Esophageal Cancer Recurrent and Metastatic Disease

After curative esophagectomy for esophageal cancer , recurrence is common and the pattern, local, regional, or distant, directs subsequent management. Among 1626 patients in a large single-institution series, 595 (37%) developed recurrence: distant/systemic in 435 (27%), regional in 125 (7.7%), and local in 35 (2.2%) ^[4]B3b. The ENSURE multicenter study of 3299 patients with esophageal adenocarcinoma reported 1310 recurrences (40%), with first recurrence at multiple distant sites (37.3%), a single distant site (33.7%), or isolated local (14.4%) ^[15]D5. In a cohort of 1147 patients, 38% recurred, distributed as distant (55%), locoregional (28%), or both (17%) ^[5]C4.

Timing of Recurrence

The highest risk period is the first two years after surgery. The recurrence rate decreases from 27 per 100 person-years in posttreatment year 1 to 4 per 100 person-years by year 6 ^[5]C4. Among patients who received neoadjuvant therapy, the rate in the first 2 years was significantly higher than in those who did not (35 vs 14 per 100 person-years, p<0.001) ^[5]C4. More than half of isolated brain metastases are diagnosed within the first year postoperatively ^[1]B2b.

Site-Specific Patterns

The site of recurrence carries prognostic significance. Patients with isolated lung oligometastases have a more indolent course, with median overall survival of 2.41 years (95% CI, 1.58-3.31), compared with 0.95 years for brain or liver oligometastases ^[12]D5. After adjustment, the hazard ratio for brain vs lung recurrence was 4.48 (95% CI, 2.24-8.99) and for liver vs lung 2.94 (95% CI, 1.48-5.82) ^[12]D5. The ENSURE study confirmed that liver-only recurrence appears earlier (median 9.0 months) and carries worse postrecurrence survival (8.3 months), whereas local-only recurrence appears later (17.8 months) and has better survival (15.9 months) ^[15]D5.

Isolated brain recurrence occurs in 2% of patients, with median survival of 0.95 years (95% CI, 0.6-1.5) ^[1]B2b. Pathologic complete response to neoadjuvant therapy is associated with longer survival after brain recurrence (median 1.56 vs 0.66 years, p=0.019) ^[1]B2b.

Predictors of Recurrence Pattern

Predictors of isolated nodal recurrence include neoadjuvant chemoradiation with total dose <45 Gy (HR 3.5, 95% CI 1.7-7.3), pathologic node-positive disease (HR 1.9, 95% CI 1.3-3.0), and lymphovascular invasion (HR 1.6, 95% CI 1.0-2.5) ^[4]B3b. Increasing age (HR 0.97 per year) and a higher number of excised lymph nodes (HR 0.98 per node) are independently associated with decreased risk of regional recurrence ^[4]B3b. The systemic immune-inflammation index (SII) at the time of recurrence diagnosis is an independent predictor of survival; a cutoff of 500 stratifies patients into low-risk (1-year OS 84.9%) and high-risk (1-year OS 28.8%) groups (HR 2.833, 95% CI 1.555-5.161) ^[10]C4.

Clinical Implications

These patterns inform surveillance strategy. The yield of surveillance imaging diminishes after the sixth year ^[5]C4, and patients with lung-only recurrence may benefit from aggressive local therapy given their more favorable prognosis ^[12]D5. The next section details the workup of suspected recurrence, including the role of imaging, endoscopy, and biopsy.

Pearl: Site of recurrence carries prognostic significance: lung oligometastases have better survival (median 2.41 years) than liver or brain (0.95 years each).

Clinicians must maintain a high index of suspicion for recurrence, given that up to 39% of patients recur after esophagectomy ^[3]B3b. The workup aims to confirm the presence, site, and extent of disease, and to identify actionable molecular alterations. A stepwise approach integrating clinical assessment, imaging, laboratory markers, and tissue sampling is recommended.

Clinical Suspicion and History

Recurrence may be detected through surveillance imaging or symptom-driven evaluation. In a cohort of 225 patients, 39.1% recurred, and detection rates were similar between surveillance imaging (SI) and history and physical examination (HPE) alone (40.6% vs 37.9%, P =.68) ^[3]B3b. However, patients whose recurrence was detected by SI had longer median overall survival (23 months vs 16 months) ^[3]B3b. Symptoms prompting evaluation include dysphagia, odynophagia, weight loss, pain, cough, or neurologic deficits.

Imaging Surveillance

Chest and abdominal CT is the most common imaging modality, detecting 45% of asymptomatic recurrences in a large series of 1147 patients ^[5]C4. FDG-PET/CT may be used for suspected distant recurrence when CT is equivocal, though its performance in this setting is not reported in the available evidence. Brain MRI should be considered in patients with pathologic complete response (PCR) after neoadjuvant therapy, as isolated brain metastasis occurs in 2% of patients after esophagectomy and may represent true isolated recurrence in PCR patients ^[1]B2b. Surveillance endoscopy has limited value, detecting only 1% of asymptomatic local recurrences ^[5]C4.

Laboratory and Biomarker Assessment

The systemic immune-inflammation index (SII) is an independent prognostic factor for patients with recurrent esophageal cancer . The SII is calculated as: platelet count (×10³/mm³) × neutrophil count (cells/mm³) / lymphocyte count (cells/mm³). Using a cutoff of 500, patients were classified as SII-low (n = 36) or SII-high (n = 58). The 1-year OS was 84.9% in the SII-low group vs 28.8% in the SII-high group (P < 0.001); the SII remained an independent prognostic factor on multivariate analysis (HR = 2.833, 95% CI: 1.555-5.161, P < 0.001) ^[10]C4. The SII-high group also had lower rates of first-line treatment introduction (63.8% vs 86.1%, P = 0.019) ^[10]C4. The REEC model (predicting recurrence and prognosis) uses alcohol consumption, TNM classification, number of lymph node station metastases, and number of lymph node metastases. At a cutoff of 1.095, the REEC had an AUC of 0.68 (P < 0.001) in the exploratory group and 0.65 (P < 0.001) in the validation group, with sensitivity of 76.68% and specificity of 51.18% ^[8]B3b.

Diagnostic Test Performance

The table below summarizes the performance of the REEC model for predicting recurrence.

Biopsy Confirmation

Tissue biopsy of suspicious lesions should be obtained whenever feasible to confirm recurrence and to test for actionable molecular alterations. Although the available evidence does not report specific protocols, biopsy is considered standard practice to distinguish recurrence from benign changes and to guide targeted therapy or immunotherapy. Cytologic or histologic confirmation is especially important for solitary lesions being considered for local aggressive treatment.

Pearl: In patients with pathologic complete response after neoadjuvant therapy, a new brain lesion should prompt brain MRI and consideration of resection, as these patients may have isolated brain metastasis with a better prognosis (median OS 1.56 vs 0.66 years, P =.019) ^[1]B2b.

Once recurrence is confirmed, the site and extent of disease, local, regional, or distant, determine the therapeutic approach. Local-regional recurrence includes disease at the anastomosis or tumor bed (local) and involvement of regional lymph nodes in the mediastinum, supraclavicular fossa, or celiac axis (regional). This distinction carries prognostic and therapeutic implications, as isolated local-regional recurrence may be amenable to definitive salvage therapy.

Incidence and Risk Factors

In large surgical series, local-regional recurrence occurs in 25-35% of patients who recur after esophagectomy . Among 1626 patients at a single institution, 7.7% developed isolated regional nodal recurrence and 2.2% had local recurrence alone ^[4]B3b. The ENSURE study of 3299 patients found that 14.4% of first recurrences were isolated local, 33.7% were single-site distant, and 37.3% were multisite ^[15]D5. The median time to recurrence is approximately 9.9 months (interquartile range 5.4-18.8 months) ^[22]B3b. Local-only recurrence tends to present later (median 17.8 months), whereas liver-only recurrence appears earlier (9.0 months) ^[15]D5.

Risk factors for regional nodal recurrence include neoadjuvant chemoradiation with a total radiation dose <45 Gy (HR 3.5, 95% CI 1.7-7.3), pathologic node-positive disease (HR 1.9, 95% CI 1.3-3.0), and lymphovascular invasion (HR 1.6, 95% CI 1.0-2.5) ^[4]B3b. Increasing age and a higher number of excised lymph nodes are independently associated with lower risk of regional recurrence ^[4]B3b. Patients who received neoadjuvant therapy have a higher early recurrence rate, 35 per 100 person-years in the first 2 years versus 14 per 100 person-years for those who did not ^[5]C4.

Detection

More than half of recurrences are detected by symptoms (50%), with 45% discovered on routine CT and only 1% by surveillance endoscopy ^[5]C4. Surveillance imaging after esophagectomy is not associated with a higher detection rate but is associated with improved overall survival once recurrence is identified (median 23 months with imaging vs 16 months with history and physical examination alone) ^[3]B3b.

Treatment Options

For patients with isolated local-regional recurrence, definitive therapy can prolong survival. A series of 27 patients with isolated nodal or distant metastases treated with curative intent reported a median overall survival of 25.2 months from recurrence, with 3-year estimated survival of 33% ^[21]C4. Among patients with regional nodal recurrence, those receiving a combination of local and systemic therapies had significantly better survival than those treated with systemic therapy alone (P < 0.001) ^[4]B3b.

Salvage surgery is feasible for select patients. Resection of solitary mediastinal lymph node recurrence achieved R0 resection in 5 of 6 patients, with a median overall survival of 43 months (range 16-82 months) ^[18]D5. Definitive chemoradiation is also effective and, importantly, its efficacy is not diminished by prior neoadjuvant chemotherapy, even when the same agents are used ^[9]B3b. For anastomotic recurrence, particularly in the gastric tube, endoscopic resection can be performed, though it carries a higher bleeding rate (17.6%) and requires lifelong surveillance for metachronous lesions ^[13]D5.

Prognosis and Prognostic Factors

Local-only recurrence carries a more favorable prognosis than distant recurrence. Median postrecurrence survival is 15.9 months for local-only recurrence, compared with 10.4 months for lung-only and 8.3 months for liver-only recurrence ^[15]D5. The systemic immune-inflammation index (SII), calculated as platelet count × neutrophil count / lymphocyte count, is an independent prognostic factor. With a cutoff of 500, the 1- and 3-year overall survival rates were 84.9% and 44.7% in the SII-low group versus 28.8% and 13.1% in the SII-high group (HR 2.833, 95% CI 1.555-5.161) ^[10]C4. Other factors associated with worse survival include COX-2 expression after neoadjuvant chemoradiotherapy ^[17]D5 and a higher Charlson comorbidity index ^[14]D5.

Pearl: For patients with isolated local-regional recurrence, especially single-station nodal disease, aggressive definitive therapy with surgery or chemoradiation can achieve prolonged survival (median >2 years) and should be pursued in appropriate candidates.

When recurrence extends beyond the locoregional field, distant metastatic disease becomes the dominant pattern. Among patients who recur after curative-intent therapy, distant metastases account for 49-55% of first recurrences, with the remainder being locoregional or mixed ^[5]C4[22]B3b. The median time to recurrence is approximately 9.9 months (interquartile range, 5.4-18.8 months), and most recurrences occur within the first 2 years ^[5]C4[22]B3b.

Patterns and Timing by Site

The site of distant metastasis significantly influences presentation and outcome. The ENSURE multicentre study of 1,310 patients with recurrence after esophagectomy for adenocarcinoma reported that liver-only recurrence appears earliest (median 9.0 months) and carries the worst prognosis (median postrecurrence survival 8.3 months), whereas lung-only recurrence emerges later (median 15.2 months) and has a more favorable course (median postrecurrence survival 10.4 months) ^[15]D5. Multiple distant sites at first recurrence are common (37.3%) and portend the poorest survival (median 7.4 months) ^[15]D5. Isolated brain metastasis occurs in 2% of patients; median overall survival after brain recurrence is 0.95 years (95% CI, 0.6-1.5 years), but patients with pathologic complete response to neoadjuvant therapy fare better (median 1.56 vs 0.66 years, P = 0.019) ^[1]B2b.

Prognostic Factors at Diagnosis

The Systemic Immune-Inflammation Index (SII), calculated as platelet count × neutrophil count / lymphocyte count, is an independent prognostic factor for recurrent esophageal cancer ^[10]C4. Using a cutoff of 500, patients with SII ≥ 500 have 1-year and 3-year overall survival rates of 28.8% and 13.1%, respectively, compared with 84.9% and 44.7% in those with SII < 500 (HR 2.833, 95% CI 1.555-5.161, P < 0.001) ^[10]C4. Higher SII also correlates with lower rates of first-line treatment introduction (63.8% vs 86.1%, P = 0.019) ^[10]C4. Other adverse prognostic factors include advanced clinical stage (Stage IV vs I: HR 21.93, P < 0.001), male sex, and Charlson-Deyo comorbidity score ≥ 2 ^[6]B3b. Prior neoadjuvant chemotherapy worsens survival when chemotherapy is given for distant recurrence (P = 0.028), whereas chemoradiotherapy for local recurrence is not similarly affected ^[9]B3b.

Systemic Therapy Approach

First-line systemic therapy for distant metastatic disease consists of a platinum/fluoropyrimidine doublet (e.g., 5-fluorouracil 800 mg/m² days 1-5 plus cisplatin 80 mg/m² day 1) with or without an immune checkpoint inhibitor ( nivolumab or pembrolizumab ) ^[10]C4. Second-line options include taxanes ( paclitaxel , docetaxel ) or checkpoint inhibitor monotherapy if not used in the first line ^[10]C4. Dietary factors may influence prognosis: higher folate intake is associated with reduced esophageal cancer-specific mortality in squamous cell carcinoma (HR 0.41, 95% CI 0.25-0.69), while alcohol consumption increases all-cause mortality (HR 1.29, 95% CI 1.07-1.55) ^[2]B2a.

Detection and Surveillance

Routine surveillance imaging (CT chest/abdomen) identifies 45% of recurrences, while half are detected by symptoms ^[5]C4. Although surveillance imaging does not improve the detection rate compared with history and physical examination alone, it is associated with longer overall survival after recurrence (median 23 months vs 16 months), suggesting that earlier detection may permit more effective salvage therapy ^[3]B3b.

Selected patients with limited distant metastases, particularly those with a long disease-free interval and single-site disease, may be candidates for definitive local therapy (surgical resection or chemoradiotherapy), which can achieve median survival exceeding 2 years ^[21]C4. This approach is further explored in the next section on oligometastatic disease.

Pearl: The SII, calculated from routine blood counts, stratifies prognosis and treatment initiation rates in patients with distant metastatic recurrence; a cutoff of 500 identifies a high-risk group with 1-year survival of only 28.8% ^[10]C4.

Although most recurrences after esophagectomy are widespread, a subset of patients present with limited metastatic burden, oligometastasis, that may be amenable to definitive local therapy. Oligometastasis represents an intermediate state between localized and systemic disease; the goal of metastasis-directed therapy is to achieve long-term control or even cure.

Site-Specific Prognosis

The presentation and prognosis of oligometastasis differ markedly by organ site. In a cohort of 104 patients with isolated solid organ recurrence after R0 esophagectomy, the median overall survival from recurrence was:

Lung oligometastases follow a more indolent course, with 80% of liver, 51% of brain, and 44% of lung recurrences occurring within the first 12 months after esophagectomy ^[12]D5. Patients with isolated lung recurrence may therefore benefit from aggressive local intervention. The ENSURE study confirms that liver-only recurrence carries the worst prognosis (median post-recurrence survival 8.3 months), whereas lung-only and local-only recurrences fare better (10.4 and 15.9 months, respectively) ^[15]D5.

Patient Selection for Definitive Local Therapy

Appropriate candidates for metastasis-directed therapy are those with:

• A single site of recurrence (single nodal station or solitary solid organ metastasis)
• A long disease-free interval (≥12 months)
• Good performance status and adequate organ function
• No evidence of widespread disease on restaging

In a retrospective series of 27 patients with isolated recurrence treated with curative intent, median overall survival from recurrence was 25.2 months (3-year estimated survival 33.0%) ^[21]C4. A long disease-free interval and recurrence limited to single nodal stations predicted longer survival. Similarly, patients with isolated regional nodal recurrence treated with combined local and systemic therapy had better survival than those receiving systemic therapy alone (P<0.001) ^[4]B3b.

Treatment Modalities

• Surgical resection: Liver resection for solitary or oligofocal metastases, combined with neoadjuvant and adjuvant chemotherapy, is feasible and can yield long-term disease-free survival (reported at 22 and 92 months in two patients) ^[11]C4. Resection of solitary mediastinal lymph node recurrence after esophagectomy can achieve R0 resection with acceptable morbidity and a median survival of 43 months (range 16-82 months) ^[18]D5.
• Definitive chemoradiation: For patients who are not surgical candidates or who have recurrences at sites not amenable to resection, definitive chemoradiation provides similar survival outcomes ^[21]C4.
• Isolated brain metastasis: In patients who achieved pathologic complete response after neoadjuvant therapy, brain oligometastasis may represent true isolated recurrence rather than the first sign of widespread disease; these patients have a median survival of 1.56 years after brain recurrence, compared with 0.66 years in those with residual nodal disease (P=0.019) ^[1]B2b.

Rationale for Aggressive Local Therapy

The site-specific differences in outcome argue against a uniform palliative approach. Lung oligometastases, in particular, have a more indolent natural history and may be optimal targets for stereotactic body radiotherapy or metastatectomy. Conversely, liver and brain oligometastases require rapid systemic therapy, though local therapy may still provide palliative benefit. The key is careful restaging with PET-CT and brain imaging to confirm the absence of additional disease before proceeding with definitive local treatment.

Pearl: For patients with isolated lung oligometastasis after esophagectomy, median survival exceeds 2 years with aggressive local therapy, supporting a paradigm shift from palliative-only care to potentially curative intervention in this selected subgroup.

Even for patients with oligometastatic disease, long-term survival is the exception rather than the rule. The prognosis after recurrence of esophageal cancer is uniformly poor, with median overall survival (OS) of 11.2 months (95% CI 5.2-17.2 months) in one modern cohort ^[10]C4, and 0.95 years (approximately) in another ^[1]B2b. The risk of recurrence is highest in the first 2 years after esophagectomy , with a rate of 27 per 100 person-years in year 1 declining to 4 per 100 person-years by year 6; among patients who received neoadjuvant therapy, the rate in the first 2 years is higher (35 vs 14 per 100 person-years) ^[5]C4.

Prognostic Factors

Several clinical and laboratory parameters stratify prognosis (Table 1). The systemic immune-inflammation index (SII), calculated as platelet count × neutrophil count / lymphocyte count, is an independent prognostic factor. Using a cutoff of 500, the 1-year OS rate was 84.9% in the SII-low group vs 28.8% in the SII-high group (HR 2.833, 95% CI 1.555-5.161) ^[10]C4. The REEC index, incorporating alcohol consumption, TNM classification, and lymph node status, predicts recurrence with a sensitivity of 76.68% and specificity of 51.18% at a cutoff of 1.095 ^[8]B3b. COX-2 expression correlates with tumor recurrence and disease-free survival (P = 0.0073), though not overall survival ^[17]D5.

Site-Specific Prognosis

Site of recurrence significantly influences survival. In the ENSURE multicenter study, median postrecurrence survival was 15.9 months for local-only recurrence, 10.4 months for lung-only, 8.3 months for liver-only, and 7.4 months for multiple sites ^[15]D5. Lung oligometastasis has a more indolent course (median OS 2.41 years) compared with brain (0.95 years) or liver (0.95 years) ^[12]D5. Among patients with isolated brain recurrence, those with pathologic complete response (pCR) to neoadjuvant therapy have better survival (1.56 vs 0.66 years, P = 0.019) ^[1]B2b.

Modifiable Risk Factors

Pre-diagnosis dietary folate intake is associated with reduced esophageal cancer-specific mortality in ESCC (HR 0.41, 95% CI 0.25-0.69), while alcohol consumption increases all-cause mortality (HR 1.29, 95% CI 1.07-1.55) ^[2]B2a. In patients with pCR after neoadjuvant therapy, older age, male gender, Charlson-Deyo comorbidity score ≥2, and clinical stage IV independently predict worse OS ^[6]B3b.

Pearl: The SII at the time of recurrence diagnosis is a simple, widely available tool that stratifies patients into widely divergent prognostic groups, a cutoff of 500 identifies those with a 1-year survival of 28.8% vs 84.9%, and can guide decisions about treatment intensity and palliative care referral ^[10]C4.

Part of the Esophageal Cancer family. Cross-cutting management is split across dedicated child pages:

• Esophageal Cancer , diagnostic page (definition, epidemiology, staging, biomarkers, prognosis)
• Esophageal Cancer Surgical Management , operations by stage, fertility-sparing options, sentinel node mapping, adjuvant triggers (Sedlis / Peters)
• Esophageal Cancer Radiation Management , EBRT + image-guided brachytherapy + concurrent chemoradiation, dose / fractionation, OAR constraints
• Esophageal Cancer Systemic Therapy , concurrent / adjuvant / metastatic chemotherapy, targeted therapy, immune checkpoint inhibitors
• Esophageal Cancer Palliative Care , early integration, symptom management, palliative procedures, end-of-life care
• Esophageal Cancer Surveillance and Follow-up , post-treatment surveillance schedule, late toxicity, survivorship, patient counselling

Pearl: Use these links to hop between management modalities; the parent Esophageal Cancer page carries diagnosis + staging that informs every decision here.