Nodal status and advanced clinical stage as determinants of response and survival after induction chemoradiotherapy in esophageal and gastroesophageal junction cancer: a single-centre retrospective cohort study

Introduction

Esophageal and gastroesophageal junction (GEJ) cancers have historically had poor outcomes, especially in advanced stages. The introduction of neoadjuvant chemoradiotherapy, as exemplified by the CROSS trial (weekly carboplatin/paclitaxel with 41.4 Gy radiation followed by surgery), significantly improved survival for resectable esophageal/GEJ tumors. In the CROSS study, trimodality therapy achieved a 29% pathologic complete response (pCR) rate and nearly doubled median overall survival (OS) compared to surgery alone (49 vs. 24 months) (1). Five-year survival rose to ~47% with chemoradiotherapy, establishing this approach as a standard of care for stage II–III disease (2). Importantly, long-term follow-up confirms a durable benefit: the 10-year OS was 38% vs. 25% with surgery alone (3). Neoadjuvant chemoradiation markedly reduces locoregional recurrence, although distant metastases remain a predominant relapse pattern (3). This underscores that while CROSS therapy improves cure rates, systemic disease control is an ongoing challenge.

Patients with more advanced clinical stage—including cT4 tumors, extensive nodal disease (cN2–3), and limited metastatic disease (cM1)—derive some benefit from neoadjuvant chemoradiation but still face suboptimal outcomes. The original CROSS trial enrolled mostly cT1–3 tumors with limited nodal spread and excluded unresectable cT4 or distant metastases (1,2). Although patients with limited nodal involvement (cN1) were included in the CROSS trial, nodal status was not a primary stratification factor and outcomes in this subgroup remain incompletely defined. Exploratory analyses suggested clinically node-negative patients may benefit more than cN1 patients, implying limited nodal involvement may represent a higher-risk subgroup; these findings remain hypothesis-generating and are not well studied in real-world cohorts (2).

Real-world series indicate that locally invasive cT4 tumors can be downstaged and resected after chemoradiotherapy, but prognosis remains guarded. For example, in cT4 esophageal squamous carcinoma, reported 3-year OS was only ~33% overall; survival improved to ~55% among patients who completed chemoradiation followed by resection, compared to 30% in patients who did not undergo resection (4). Similarly, a heavy nodal burden (cN2–3, corresponding to ≥3 or ≥7 positive nodes) is a strong adverse prognostic factor—such patients have significantly poorer survival than those with limited nodal disease (5). Indeed, modern staging classifies extensive nodal metastasis as stage IVA, reflecting its high-risk nature (5). Finally, the presence of distant metastasis (cM1) at diagnosis generally precludes curative CROSS therapy. Patients with metastatic disease at diagnosis are generally managed with systemic therapy alone, with median survival of 12–18 months in most series, highlighting the need for more effective preoperative systemic treatments in borderline metastatic cases.

Given these challenges, recent trials have explored whether intensifying systemic therapy can improve outcomes for advanced-stage esophageal/GEJ cancers. An ongoing debate is whether modern perioperative chemotherapy alone can achieve outcomes comparable to, or better than, the CROSS approach. In adenocarcinoma, some randomized trials (e.g., Neo-AEGIS) found no significant difference in survival between neoadjuvant chemoradiation and modern perioperative chemotherapy regimens (6), whereas the more recent ESOPEC trial reported superior survival with perioperative chemotherapy alone, compared with chemoradiation (7). Collectively, these findings highlight the importance of effective systemic control, suggesting that many patients die from occult metastatic disease regardless of upfront radiation.

Considering this, integration of immunotherapy into preoperative treatment has shown promise. In the phase III MATTERHORN trial, the addition of the PD-L1 inhibitor durvalumab to perioperative FLOT chemotherapy significantly improved event-free survival, reducing the risk of progression or death by ~29% compared to FLOT alone [hazard ratio (HR) =~0.71] (8). Together, these advances—including perioperative chemo-immunotherapy and adjuvant immunotherapy strategies—are beginning to augment the traditional CROSS paradigm, aiming to further improve long-term survival in patients with advanced locoregional disease or high-risk clinical features. We therefore evaluated the real-world effectiveness of the CROSS regimen across a spectrum of baseline disease burdens: stratifying patients by nodal status (cN0 vs. cN1) and by advanced clinical features (cT4, cN2–3, or oligometastatic cM1), and assessed resection rates, pCR, and long-term survival. We stratified patients using a three-tier scheme reflecting escalating tumor burden—node-negative (group 1), limited nodal (group 2), and locally advanced or oligometastatic disease (group 3)—because nodal status and cT4/cM1 status are the dominant drivers of stage and prognosis in this disease and define the populations progressively excluded from CROSS-based trials, providing a clinically intuitive framework for examining real-world effectiveness across the risk spectrum. Although ESOPEC suggests perioperative FLOT may be superior for resectable adenocarcinoma, CROSS-based chemoradiotherapy remains in wide use, and real-world outcomes across nodal and clinical-stage strata—particularly in trial-excluded patients—remain poorly characterized; our analysis is intended to help fill this gap rather than to compare treatment regimens directly. We present this article in accordance with the STROBE reporting checklist (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0473/rc).

Methods

We conducted a retrospective cohort study of all patients with esophageal or GEJ cancer treated with neoadjuvant chemoradiation using the CROSS regimen at a Canadian regional cancer center (Windsor Regional Hospital) between January 2012 and December 2022. Patients were identified through institutional pathology and systemic therapy databases and cross-referenced with electronic medical records to ensure complete case capture. Demographic, clinical, staging, treatment, and outcome variables were abstracted into a standardized analytic dataset. To minimize selection bias, all consecutive eligible patients treated with CROSS-based chemoradiation during the study period were included, without exclusion based on clinical stage or outcome. Clinical staging was determined using pretreatment imaging and multidisciplinary tumor board review and classified according to the American Joint Committee on Cancer (AJCC) staging system. For risk stratification, patients were categorized into three mutually exclusive groups based on a priori clinical criteria:

• Group 1: cT1–3N0M0;
• Group 2: cT1–3N1M0;
• Group 3: cT4 and/or cN2–3 and/or cM1 disease.

These groupings were chosen to reflect increasing baseline disease burden and to distinguish patients who are eligible for CROSS-based therapy (groups 1 and 2) from those with locally advanced or oligometastatic presentations traditionally excluded from CROSS trials (group 3).

All patients received neoadjuvant chemoradiotherapy according to the CROSS protocol: weekly carboplatin [area under the curve (AUC) =2] and paclitaxel (50 mg/m²) administered concurrently with radiotherapy (41.4 Gy in 23 fractions). Surgical resection with curative intent was offered based on Eastern Cooperative Oncology Group (ECOG) performance status, response to therapy, disease resectability, and multidisciplinary review. Pathologic staging was recorded for patients who underwent surgical resection, and pCR was defined as ypT0N0. Patients were followed from the date of diagnosis through last contact or death, with follow-up data obtained from electronic medical records and outpatient clinic documentation. Patients alive at last contact were censored at the date of most recent follow-up.

Missing values for baseline covariates were handled descriptively in Table 1 and were not imputed. In the multivariable Cox models, ECOG performance status was modelled as three categories (0–1, 2–3, unknown) to retain all 138 patients and avoid complete-case exclusion of the 22 (16%) with missing ECOG. Variables with a substantial proportion of missing data [e.g., human epidermal growth factor receptor 2 (HER2), microsatellite stable (MSS)] were reported descriptively and not included as covariates.

The flow of patients from database identification through group allocation is summarized in Figure S1 (participant flow diagram). Of 147 patients screened, 9 were excluded: 3 who underwent surgery before or instead of neoadjuvant chemoradiotherapy, 5 who received palliative-intent or non-CROSS therapy rather than curative-intent concurrent CROSS-based chemoradiotherapy (including 1 who declined radiotherapy), and 1 whose eligibility could not be confirmed owing to insufficient documentation, yielding the final analytic cohort of 138.

The primary outcomes were OS and progression-free survival (PFS), measured from the date of diagnosis. Secondary outcomes included resection rate, and pCR.

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Research Ethics Board of Windsor Regional Hospital (No. 24-485), with a waiver of informed consent due to the retrospective design. Patient confidentiality was maintained.

Statistical analysis

Continuous variables were summarized as medians with interquartile ranges (IQRs) and categorical variables as counts with percentages. Categorical comparisons across the three prespecified clinical risk groups were performed using χ² tests or Fisher’s exact tests, as appropriate. OS and PFS were both measured from the date of diagnosis; patients alive at last contact were censored at the date of most recent follow-up. OS and PFS were estimated using Kaplan-Meier methodology and compared across clinical groups using log-rank tests. Median follow-up was calculated using the reverse Kaplan-Meier method. HRs and 95% confidence intervals (CIs) were derived using Cox proportional-hazards regression models adjusted for age, sex, and ECOG performance status, with group 1 as the reference category. Missing values for baseline covariates were handled descriptively in Table 1 and were not imputed. Variables with a substantial proportion of missing data (HER2, MSS status) were reported descriptively and not included as covariates. The proportional-hazards assumption was assessed globally and for individual covariates. In a sensitivity analysis, risk-group models were additionally adjusted for treatment era. Two-sided statistical significance was defined as P<0.05. Analyses were performed in RStudio 2025.05.1+513.

Results

Between January 2012 and December 2022, 138 consecutive patients with esophageal or GEJ cancer received CROSS-based neoadjuvant chemoradiation at Windsor Regional Hospital and formed the analytic cohort. All 138 contributed to survival analyses; the subset undergoing surgical resection formed the denominator for pathologic outcomes. Patients were classified into three prespecified clinical groups: group 1, cT1–3N0M0 (n=68); group 2, cT1–3N1M0 (n=54); and group 3, cT4 and/or N2–3 and/or M1 disease (n=16). Baseline demographic and tumor characteristics were broadly similar across the three groups, with no significant differences in age, sex, ECOG performance status, tumor location, histologic subtype, grade, or HER2 status (Table 1). Rates of surgical resection varied significantly by clinical group, with a stepwise reduction from 70.6% (48/68) in group 1 to 57.4% (31/54) in group 2 and 25.0% (4/16) in group 3 (P=0.004).

The median follow-up for the overall cohort was 56 months (95% CI: 54 to 76). During follow-up, 86 deaths occurred (39 in group 1, 37 in group 2, and 10 in group 3). Median OS was 46 months (95% CI: 31 to not estimable) in group 1, 19 months (95% CI: 16 to 59) in group 2, and 17 months (95% CI: 10 to not estimable) in group 3. The estimated 1-, 3-, and 5-year OS probabilities were 79%, 56%, and 42% in group 1; 72%, 37%, and 31% in group 2; and 54%, 38%, and 38% in group 3, respectively. In unadjusted analysis, there was a nonsignificant trend toward worse OS with increasing stage group [global log-rank χ²=4.9; 2 degrees of freedom (df); P=0.09] (Figure 1).

Figure 1 OS by clinical stage group. Kaplan-Meier estimates of OS stratified by the same three clinical risk groups: cT1–3N0M0 (n=68), cT1–3N1M0 (n=54), and cT4/N2–3/M1 (n=16). Median OS was 46, 19, and 17 months, respectively. Differences in OS showed a nonsignificant trend toward poorer outcomes with increasing disease burden (log-rank P=0.09). Tick marks indicate censored observations. OS, overall survival.

In the multivariable Cox model adjusted for age, sex, and ECOG performance status, the HRs for death were 1.57 (95% CI: 0.98 to 2.52; P=0.06) for group 2 and 1.85 (95% CI: 0.91 to 3.78; P=0.09) for group 3, as compared with group 1. ECOG performance status 2–3 (vs. 0–1) was independently associated with higher mortality (HR =1.95; 95% CI: 1.03 to 3.73; P=0.04), whereas age and sex were not. The proportional-hazards assumption was satisfied for this model (global P=0.88). pCR (defined as ypT0N0) occurred in 18.8% of patients in group 1, 12.9% in group 2, and 0% in group 3 (P=0.77). Full multivariable models for OS and PFS are presented in Table S1. Risk-group associations were essentially unchanged after additional adjustment for treatment era (for example, group 3 PFS HR =2.91 and group 2 PFS HR =1.85), indicating that outcomes were stable across the study period and were not driven by temporal shifts in practice.

Progression occurred in 102 patients (45 in group 1, 43 in group 2, and 14 in group 3). Median PFS was 34 months (95% CI: 22 to 51) in group 1, 13 months (95% CI: 9 to 19) in group 2, and 6 months (95% CI: 4 to not estimable) in group 3. The 1-, 3-, and 5-year PFS estimates were 78%, 47%, and 33% in group 1; 55%, 25%, and 22% in group 2; and 31%, 19%, and 19% in group 3, respectively. Unadjusted curves showed a clear gradient of shorter PFS with increasing stage group (global log-rank χ²=13.3; 2 df; P=0.001) (Figure 2).

Figure 2 PFS by clinical stage group. Kaplan-Meier estimates of PFS among 138 patients treated with CROSS chemoradiation, stratified into three prespecified clinical groups: cT1–3N0M0 (n=68), cT1–3N1M0 (n=54), and cT4/N2–3/M1 (n=16). Median PFS was 34, 13, and 6 months, respectively. Survival curves diverged significantly across groups (log-rank P=0.001). Tick marks indicate censored observations. PFS, progression-free survival.

In the adjusted Cox model, PFS was significantly worse in both higher-risk groups: the HRs for progression or death were 1.86 (95% CI: 1.20 to 2.89; P=0.006) for group 2 and 2.84 (95% CI: 1.54 to 5.25; P<0.001) for group 3, as compared with group 1. ECOG performance status 2–3 remained independently prognostic for shorter PFS (HR =2.05; 95% CI: 1.14 to 3.69; P=0.02), whereas age and sex were not significantly associated with outcome. There was some evidence of nonproportionality for age (P=0.02), but the global test indicated no major violation of the proportional-hazards assumption (P=0.14), and the assumption was met for the primary risk-group comparisons.

Among patients with a documented site of first relapse (n=77), recurrence involved a distant (systemic) site in 83% (64/77), whereas isolated locoregional recurrence accounted for only 17% (13/77), consistent with a predominantly systemic pattern of failure.

Discussion

In this real-world cohort of patients with esophageal and GEJ cancer treated with CROSS-based neoadjuvant chemoradiation, we observed clear prognostic separation based on baseline nodal status and clinical disease stage. Patients with node-negative disease at presentation (cT1–3N0M0) demonstrated substantially better long-term outcomes than those with node-positive or T4/oligometastatic disease. Although the original CROSS trial enrolled only resectable cT1N1 or T2–3N0–1 tumors and excluded T4, cN2–3, or oligometastatic M1 cases, these higher-risk presentations are common in contemporary practice. Our study provides insight into the effectiveness of CROSS-based therapy in these underrepresented, high-risk populations.

PFS showed the most pronounced gradient across groups. Median PFS declined from 34 months in node-negative disease to 13 months in cN1 tumors and 6 months in cT4/cN2–3/cM1 presentations. The multivariable Cox model confirmed this pattern, with significantly higher hazards of progression in both group 2 (HR =1.86) and group 3 (HR =2.84) compared with node-negative patients. OS demonstrated a similar trend: median OS was 46 months for node-negative disease, compared with 19 and 17 months for groups 2 and 3, respectively. Adjusted analyses suggested a stepwise increase in mortality risk, though statistical significance was not reached, likely reflecting limited sample size. For the highest-risk group, the adjusted HR for death was 1.85 (95% CI: 0.91–3.78; P=0.09); this estimate is imprecise, and the non-significant result should be regarded as hypothesis-generating. Notably, median follow-up exceeded 4.5 years, ensuring robust capture of long-term outcomes.

pCR rates also differed across groups, mirroring clinical risk profiles. The pCR rate was 19% among resected low-risk patients, modestly lower at 13% in cN1 tumors, and 0% in those with cT4/cN2–3/cM1 disease. Although pCR differences were not statistically significant, the absolute absence of pCR in the highest-risk group underscores the relative biologic resistance of these tumors to CROSS-based therapy. These pCR rates contrast with the 29% reported in the original CROSS trial (1) and align more closely with contemporary real-world series, which consistently show lower pCR rates, particularly in more advanced disease.

The poor outcomes observed in the cN1 subset in our cohort add to a growing body of evidence questioning the sufficiency of neoadjuvant chemoradiation for node-positive adenocarcinoma. In the long-term results reported by Shapiro et al., exploratory subgroup analyses demonstrated a substantial survival benefit of CROSS-based therapy in patients with clinically node-negative disease, whereas no statistically significant survival advantage was observed among patients with cN1 tumors (2). Although these analyses were not powered to formally test effect modification by nodal status, they raise the possibility that limited nodal involvement represents a biologically higher-risk subgroup with attenuated responsiveness to local-dominant therapy (2). The ESOPEC trial demonstrated a survival advantage for perioperative FLOT chemotherapy over the CROSS regimen, suggesting that systemic intensification is critical for addressing micrometastatic disease (7). Our findings support this paradigm shift: early distant relapses driving the poor PFS in our cN1 and N2–3 cohorts imply that the primary threat to these patients is systemic rather than locoregional. Consequently, for patients with any clinical evidence of nodal involvement, a shift toward perioperative chemotherapy (as in the ESOPEC trial) (7) or the addition of immune checkpoint inhibition—as explored in the MATTERHORN trial with durvalumab (8)—may be biologically necessary to improve cure rates. These cN1 observations should nonetheless be interpreted cautiously: they derive from a modest, single-centre cohort, the adjusted OS differences did not reach statistical significance, and the apparent convergence of cN1 with advanced-stage outcomes requires prospective confirmation.

An important and clinically relevant question concerns whether oligometastatic or stage IV-limited disease behaves distinctly from bulky stage III tumors when treated with definitive chemoradiation. In our exploratory analyses within the high-risk cohort, outcomes of patients with cM1 or cN3 disease appeared similar to those with cT4 or cN2 presentations, with uniformly poor pCR rates and short PFS. Although the sample size limits definitive conclusions, our findings parallel observations from small retrospective studies suggesting that locoregionally aggressive and oligometastatic disease may share similarly poor responsiveness to neoadjuvant chemoradiation, despite heterogeneity in metastatic patterns.

Similarly, for well-selected patients with oligometastatic disease, integration of systemic therapy with ablative local treatments-including stereotactic radiation or targeted surgical approaches—has achieved durable survival in a minority of patients (9,10). These multidisciplinary approaches may represent a rational alternative to CROSS alone, especially given the poor outcomes observed in our high-risk cohort.

This study has several limitations. The retrospective nature of the study introduces the possibility of unmeasured confounding and variation in staging, treatment selection, and follow-up intensity. The cohort was composed predominantly of adenocarcinoma and GEJ tumors, reflecting contemporary North American epidemiology, but limiting the applicability of these findings to squamous cell carcinoma or proximal esophageal cancers. In addition, the number of patients with T4 or oligometastatic disease was modest, restricting the precision of subgroup analyses. Nonetheless, the study spans more than a decade of real-world practice, includes a heterogeneous population seldom represented in clinical trials, and provides long-term outcome data with extended follow-up, allowing an assessment of durable disease control and late relapse patterns. Our dataset did not capture postoperative complications, so we could not assess whether surgical morbidity contributed to the survival differences. Site-of-relapse information, although not recorded for every patient, was available for 77 patients and showed a distant (systemic) component as the first site of relapse in 83%, supporting—though not definitively proving—the inference that failure after CROSS-based therapy is predominantly systemic. Molecular data were also substantially incomplete (HER2 unknown in 39% and MSS status unknown in 88% of patients), so molecular subgroups were not emphasized and these missingness rates preclude firm biomarker inferences. Finally, with only 16 patients in group 3, comparative estimates for this group are statistically unstable and should be considered exploratory and hypothesis-generating throughout.

Although single-centre data limit broad extrapolation, our cohort reflects contemporary North American real-world practice: adenocarcinoma-predominant, GEJ-biased, treated at a regional cancer centre with multidisciplinary review. The findings are likely generalisable to similar academic and community settings using the CROSS regimen but may not extend to squamous-predominant populations or practices where perioperative chemotherapy is the dominant approach.

Conclusions

In this real-world cohort, greater baseline clinical disease burden following CROSS-based chemoradiation was associated with progressively worse outcomes: pCR was absent in the highest-risk group, resection rates declined stepwise, and PFS was significantly shorter with increasing nodal and clinical stage. Notably, outcomes in cN1 disease approached those of cT4/cN2–3/cM1 disease, suggesting that even limited nodal involvement may mark a biologically aggressive subgroup inadequately served by local-dominant therapy. These findings support a shift toward systemic-intensified perioperative strategies, including perioperative chemotherapy and chemo-immunotherapy, in node-positive and advanced-stage esophageal/GEJ cancer. Prospective studies are needed to define optimal treatment pathways and clarify the role of chemo-immunotherapy in improving long-term survival for these high-risk populations.

Acknowledgments

The authors thank all staff and colleagues who contributed to this project.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0473/rc

Data Sharing Statement: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0473/dss

Peer Review File: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0473/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0473/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Research Ethics Board of Windsor Regional Hospital (No. 24-485), and individual consent for this retrospective analysis was waived. Patient confidentiality and anonymity were strictly maintained.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. van Hagen P, Hulshof MC, van Lanschot JJ, et al. Preoperative chemoradiotherapy for esophageal or junctional cancer. N Engl J Med 2012;366:2074-84. [Crossref] [PubMed]
2. Shapiro J, van Lanschot JJB, Hulshof MCCM, et al. Neoadjuvant chemoradiotherapy plus surgery versus surgery alone for oesophageal or junctional cancer (CROSS): long-term results of a randomised controlled trial. Lancet Oncol 2015;16:1090-8. [Crossref] [PubMed]
3. Eyck BM, van Lanschot JJB, Hulshof MCCM, et al. Ten-Year Outcome of Neoadjuvant Chemoradiotherapy Plus Surgery for Esophageal Cancer: The Randomized Controlled CROSS Trial. J Clin Oncol 2021;39:1995-2004. [Crossref] [PubMed]
4. Gao LR, Li C, Han W, et al. Survival benefit of surgery in patients with clinical T4 esophageal cancer responded to neoadjuvant chemoradiotherapy/radiotherapy. Int J Radiat Oncol Biol Phys 2022;114:e159.
5. Olthof PB, Noordman BJ, Henckens SPG, et al. The dilemma of ypN3 esophageal cancer. Eur J Surg Oncol 2025;51:110224. [Crossref] [PubMed]
6. Reynolds JV, Preston SR, O’Neill B, et al. Trimodality therapy versus perioperative chemotherapy in the management of locally advanced adenocarcinoma of the oesophagus and oesophagogastric junction (Neo-AEGIS): an open-label, randomised, phase 3 trial. Lancet Gastroenterol Hepatol 2023;8:1015-27. [Crossref] [PubMed]
7. Hoeppner J, Brunner T, Schmoor C, et al. Perioperative Chemotherapy or Preoperative Chemoradiotherapy in Esophageal Cancer. N Engl J Med 2025;392:323-35. [Crossref] [PubMed]
8. Janjigian YY, Al-Batran SE, Wainberg ZA, et al. Perioperative Durvalumab in Gastric and Gastroesophageal Junction Cancer. N Engl J Med 2025;393:217-30. [Crossref] [PubMed]
9. Kroese TE, Jorritsma NKN, van Laarhoven HWM, et al. Stereotactic radiotherapy or metastasectomy for oligometastatic esophagogastric cancer: A nationwide population-based cohort study. Clin Transl Radiat Oncol 2022;37:109-15. [Crossref] [PubMed]
10. Franceschini D, Teriaca MA, Di Cristina L, et al. Stereotactic radiation therapy for oligometastatic esophagogastric adenocarcinoma: outcome and prognostic factors. Br J Radiol 2023;96:20220771. [Crossref] [PubMed]