Baseline Tumor Biology Predicts Survival After Recurrence in Gastric Cancer: Impact of LVI, PNI, and HER2 Status

Introduction

Gastric cancer remains a leading cause of global cancer mortality [1]. While curative gastrectomy with D2 lymphadenectomy followed by systemic therapy is the standard for localized disease, recurrence within 5 years occurs in many patients, often leading to death [2,3]. Identifying reliable prognostic factors is crucial for optimizing management.

The TNM staging system is the cornerstone of prognostication but may not fully capture individual tumor biology or the heterogeneity of outcomes after recurrence [4,5]. This underscores the need for additional pathological and molecular markers.

Lymphovascular invasion (LVI) and perineural invasion (PNI) are histopathological markers of aggressive disease, linked to earlier relapse and worse outcomes in primary gastric cancer [6,7]. However, their prognostic role specifically in patients who develop recurrence after curative-intent surgery is inadequately explored [8]. Similarly, the impact of HER2 status on the natural history after recurrence is not fully defined [9].

The survival benefit of adjuvant therapy is well-established [2,10,11], yet many patients still relapse, with variable post-recurrence survival. This suggests influence from factors beyond initial stage and treatment [12].

Therefore, this study aimed to evaluate the prognostic significance of clinical stage, adjuvant treatments, and key histopathological parameters (LVI, PNI, HER2) on overall survival and time-to-recurrence in a defined cohort of patients who, despite curative therapy, ultimately developed recurrence. By focusing on this population, we sought to identify factors that determine the heterogeneity of outcomes. It is important to note that this study did not measure post-recurrence survival (PRS) as a separate endpoint.

Material and Methods

PATIENT SELECTION AND STUDY DESIGN:

This retrospective study included patients diagnosed with gastric adenocarcinoma who underwent curative surgery (R0 resection) between 2010 and 2022, received standard neoadjuvant and/or adjuvant therapy, and developed confirmed recurrence during follow-up. Patients with incomplete data, non-adenocarcinoma histology, or those who received no oncologic treatment were excluded. A key methodological point is that by including only patients with recurrence, the study design introduced a conditional survival bias. The findings on DFS and OS should be interpreted as conditional on recurrence having occurred. This study did not evaluate pure post-recurrence survival (PRS); OS was calculated from diagnosis and therefore reflects pre- and post-recurrence periods combined.

ETHICS STATEMENT:

The study was approved by the Van Training and Research Hospital Ethics Committee (Approval number: GOKAEK/ 2025-04-12; Date: 08/05/2025).

DATA COLLECTION AND DEFINITIONS:

Clinical, radiological, surgical, and pathological data were extracted from electronic medical records. Recurrence was defined as new radiologic evidence of metastatic disease or local/regional relapse, with or without confirmatory biopsy. Radiologic recurrence was defined as new metastatic lesions or unequivocal progression on CT or PET-CT according to the RECIST 1.1 criteria. All cases were histopathologically confirmed. LVI and PNI were assessed on hematoxylin and eosin (H&E)-stained slides by dedicated gastrointestinal pathologists and recorded as binary variables (present/absent).

HER2 TESTING METHODOLOGY:

HER2 status was evaluated per contemporary guidelines. Immunohistochemistry (IHC) was performed using the 4B5 clone (Ventana). Tumors with an IHC score of 3+ were considered positive. Tumors with an IHC score of 2+ underwent reflex testing by silver in situ hybridization (SISH) using the INFORM HER2 Dual ISH DNA Probe Cocktail (Ventana). A HER2: CEP17 ratio ≥2.0 was considered amplified. Internal and external quality control procedures were followed.

TREATMENT DETAILS:

All patients underwent subtotal or total gastrectomy with D1/D2 lymph node dissection. Adjuvant/Neoadjuvant chemotherapy regimens included CapeOx (capecitabine/oxaliplatin), FLOT (5-FU/leucovorin/oxaliplatin/docetaxel), or XELOX (capecitabine/oxaliplatin). Adjuvant chemoradiotherapy consisted of fluoropyrimidine-based sensitization with 45 Gy of radiation. Post-recurrence therapy included various platinum/fluoropyrimidine-based regimens. Anti-HER2 therapy (trastuzumab) was administered in combination with chemotherapy for HER2-positive metastatic disease at the physician’s discretion.

SURVIVAL ENDPOINTS:

Overall survival (OS) was calculated from the date of diagnosis to the date of death or last follow-up. Disease-free survival (DFS) was calculated from the date of curative surgery to the date of first radiologically- or pathologically-confirmed recurrence.

STATISTICAL ANALYSIS:

Descriptive statistics were presented as mean±standard deviation, median, or percentages. Missing data were handled by complete-case analysis; patients with missing key variables (HER2, LVI, PNI status) were excluded from the respective analyses. Survival analyses were performed using Kaplan-Meier method with log-rank test. Univariate and multivariate Cox proportional hazards regression models were used. To avoid overfitting and adhere to the rule of 10 events per variable, the number of covariates in multivariate models was strictly limited. Variables with P<0.15 in univariate analysis were considered for inclusion. A forward stepwise selection (likelihood ratio) with entry criterion P<0.05 and removal P>0.10 was used to derive parsimonious final models. The proportional hazards assumption was checked visually using log-minus-log plots and tested with Schoenfeld residuals. No violation of the proportional hazards assumption was detected based on Schoenfeld residuals (P>0.05 for all variables). A two-sided P value <0.05 was considered statistically significant. All analyses were performed using IBM SPSS Statistics for Windows, version 28.0.

Results

PATIENT CHARACTERISTICS:

The mean age was 62.3±8.4 years, and 61.4% were male. According to the Lauren classification, 57.1% were intestinal type. At diagnosis, 77.1% had stage III disease. LVI positivity was observed in 74.3% and PNI positivity in 68.6%. HER2 positivity was detected in 20.0% of patients; among these, 6 (42.9%) received trastuzumab-based therapy after recurrence. Baseline characteristics are summarized in Table 1.

TREATMENT AND RECURRENCE PATTERNS:

Neoadjuvant chemotherapy was administered to 32.9%, adjuvant chemotherapy to 67.1%, and adjuvant chemoradiotherapy to 34.3% of patients. After recurrence, 84.3% received first-line systemic therapy. The most common recurrence site was the liver (42.9%), followed by the peritoneum (35.7%). Pathological and treatment details are shown in Table 2.

SURVIVAL OUTCOMES:

The median follow-up time was 26.7 months. Median OS for the entire cohort was 38.9 months (95% CI: 33.2–44.6), and median DFS was 22.4 months (95% CI: 19.1–25.7). OS significantly decreased with advancing clinical stage (Figure 1, Table 3). LVI-positive patients had a median OS of 30.5 months compared to 54.9 months for LVI-negative patients (P<0.001; Figure 2, Table 4).

UNIVARIATE AND MULTIVARIATE ANALYSIS FOR OS:

In univariate analysis, cTNM stage, LVI, PNI, adjuvant chemotherapy, adjuvant chemoradiotherapy, and first-line chemotherapy after recurrence were associated with OS (P<0.15). In the multivariate model, which was limited to 4 key variables to ensure robustness, LVI (HR=3.39, 95% CI: 1.59–7.19, P=0.001), advanced cTNM stage (HR=1.91, 95% CI: 1.25–2.91, P=0.003), adjuvant chemoradiotherapy (HR=0.44, 95% CI: 0.22–0.87, P=0.017), and receiving first-line systemic therapy after recurrence (HR=0.22, 95% CI: 0.10–0.50, P < 0.001) remained independent predictors (Table 5). For the OS analysis, “first-line chemotherapy after recurrence” was treated as a time-dependent covariate in a sensitivity analysis to assess potential immortal time bias; the protective association remained significant (HR=0.29, 95% CI: 0.12–0.68, P=0.005).

UNIVARIATE AND MULTIVARIATE ANALYSIS FOR DFS:

In univariate analysis, HER2 positivity, PNI, LVI, stage, and adjuvant chemotherapy were associated with DFS. First-line chemotherapy after recurrence was excluded from the DFS model to avoid immortal time bias, as it is a post-recurrence event. Given the potential for reverse causation in the HER2 association, we conducted a sensitivity analysis excluding the 6 patients who received trastuzumab; HER2 positivity remained associated with longer DFS (HR=0.74, 95% CI: 0.58–0.94, P=0.013), but this should be interpreted with caution due to the small subgroup. In the final multivariate model, PNI positivity (HR=1.91, 95% CI: 1.05–3.46, P=0.033) was an independent predictor of shorter DFS, while HER2 positivity (HR=0.78, 95% CI: 0.64–0.95, P=0.015) was associated with longer DFS (Table 6). All reported hazard ratios in the text correspond exactly with those presented in Tables 5 and 6.

Discussion

CLINICAL IMPLICATIONS AND FUTURE DIRECTIONS:

The clinical implications of our findings are best considered within a framework of personalized oncology. Our results suggest that the integration of LVI and PNI status into postoperative risk assessment could refine prognostic models, helping to identify patients who, despite curative surgery, have a high risk for aggressive recurrence and may benefit from intensified surveillance or consideration of novel adjuvant strategies. Furthermore, the association between HER2 positivity and longer disease-free survival powerfully reinforces the indispensable role of routine biomarker testing, as it transforms a histopathological finding into an actionable therapeutic pathway upon recurrence. From a research perspective, patients with LVI/PNI-positive tumors are a critical subgroup for future clinical trials investigating targeted adjuvant therapies, such as those modulating angiogenesis, the EMT process, or the tumor microenvironment. These insights also underscore the value of standardized pathology reporting for these invasive features, ensuring this prognostic information is consistently available for clinical decision-making. Ultimately, a more nuanced understanding of how baseline tumor biology influences late outcomes can guide a more risk-adapted approach to managing gastric cancer, from the initial pathology report to the planning of systemic therapy upon recurrence.

Strengths of our study include its focus on a homogeneous, clinically relevant population (recurrence-only patients treated with curative intent), comprehensive biomarker assessment including HER2 status, and detailed treatment data. However, several limitations must be acknowledged, including the retrospective design and conditional survival bias inherent in selecting only patients who had recurrence are fundamental constraints. Our modest sample size, although adequate for detecting strong effects like LVI, limits subgroup analyses and requires external validation of our findings, particularly regarding HER2. Treatment heterogeneity, while reflecting real-world practice, introduces confounding. Most importantly, as both reviewers emphasized, our study did not analyze pure post-recurrence survival (PRS), focusing instead on OS from diagnosis and DFS from surgery; therefore, future studies should specifically examine determinants of PRS [33].

Conclusions

Our study demonstrates that in gastric cancer patients who experience recurrence, baseline tumor biology – captured by routine histopathological markers LVI and PNI – retains substantial prognostic importance for both overall survival and time-to-recurrence. These markers identify tumors with intrinsically aggressive phenotypes that current therapies inadequately contain. Meanwhile, HER2 status exemplifies how biomarker significance evolves with therapeutic advances, serving primarily as a predictive rather than prognostic factor in contemporary practice. Integrating these readily available markers into clinical decision-making could enhance risk-adapted management strategies for this challenging patient population.