMB using other platforms have not been identified or linked to the FoundationOne data. (See "Tissue-agnostic cancer therapy: DNA mismatch repair deficiency, tumor mutational burden, and response to immune checkpoint blockade in solid tumors", section on 'Approach to testing for high levels of TMB'.)

Timing of immunotherapy — As noted above, there is disagreement as to the appropriate timing for a trial of a PD-1 inhibitor in patients with advanced esophagogastric cancer that overexpresses PD-L1 or is MSI-H/dMMR or TMB-high (TMB-H), and has not received frontline PD-1 targeted immunotherapy. In the United States, pembrolizumab is approved for third-line treatment of PD-L1-overexpressing adenocarcinomas after failure of two separate chemotherapy regimens, for advanced or metastatic SCC of the esophagus that expresses PD-L1 at a CPS ≥10 and has progressed after one or more prior lines of systemic therapy, for dMMR/MSI-H or TMB-H tumors that had progressed following prior treatment and for which there are no satisfactory alternative treatment options.

Three trials have directly compared pembrolizumab or nivolumab versus cytotoxic chemotherapy in the second-line setting:

●Pembrolizumab

•In the KEYNOTE-061 trial, 592 patients with advanced gastric or EGJ cancer that had progressed on chemotherapy using a platinum and fluoropyrimidine combination were randomly assigned to pembrolizumab (200 mg every three weeks for up to two years) or paclitaxel monotherapy (80 mg/m² on days 1, 8, and 15 every 28 days) [69]. Patients were initially enrolled without regard to PD-L1 overexpression, but enrollment was restricted to those with a CPS for PD-L1 ≥1 after an interim analysis suggested inferior results with pembrolizumab in those with a CPS <1. When the analysis was limited to the 395 randomly allocated patients with a CPS ≥1, pembrolizumab did not significantly prolong overall survival (median 9.1 versus 8.3 months, HR 0.82, 95% CI 0.66-1.03), and the objective response rates were similar (16 versus 14 percent). However, the adverse event profile was better with pembrolizumab (grade 3 to 5 treatment-related adverse events in 14 versus 35 percent). The main differences were in fatigue, neuropathy, and cytopenia. In a post hoc analysis, which should be considered hypothesis-generating only, the pembrolizumab treatment effect appeared to be greater for patients with a PD-L1 CPS of 10 or higher and for patients whose tumors were MSI-H irrespective of the CPS.

In a later update of these results, presented at the 2020 ASCO annual meeting, the duration of response with pembrolizumab was significantly better than with paclitaxel (median 19.1 versus 5.2 months), and the survival benefit for pembrolizumab in PD-L1 positive cases persisted [70]. The survival improvement with second-line pembrolizumab for those with CPS ≥5 was 10.4 versus 8.3 months (HR 0.72, 95% CI 0.53-0.99), while for CPS ≥10, it was 10.4 versus 8.0 months (HR 0.69, 95% CI 0.46-1.05).

•The superiority of pembrolizumab over systemic chemotherapy (investigators' choice of paclitaxel, docetaxel, or irinotecan) for second-line therapy was also shown in the phase III KEYNOTE-181 trial, which enrolled 628 patients with advanced/metastatic SCC or adenocarcinoma of the esophagus or EGJ, Siewert type I [48]. As noted above, among the 222 patients who had PD-L1-positive tumors (PD-L1 CPS ≥10, defined as the number of PD-L1-positive cells divided by the total number of tumor cells and then multiplied by 100), median overall survival was superior with pembrolizumab (9.3 versus 6.7 months), twice as many individuals remained alive at 12 months (43 versus 20 percent), and there were fewer grade 3 to 5 drug-related adverse events (18 versus 41 percent); survival was not impacted. In subset analysis, among those with CPS ≥10, a survival benefit favoring pembrolizumab was evident for SCC (HR 0.64, 95% CI 0.46-0.90) but not adenocarcinoma. (See 'Esophageal squamous cell cancer' above.)

Largely based on these data, in July 2019, the FDA approved pembrolizumab for patients with recurrent locally advanced or metastatic SCC of the esophagus whose tumors express high levels of PD-L1 (CPS ≥10) and who have disease progression after one or more prior lines of systemic therapy (ie, second-line therapy). We would generally not pursue anti-PD-1 monotherapy for second-line therapy or beyond if disease progression occurred during prior PD-1 targeted therapy. (See 'Esophageal squamous cell cancer' above.)

In our view, these data also provide support for the routine use of pembrolizumab in the second-line setting for individuals with highly PD-L1-overexpressing (CPS 10 or more) adenocarcinomas of the esophagus or EGJ, Siewert type I (figure 1). Other clinicians might pursue second-line pembrolizumab off label for a patient with a dMMR tumor of the stomach or EGJ, given the subgroup analysis of the KEYNOTE-061 trial, which showed a significant survival benefit for pembrolizumab over a taxane in this setting.

Of note, both the FDA and the European Medicines Agency have approved a new dosing regimen of 400 mg every six weeks for pembrolizumab across all currently approved adult indications, in addition to the current 200 mg every three weeks dosing regimen [71,72]. The extension of the dosing interval may be of particular benefit during the coronavirus disease 2019 (COVID-19) pandemic.

●Nivolumab

•Support for second-line nivolumab over chemotherapy was also provided by the phase 3 ATTRACTION-3 trial, in which 419 patients with previously treated (at least one fluoropyrimidine- and platinum-based regimen) advanced esophageal SCC were randomly assigned to nivolumab (240 mg every two weeks) or a single-agent taxane (investigators' choice of paclitaxel or docetaxel) [49]. Patients were unselected for biomarker expression. Overall survival was significantly better with nivolumab (median 10.9 versus 8.4 months, HR for death 0.77, 95% CI 0.62-0.96), and fewer patients had grade 3 or 4 treatment-related adverse effects. (See 'Esophageal squamous cell cancer' above.)

Nivolumab is approved in the United States for treatment of advanced squamous cell esophageal cancer after prior fluoropyrimidine and platinum-based chemotherapy, regardless of biomarker expression. In addition, in Japan, nivolumab is approved for the treatment of unresectable, advanced, or recurrent gastric cancer that has progressed after "conventional chemotherapy." We would generally not pursue anti-PD-1 monotherapy for second-line therapy or beyond if disease progression occurred during prior PD-1 targeted therapy.

Targeting CTLA-4 — Another immune checkpoint that may be exploited for immunotherapy is cytotoxic T lymphocyte-associated protein 4 (CTLA-4), which negatively regulates T cell effector responses. By blocking the interaction between CTLA-4 and its ligands, the anti-CTLA-4 monoclonal antibody ipilimumab promotes antitumor responses through T cell activation and tumor infiltration. (See "Principles of cancer immunotherapy", section on 'Checkpoint inhibitor immunotherapy'.)

Unfortunately, benefit for ipilimumab monotherapy in advanced gastric cancer could not be shown in a randomized phase II trial in which 114 patients with unresectable, locally advanced or metastatic gastric or EGJ cancer who achieved at least stable disease with first-line chemotherapy were randomly assigned to ipilimumab (10 mg/kg every three weeks for four doses, then 10 mg/kg every 12 weeks for up to three years) or best supportive care [73]. The patients receiving best supportive care could have received the same fluoropyrimidine that they received during first-line chemotherapy as maintenance but no other systemic anticancer therapy. Immune-related response and progression criteria (table 2) were used to define immune-related PFS (irPFS), the primary endpoint. There was no improvement in either irPFS (2.92 versus 4.90 months) or median overall survival (12.7 versus 12.1 months) with ipilimumab.

There is continued excitement about the potential benefit of these treatments for patients with advanced gastric cancers.

Treatments restricted to adenocarcinomas

Trifluridine-tipiracil — Efficacy for trifluridine-tipiracil was suggested in the TAGS trial, but the comparator arm was placebo and not an alternative chemotherapy regimen [74]. In this trial of 507 heavily pretreated patients (over 60 percent of the patients in each group had received three or more prior chemotherapy regimens) with adenocarcinoma of the stomach or EGJ, trifluridine-tipiracil (35 mg/m² orally twice daily on days 1 through 5 and 8 through 12 of each 28-day cycle) significantly improved overall survival over placebo (median 5.7 versus 3.6 months) and was reasonably well tolerated. The most frequent grade 3 or higher adverse effects were neutropenia and anemia in the trifluridine-tipiracil group.

Largely based on these data, in February 2019, the FDA approved trifluridine-tipiracil for adult patients with metastatic gastric or gastroesophageal junction adenocarcinoma previously treated with at least two prior lines of chemotherapy that included a fluoropyrimidine, a platinum, either a taxane or irinotecan, and if appropriate, human epidermal growth factor receptor 2 (HER2)-targeted therapy.

An important point is that oral absorption of trifluridine-tipiracil appears to be adequate even in the setting of prior gastrectomy [75].

Treatments targeting VEGF — Elevated serum and tumor levels of vascular endothelial growth factor (VEGF) are associated with a poor prognosis in patients with resectable gastric adenocarcinoma [76-78]. In animal models of gastric adenocarcinoma, inhibition of vascular endothelial growth factor receptor (VEGFR)-2 reduces tumor growth and vascularity [79]. The importance of VEGFR-2 signaling as a therapeutic target in advanced gastric and EGJ adenocarcinomas was confirmed in the placebo-controlled phase III REGARD trial, which demonstrated a modest but significant survival benefit for the VEGFR-2 inhibitor ramucirumab after progression on first-line chemotherapy.

Ramucirumab — We suggest ramucirumab plus paclitaxel for most patients with advanced or metastatic gastric or EGJ adenocarcinomas who have disease progression on or after prior treatment with fluoropyrimidine- or platinum-containing chemotherapy and retain an excellent performance status. For patients with an adequate performance status or those for whom limiting treatment-related toxicity is an important goal, ramucirumab monotherapy is an acceptable alternative.

Ramucirumab is a recombinant monoclonal antibody of the immunoglobulin G1 (IgG1) class that binds to VEGFR-2, blocking receptor activation. At least two trials have shown a survival benefit for therapy with ramucirumab, either as monotherapy (compared with best supportive care) or in combination with paclitaxel (over paclitaxel alone) in patients with previously treated, advanced gastric or EGJ adenocarcinoma [4,5]:

●In the phase III REGARD trial, 355 patients with previously treated, advanced or metastatic gastric or EGJ adenocarcinoma were randomly assigned to best supportive care plus either ramucirumab (8 mg/kg intravenously every two weeks) or placebo [4]. Although the benefits were modest, patients treated with ramucirumab had significantly better median PFS (2.1 versus 1.3 months) and overall survival (5.2 versus 3.8 months, HR 0.78, 95% CI 0.60-0.99). The rate of objective tumor response was 8 versus 3 percent, but the overall disease control rate (objective response plus stable disease) was significantly higher with ramucirumab (49 versus 23 percent). The most common adverse event (all grades) in ramucirumab-treated patients was hypertension (16 versus 8 percent in the placebo group, 8 versus 3 percent grade 3 or worse). Ramucirumab was not associated with increased bleeding, venous thromboembolism, perforation, fistula formation, or proteinuria. There were three grade ≥3 arterial thromboembolism events in the ramucirumab arm versus none with placebo. (See "Toxicity of molecularly targeted antiangiogenic agents: Non-cardiovascular effects" and "Toxicity of molecularly targeted antiangiogenic agents: Cardiovascular effects".)

●A survival benefit for ramucirumab was also demonstrated in the phase III RAINBOW trial, which compared weekly paclitaxel (80 mg/m² on days 1, 8, and 15 of each 28-day cycle) plus either ramucirumab (8 mg/kg intravenously every two weeks) or placebo in 665 patients with metastatic gastric or EGJ adenocarcinoma who had disease progression on or within four months after first-line platinum- and fluoropyrimidine-based combination therapy [80]. Median overall survival was significantly better with ramucirumab (9.6 versus 7.4 months, HR 0.807, 95% CI 0.678-0.962), as was PFS (4.4 versus 2.9 months) and the objective response rate (28 versus 16 percent). Grade 3 or worse neutropenia was more common with ramucirumab (41 versus 19 percent), but rates of febrile neutropenia were low and similar in both groups (3 versus 2 percent). Rates of grade ≥3 hypertension were 14 versus 2 percent.

A similarly designed trial conducted in China (RAINBOW-Asia) in 440 patients with previously treated advanced gastric or EGJ adenocarcinoma concluded that combined therapy with ramucirumab plus weekly paclitaxel improved median PFS (4.14 versus 3.15 months) but not overall survival (median 8.71 versus 7.92 months, HR 0.963, 95% CI 0.771-1.203) compared with weekly paclitaxel alone, and rates of grade 3 or worse neutropenia (54 versus 39 percent) and febrile neutropenia (6 versus <1 percent) were worse [81].

Largely based on the REGARD trial results, the FDA approved single-agent ramucirumab for patients with advanced or metastatic gastric or EGJ cancer and disease progression during or after prior treatment with fluoropyrimidine- or platinum-containing chemotherapy. In November 2014, the combination of ramucirumab plus paclitaxel was also approved by the FDA for treatment of advanced gastric or EGJ cancer.

Whether ramucirumab plus paclitaxel is a preferred strategy over ramucirumab monotherapy is unknown given the lack of comparator trials. The REGARD and RAINBOW trials are not directly comparable in that patients in the REGARD trial were more heavily pretreated. In our view, ramucirumab plus weekly paclitaxel is a reasonable alternative to ramucirumab monotherapy for patients who are willing to accept more treatment-related toxicity.

Other agents

●Bevacizumab – The benefit of bevacizumab, a monoclonal antibody that binds to soluble VEGF and prevents binding to VEGFR, for gastric and EGJ adenocarcinomas is uncertain. Promising results were reported in a phase II study of bevacizumab in combination with cisplatin plus irinotecan in patients with gastric or EGJ adenocarcinoma [82]; however, a survival benefit for adding bevacizumab to capecitabine plus cisplatin could not be shown in the global phase III Avastin in Gastric Cancer (AVAGAST) trial [83]. A similar lack of benefit for the addition of bevacizumab to capecitabine plus cisplatin in Asian patients was also noted in a preliminary report of the AVATAR study presented at the 2012 ASCO Gastrointestinal Cancers Symposium [84].

Until further information is available, we suggest not using bevacizumab for patients with advanced esophagogastric cancer. Biomarkers may prove useful for identifying patients with advanced gastric adenocarcinoma who might benefit from the addition of bevacizumab to cytotoxic chemotherapy, but this is not yet established [85].

●Apatinib – Apatinib (rivoceranib) is an orally active VEGFR-2 inhibitor that is approved for second-line treatment of advanced gastric cancer in China; it is not available in the United States or Europe. The data on activity in advanced gastric cancer are conflicting:

•Activity was shown in a multicenter, randomized, double-blind trial in which 270 patients in China with advanced gastric cancer and prior failure on second-line chemotherapy were randomly assigned in a 2:1 ratio to apatinib (850 mg daily) or placebo [86]. Objective response rates were low with apatinib (3 versus 0 percent with placebo), but mean overall survival was modestly but significantly prolonged (median 6.5 versus 4.7 months).

•On the other hand, activity for apatinib after failure of two prior chemotherapy regimens was not shown in a preliminary report of the placebo-controlled multinational phase III ANGEL trial [87]. The median overall survival was not significantly better with apatinib (5.78 versus 5.13 months, HR 0.93, p = 0.485).

●Regorafenib – Regorafenib is an orally active inhibitor of angiogenic (including VEGFR-1 to VEGFR-3), stromal, and oncogenic receptor tyrosine kinases. Activity was shown in a multicenter, randomized, double-blind phase II trial in which 152 patients with advanced gastric cancer and prior failure on first- or second-line chemotherapy were randomly assigned in a 2:1 ratio to regorafenib (160 mg daily) or placebo [88]. Median PFS was significantly longer with regorafenib (2.6 versus 0.9 months, HR 0.4, 95% CI 0.28-0.59), and there was a trend toward improved overall survival (median 5.8 versus 4.5 months, p = 0.47).

In our view, these data require confirmation before regorafenib can be considered a standard therapy for advanced gastric cancer.

●Sunitinib and sorafenib – Sunitinib and sorafenib are small-molecule TKIs that inhibit VEGFR-1, VEGFR-2, and VEGFR-3, as well as other tyrosine kinases; they are referred to as "multitargeted" TKIs. Early studies have shown mixed results:

•In a phase II trial of sunitinib monotherapy for second-line treatment of metastatic gastric adenocarcinoma, only 2 of 78 patients had an objective partial response (3 percent), while 25 others (32 percent) had stable disease [89]. Median overall survival was 6.8 months. Grade 3 or worse thrombocytopenia and neutropenia were reported in 35 and 30 percent, respectively.

•The combination of sorafenib (400 mg orally, twice daily, continuously), docetaxel (75 mg/m² intravenously on day 1), and cisplatin (75 mg/m² intravenously on day 1) was studied in a phase II trial of 44 chemotherapy-naïve patients with locally advanced or metastatic (80 percent) gastric or EGJ adenocarcinoma [90]. Objective responses were noted in 41 percent, and median overall survival was 13.6 months; the major grade 3 or 4 toxicity was neutropenia (64 percent).

•However, benefit for sunitinib could not be confirmed in the only randomized trial, in which sunitinib (37.5 mg every day) plus docetaxel (60 mg/m² every three weeks) was compared with docetaxel monotherapy (60 mg/m² every three weeks) for second-line treatment of patients with unresectable or metastatic gastric cancer (n = 107) [91]. Although combined therapy was associated with a significantly higher objective response rate, neither time to disease progression (the primary endpoint, 3.9 versus 2.6 months) nor overall survival (8.0 versus 6.6 months) was significantly improved.

●Aflibercept – Intravenous aflibercept, a recombinant fusion protein consisting of the VEGF-binding portions from human VEGFR-1 and VEGFR-2 fused to the fragment crystallizable (Fc) portion of human IgG1, functions as a decoy receptor that prevents intravascular and extravascular VEGF-A, VEGF-B, and placental growth factor (PlGF) from binding to their receptors. Unfortunately, benefit from the addition of aflibercept to FOLFOX could not be shown in a randomized phase II trial [92].

HER2-positive disease and continued targeting of HER2 after progression — The best treatment for HER2-positive adenocarcinomas after progression on a first-line trastuzumab-containing regimen is not established. To date, there has been no strong evidence to support continued targeting of HER2 with trastuzumab, lapatinib, or ado-trastuzumab emtansine, and we suggest not pursuing these approaches. A notable weakness in most of the studies evaluating second-line therapy targeting HER2 is that HER2 status was not reconfirmed prior to initiating the second-line HER2-directed therapy, and loss of HER2 expression could have contributed to the negative results of other HER2-targeted approaches. Some clinicians continue second-line trastuzumab in patients who are shown to retain HER2 positivity after progression on first-line trastuzumab, although this is controversial. Most recently, potential benefit has been shown for the antibody-drug conjugate fam-trastuzumab deruxtecan after progression on first-line trastuzumab, and this is a reasonable alternative approach for HER2-positive advanced adenocarcinoma, although treatment-related toxicity may be prominent.

Trastuzumab plus chemotherapy is a standard first-line treatment for human epidermal growth factor receptor 2 (HER2)-positive advanced esophagogastric cancer. (See "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Benefit of trastuzumab'.)

Several anti-HER2 therapies have been examined in the second-line setting after failure of an initial trastuzumab-containing regimen:

●Trastuzumab, lapatinib, or ado-trastuzumab emtansine – The benefit of continuing to target HER2 using trastuzumab, lapatinib, or ado-trastuzumab emtansine after progression on an initial trastuzumab-containing regimen has been addressed in the following trials, all of which failed to show benefit:

•The benefit of continued trastuzumab after progression on an initial trastuzumab-containing regimen was studied in the phase II T-ACT trial in which 91 patients refractory to trastuzumab plus a fluoropyrimidine and platinum were randomly assigned to paclitaxel with or without trastuzumab [93]. There was no benefit from trastuzumab continued beyond progression. Notably, 69 percent of the patients whose tumor tissues were available lost tumoral HER2 positivity after progression on the prior trastuzumab-containing chemotherapy. No biomarker could be identified that was associated with efficacy of trastuzumab beyond progression.

•The benefit of lapatinib, an inhibitor of both EGFR and HER2, in conjunction with weekly paclitaxel versus weekly paclitaxel for second-line therapy was evaluated in the Asian TyTAN trial, a two-phase, parallel-group phase III study that included an open-label dose optimization pilot study (n = 12) followed by a randomized trial (n = 261) in which patients with HER2-overexpressing gastric cancer were randomly assigned to paclitaxel (80 mg/m² weekly) with or without lapatinib (1500 mg once daily) [94]. The addition of lapatinib to weekly paclitaxel provided no significant benefit in terms of overall survival (median 11 versus 8.9 months) or PFS (5.4 versus 4.4 months). However, the risk of death or disease progression was significantly lower in patients with IHC 3+ tumors who were treated with lapatinib, while it was not significantly different in those with IHC 0/1+ or IHC 2+ tumors. Importantly, few patients in either arm had received prior trastuzumab (7 percent of the lapatinib patients versus 15 percent of the placebo group). As a result, whether these results can be extrapolated to populations receiving initial trastuzumab is unclear.

•A similar lack of benefit for second-line lapatinib was shown in a randomized phase II trial; of the 37 patients randomly assigned to lapatinib with or without capecitabine, 24 (65 percent) had received prior trastuzumab [95]. Only two patients (11 percent) achieved an objective partial response, both in the combined therapy group. The median time to progression was only 42 days with lapatinib monotherapy and 86 days with combined therapy.

•There was also a lack of benefit for ado-trastuzumab emtansine (an antibody-drug conjugate composed of trastuzumab, a thioether linker, and a microtubule inhibitor) compared with a taxane alone for second-line therapy in the randomized phase II/III GATSBY trial [96].

One potential reason for the lack of efficacy of these second-line therapies targeting HER2 is loss of tumoral HER2 expression, which has been documented in 16 to 69 percent of such patients [93,97-99]. However, in at least one trial, no biomarker, including HER2 overexpression, could be identified that was associated with efficacy of trastuzumab beyond progression [93].

●Fam-trastuzumab deruxtecan – Fam-trastuzumab deruxtecan is an antibody-drug conjugate composed of an anti-HER2 antibody, a cleavable tetrapeptide-based linker, and a cytotoxic topoisomerase I inhibitor; it is approved for locally advanced or metastatic HER2-positive gastric or esophagogastric junction adenocarcinoma who have received a prior trastuzumab-based regimen. (See "Systemic treatment for HER2-positive metastatic breast cancer", section on 'Fam-trastuzumab deruxtecan'.)

In contrast to the above negative trials, benefit for fam-trastuzumab deruxtecan in previously treated HER2-positive gastric cancer was suggested in the open label randomized phase II DESTINY-Gastric01 trial, in which 187 patients with confirmed HER2-positive gastric cancer that had progressed while receiving two prior therapies, including trastuzumab, were randomly assigned to fam-trastuzumab deruxtecan or clinician's choice of chemotherapy (irinotecan [89 percent] or paclitaxel [11 percent]) [100]. HER2 status was assessed on the basis of the most recent archival tissue, and not systematically reconfirmed prior to enrollment. The fam-trastuzumab deruxtecan group had a significantly higher objective response rate (51 versus 14 percent), and significantly longer median overall survival (12.5 versus 8.4 months, HR for death 0.59, 95% CI 0.39-0.88) and median PFS (5.6 versus 3.5 months, HR 0.47, 95% CI 0.31-0.71). Objective response rates seemed higher in the subgroup with the highest level of HER2 expression (3+ or higher, 58 percent versus 29 percent in the control group), but the small numbers of patients in this subgroup precludes definitive conclusions.

Nausea, malaise, diarrhea, and vomiting occurred in over 25 percent of patients receiving fam-trastuzumab deruxtecan, but were mostly mild. Serious (grade 3 or 4) treatment-related toxicity included myelosuppression (neutropenia [51 percent], anemia [38 percent], thrombocytopenia [10 percent]), and interstitial lung disease (ILD; 10 percent); there was one treatment-related death classified as pneumonia. While there were no clinical adverse events of heart failure, 8 percent had asymptomatic grade 2 decrease in left ventricular ejection fraction (table 3).

Largely based on these data, the FDA approved fam-trastuzumab deruxtecan for treatment of adults with locally advanced, or metastatic gastric or esophagogastric cancer junction adenocarcinoma who have received a prior trastuzumab-based regimen [101]. The recommended dose for gastric cancer is 6.4 mg/kg given as an IV infusion, once every three weeks until disease progression or unacceptable toxicity. The United States Prescribing Information for fam-trastuzumab deruxtecan recommends that clinicians monitor for and promptly investigate signs and symptoms including cough, dyspnea, fever, and other new or worsening respiratory symptoms, and that the drug be permanently discontinued for ≥grade 2 ILD/pneumonitis (table 4) or severe left ventricular dysfunction. Dose reduction guidelines are provided for severe myelosuppression.

SPECIAL CONSIDERATIONS DURING THE COVID-19 PANDEMIC — The COVID-19 pandemic has increased the complexity of cancer care. Important issues in areas where viral transmission rates are high include balancing the risk from delaying diagnostic evaluation and cancer treatment versus harm from COVID-19, minimizing the number of clinic and hospital visits to reduce exposure whenever possible, mitigating the negative impacts of social distancing on delivery of care, and appropriately and fairly allocating limited health care resources. Recommendations for vaccination of cancer patients and cancer care during active phases of the COVID-19 pandemic are discussed separately. (See "COVID-19: Considerations in patients with cancer".)

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 and patient selection – The major goals of systemic chemotherapy for advanced esophagogastric cancer are to provide symptom palliation, improve quality of life, and prolong survival. (See 'Goals of therapy' above.)

The criteria to select patients for second-line systemic therapy have not been established. For patients who retain an adequate performance status after progression on first-line therapy, we make this decision on a case-by-case basis, after consulting with the patient and assessing his or her overall medical condition and disease status. (See 'Patient selection' above.)

●Regimen choice – There is no standard approach to second-line therapy or beyond after failure of the first-line regimen. Performance status; comorbidity; patient preference; symptom burden; quality of life; histologic type; molecular analysis for human epidermal growth factor receptor 2 (HER2) overexpression, deficient mismatch repair (dMMR), high levels of microsatellite instability (MSI-H), and overexpression of programmed cell death ligand 1 (PD-L1); and the nature of the first-line regimen are key considerations. We prefer enrollment in a clinical trial if available. If trials are unavailable or patients are ineligible, a general approach is outlined in the algorithm (algorithm 1):

•Supportive care alone – For patients with a poor performance status, poorly controlled comorbidity, or a preference for no additional therapy, we suggest supportive care alone.

•Adenocarcinoma

-HER2-positive – The optimal second-line treatment strategy for patients with HER2-overexpressing advanced gastroesophageal adenocarcinoma is not established. There is no strong evidence to support continued targeting of HER2 with trastuzumab, lapatinib, or ado-trastuzumab emtansine, and we suggest not pursuing these approaches (Grade 2B).

Potential benefit has been shown for the antibody-drug conjugate fam-trastuzumab deruxtecan after progression on first-line trastuzumab, and this approach is reasonable, although myelosuppression and interstitial lung disease are notable toxicities. (See 'HER2-positive disease and continued targeting of HER2 after progression' above.)

Another alternative is cytotoxic chemotherapy.

-PD-L1 overexpression, dMMR/MSI-H, or high TMB – Another option for treatment at progression for patients who have advanced tumors with dMMR, MSI-H, high levels of tumor mutation burden (TMB), or high levels of overexpression of PD-L1 (combined positive score [CPS] 10 or higher) is immunotherapy with pembrolizumab (table 5), if an immune checkpoint inhibitor was not administered for front-line therapy (See 'Biomarkers and benefit from PD-1 inhibitors' above.)

Notably, In Japan, nivolumab is approved for the treatment of unresectable, advanced, or recurrent gastric cancer that has progressed after "conventional chemotherapy," without regard to specific biomarker expression. (See 'PD-1 and PD-L1 inhibitors' above.)

For patients who received an immune checkpoint inhibitor as initial therapy for advanced disease, either as monotherapy, or with cytotoxic chemotherapy, we would generally not pursue rechallenge with the same or a different checkpoint inhibitor if disease progression occurred during prior programmed cell death receptor 1 (PD-1) targeted therapy. (See "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Frontline immunotherapy'.)

Rechallenge may be considered in the context of a clinical trial, if prior anti-PD-1 immunotherapy was discontinued due to a specific toxicity that has since resolved, and possibly for patients with progression at few sites (oligoprogression) after a prior objective response or prolonged period of stable disease during initial PD-1 targeted immunotherapy.

-Cytotoxic chemotherapy – For most patients with HER2-negative gastric or gastroesophageal junction adenocarcinoma who have disease progression on or after prior treatment with fluoropyrimidine- or platinum-containing chemotherapy, are not eligible for immunotherapy or who had disease progression during front-line PD-1 targeted immunotherapy, retain an excellent performance status, have a favorable comorbidity profile, and have a preference for intensive treatment and/or have a high symptom burden or another need for a higher response rate, we suggest ramucirumab plus paclitaxel (Grade 2A). (See 'Ramucirumab' above.)

For patients with an adequate performance status and comorbidity profile or those for whom limiting treatment-related toxicity is an important goal, we suggest monotherapy rather than combination chemotherapy (Grade 2C). Options include irinotecan, weekly paclitaxel, weekly nanoparticle albumin-bound paclitaxel (nabpaclitaxel), apatinib (where available), or ramucirumab monotherapy. (See 'Cytotoxic chemotherapy' above and 'Treatments targeting VEGF' above.)

Trifluridine-tipiracil is a reasonable third-line option for patients with an adenocarcinoma that is PD-L1 negative and has proficient mismatch repair, and a fourth-line option for patients with a PD-L1-positive/dMMR adenocarcinoma. (See 'Trifluridine-tipiracil' above.)

•Squamous cell cancer

-For patients with squamous cell carcinoma (SCC) who have disease progression on or after prior treatment with fluoropyrimidine- or platinum-containing chemotherapy, are not eligible for immunotherapy or had disease progression during prior PD-1 targeted immunotherapy, retain an excellent performance status, have a favorable comorbidity profile and preference for intensive treatment and/or a high symptom burden or another need for a higher response rate, we suggest a combination regimen containing agents not used in the first-line setting (Grade 2C). Options include FOLFIRI (irinotecan plus leucovorin and short-term infusional fluorouracil [FU]) or taxane-based therapy after FOLFOX (oxaliplatin plus leucovorin and short-term infusional FU) or cisplatin plus FU. (See 'Cytotoxic chemotherapy' above.)

-For patients with an adequate performance status and comorbidity profile or those for whom limiting treatment-related toxicity is an important goal, we suggest monotherapy rather than combination chemotherapy (Grade 2C). Options include irinotecan, weekly paclitaxel, or weekly nabpaclitaxel. (See 'Cytotoxic chemotherapy' above and 'Treatments targeting VEGF' above.)

-For patients who are eligible for immunotherapy, an option for second-line therapy, regardless of PD-L1 or MMR or TMB status, is nivolumab monotherapy, or, where available, camrelizumab. (See 'Esophageal squamous cell cancer' above.)

-For patients who received an immune checkpoint inhibitor as initial therapy for advanced disease, either as monotherapy, or with cytotoxic chemotherapy, we would generally not pursue rechallenge with the same or a different checkpoint inhibitor if disease progression occurred during prior PD-1 targeted therapy. (See "Initial systemic therapy for locally advanced unresectable and metastatic esophageal and gastric cancer", section on 'Frontline immunotherapy'.)

Rechallenge may be considered in the context of a clinical trial, if prior anti-PD-1 immunotherapy was discontinued due to a specific toxicity that has since resolved, and possibly for patients with progression at few sites (oligoprogression) after a prior objective response or prolonged period of stable disease during initial PD-1 targeted immunotherapy.

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Johanna Bendell, MD, who contributed to an earlier version of this topic review.