Residual Tumor Burden After Immunotherapy Predicts Survival in Esophageal Cancer

For patients with locally advanced esophageal squamous cell carcinoma, the rise of neoadjuvant immunochemotherapy has transformed what was once one of the most lethal malignancies of the upper gastrointestinal tract into a disease where meaningful tumor shrinkage before surgery is now routine. Yet the treatment does not work equally well for everyone, and a new real-world study from Fujian Cancer Hospital in China has mapped out, in unusually granular detail, what happens to the substantial fraction of patients whose tumors do not completely disappear under the microscope before they go to the operating table. The findings, published in BMC Cancer, suggest that the degree of residual tumor and the status of lymph nodes after treatment carry powerful prognostic weight, and that these measures could reshape how oncologists tailor postoperative care.

The study focused on a group that has received comparatively little attention in the immunotherapy era: patients who achieved a non-pathological complete response, abbreviated non-pCR, after receiving neoadjuvant immunochemotherapy, or NICT. This treatment strategy combines chemotherapy with immune checkpoint inhibitors, drugs that unleash the body’s own T cells against cancer by blocking inhibitory receptors such as PD-1. When NICT works spectacularly, the entire tumor is eradicated before surgery, a state called pathological complete response, and these patients enjoy excellent long-term outcomes. But when cancer cells persist in the resected specimen, the prognosis becomes far less certain, and clinicians have lacked clear guidance on how to stratify risk within this heterogeneous population.

To address that gap, the research team led by Hao He, Pengqiang Gao, and Junpeng Lin, with Shuoyan Liu and Feng Wang as corresponding authors, retrospectively analyzed 178 patients with thoracic esophageal squamous cell carcinoma who underwent R0 esophagectomy, meaning a complete surgical resection with clear margins, between January 2020 and December 2023 after receiving NICT. All patients had residual disease on final pathology. The investigators divided them into two groups based on tumor regression grade, a standardized pathological scoring system that quantifies how much of the original tumor was destroyed by preoperative therapy. The TRGa group comprised 84 patients with residual tumor occupying between zero and fifty percent of the original tumor bed, while the TRGb group comprised 94 patients in whom more than half of the tumor remained viable.

The pathological differences between these two groups were striking. Patients in the TRGb group, whose tumors had responded poorly to the preoperative immunochemotherapy, carried far more aggressive biological features at the time of surgery. Deep tumor invasion into the esophageal wall, classified as ypT3-4 stage, was present in 77.7 percent of TRGb patients compared with just 26.2 percent of TRGa patients, a difference that was highly statistically significant. Perineural invasion, in which cancer cells wrap around nerve fibers and gain a conduit for spread, occurred in 44.7 percent of the poor responders versus 6.0 percent of the better responders. Lymphovascular invasion, the presence of tumor cells inside lymphatic or blood vessels and a well-known harbinger of metastasis, was documented in 63.8 percent of TRGb patients compared with 27.4 percent of TRGa patients. Each of these comparisons reached a P value below 0.001, indicating that the association is very unlikely to be a statistical fluke.

These pathological differences translated directly into survival outcomes. The TRGb group demonstrated significantly inferior disease-free survival, the length of time after surgery without any evidence of cancer returning, as well as inferior overall survival. In the language of survival analysis, the curves for the two groups diverged in a statistically meaningful way, with P values of 0.038 for disease-free survival and 0.003 for overall survival. What makes this observation clinically important is that both groups had received the same modern immunotherapy-based regimen and both had undergone the same operation, yet the simple measure of how much tumor remained after treatment cleanly separated patients into distinct prognostic categories. Tumor regression grade, in other words, is not merely a descriptive label but a genuine predictor of who will live and who will relapse.

When the researchers dug deeper using multivariate analysis, a statistical technique that adjusts for multiple factors simultaneously to identify independent predictors, one variable stood out above all others: ypN3, defined as the presence of seven or more regional lymph node metastases on post-treatment pathology. Patients whose resected specimens contained this burden of nodal disease faced a substantially elevated risk of death and recurrence, independent of other clinical and pathological characteristics. This finding is notable because it suggests that even in the immunotherapy era, the extent of lymph node involvement after neoadjuvant treatment remains the dominant driver of long-term outcomes, echoing decades of surgical oncology research while confirming that the principle holds when immune checkpoint inhibitors are added to the equation.

The recurrence patterns themselves told an equally compelling story. Across the entire cohort, 63 of the 178 patients, or 35.4 percent, experienced a recurrence of their disease during follow-up. But the geography of that recurrence differed sharply between the response groups. Patients in the TRGb group developed distant metastases, the spread of cancer to distant organs such as the lungs, liver, or bones, at a rate of 23.4 percent, compared with only 11.9 percent among the TRGa patients, a difference that reached statistical significance with a P value of 0.046. Distant metastasis is the most feared pattern of recurrence because it generally precludes curative salvage treatment, so identifying which patients are prone to it is essential for planning surveillance and adjuvant therapy.

Lymph node status after treatment also shaped the pattern of failure in a way that carries direct implications for how surgeons and radiation oncologists think about treatment volumes. Patients who remained lymph node positive after neoadjuvant therapy, designated ypN+, developed extra-regional lymph node metastases, meaning cancer spread to lymph node stations outside the originally involved regional basin, at a rate of 10.5 percent. Among patients who were node negative after treatment, the ypN0 group, that figure was only 1.1 percent, a nearly tenfold difference that was statistically significant at P equals 0.007. This suggests that persistent nodal disease after immunotherapy is not just a marker of burden within the chest but a signal that the cancer has acquired the capacity to seed distant lymphatic territories, a behavior that standard surveillance protocols may not adequately capture.

Taken together, the results of this single-center study offer a practical framework for risk stratification in a population that is growing rapidly as NICT becomes standard of care for locally advanced esophageal squamous cell carcinoma. The authors conclude that postoperative ypN status should be used to guide risk-adapted adjuvant therapy and to intensify surveillance for high-risk patients. In practical terms, a patient who emerges from surgery with heavy residual tumor, perineural or lymphovascular invasion, or extensive nodal involvement, particularly ypN3 disease, may be an appropriate candidate for escalated postoperative treatment strategies or closer imaging follow-up, whereas patients with minimal residual disease and cleared nodes might safely avoid the toxicity of unnecessary intensification.

The study does carry the inherent limitations of its design: it was retrospective, conducted at a single institution, and included patients treated over a four-year window during which immunotherapy practice was still evolving. Prospective validation in multi-institutional cohorts will be needed before these risk categories can be formally embedded in treatment guidelines. Nevertheless, as immune checkpoint inhibitors continue to reshape the preoperative treatment of esophageal cancer worldwide, this real-world evidence provides oncologists with something they have sorely lacked: a data-driven map of what recurrence actually looks like in patients who do not achieve a complete pathological response, and a clear signal that the pathology report after surgery contains some of the most prognostically valuable information a treating team will ever receive.

Subject of Research: Survival and recurrence patterns in non-pathological complete response after neoadjuvant immunochemotherapy for esophageal squamous cell carcinoma

Article Title: Patterns of survival and recurrence in non-pathological complete response to neoadjuvant immunochemotherapy for esophageal squamous cell carcinoma: a real-world single-center study

Article References: He, H., Gao, P., Lin, J., Wang, P., Liu, S., & Wang, F. (2026). Patterns of survival and recurrence in non-pathological complete response to neoadjuvant immunochemotherapy for esophageal squamous cell carcinoma: a real-world single-center study. BMC Cancer. https://doi.org/10.1186/s12885-026-17041-8

Image Credits: AI Generated

DOI: 10.1186/s12885-026-17041-8

Keywords: esophageal squamous cell carcinoma, neoadjuvant immunochemotherapy, immune checkpoint inhibitors, pathological complete response, tumor regression grade, ypN3, disease-free survival, overall survival, recurrence patterns, distant metastasis, prognostic factors, esophagectomy