Tumor regression grade combined with post‐therapy lymph node status: A novel independent prognostic factor for patients treated with neoadjuvant therapy followed by surgery in locally advanced gastroesophageal junction and gastric carcinoma

Abstract Background Tumor regression grade (TRG) is a measure of histopathological response to neoadjuvant therapy (NAT). Post‐therapy lymph node (ypN) metastasis was reported as a prognostic factor. However, the evaluation of the treatment effectiveness of NAT has not been well studied. Here, we explored whether TRG combined with ypN status could be a prognostic factor for gastroesophageal junction (GEJ) and gastric cancer (GC). Besides, we aimed at making clear the association of different neoadjuvant regimens with different TRG and ypN status. Methods 376 patients with GEJ or GC accepting NAT in Peking University Cancer Hospital were retrospectively collected from January 1, 2003 to June 30, 2021. According to TRG and ypN status, patients were innovatively categorized into four groups: TRG0N0, TRG1‐3N0, TRG0‐1N+, and TRG2‐3N+. We applied Kaplan–Meier method and log‐rank test to testify the differences in disease free survival (DFS) and overall survival (OS) among four groups. Univariate and multivariate analyses were performed to examine the relationships between TRG combined with ypN status and prognosis. Results We observed significant survival differences among the four groups (p < 0.001, respectively). Median DFS and OS of patients with TRG0N0, TRG1‐3N0, and TRG0‐1N+ were not reached, whereas these of patients with TRG2‐3N+ were 17.37 months (95% CI, 14.14–20.60 months) and 39.97 months (95% CI, 27.05–52.89 months). TRG combined with ypN status was still an independent predictor for both DFS (p < 0.001) and OS (p < 0.001) in multivariate analysis. Chi‐squared test showed TRG combined with ypN status was significantly associated with different preoperative treatments (p < 0.001). Patients receiving immunotherapy achieved the highest TRG0N0 rate (31.9%). Conclusion Our results demonstrate that TRG combined with ypN status is a novel independent predictor of both DFS and OS in resectable, locally advanced GEJ and GC. Neoadjuvant immunotherapy achieved the highest TRG0N0 rate.


| INTRODUCTION
Gastric cancer (GC) is the fourth leading cause of cancerrelated mortality worldwide and remains a significant global health threat. 1 Patients in high-incidence regions are usually diagnosed with GC at the advanced stage.the MAGIC and FNCLCC/FFCD studies proved that neoadjuvant therapy (NAT) significantly improved the survival of patients with locally advanced GC. [2][3][4] RESOLVE study further laid the foundation of NAT followed by D2 gastrectomy for GC. 5 NAT has become a standard treatment for resectable gastroesophageal junction (GEJ) and GC based on the National Comprehensive Cancer Network (NCCN) and European Society for Medical Oncology (ESMO) guidelines, 6,7 as well as Chinese Society of Clinical Oncology (CSCO) guidelines. 8Despite receiving standard treatments, half of the patients face the risk of tumor recurrence and death of carcinoma.Thus, predicting the long-term prognosis of patients receiving NAT is an urgent need.
To date, no prognostic factor has filled the predictive vacancy for patients undergoing NAT and radical gastrectomy.The American Joint Committee on Cancer (AJCC)/Union for International Cancer Control (UICC) proposed neoadjuvant therapy TNM staging (ypTNM) as a prognostic marker for patients accepting NAT.Unfortunately, patients who achieve pCR or ypT0 with post-therapy lymph node (ypN) positive after NAT are excluded from the ypTNM system for GC, which makes it incomplete.Tumor regression grade (TRG) is defined as the pathological evaluation of tumors resected after NAT to perform a quantitative analysis of the therapeutic effect on tumors.Notably, ypN status is also a response to NAT.0][11][12][13][14][15][16] According to a few studies, ypN status instead of TRG is an independent prognostic factor for GEJ and GC. 10,13,14However, these studies analyzed the prognostic value of ypN status and TRG separately, which is not sufficient.Dichotomizing patients into ypN-negative (ypN0) and ypN-positive (ypN+) groups is also far from enough.By achieving pathological complete response (pCR), being TRG 0 and ypN0, patients can gain a better outcome than those who do not achieve pCR. 17From a study by Smyth, we can find that the overall survival (OS) of patients with Mandard TRG 1-2 in the ypN+ group was longer than that of patients with Mandard TRG 3-5 in GC. 10 Researches precisely assessing the prognosis after NAT are thus warranted.Furthermore, both TRG and ypN status could reflect the effectiveness of preoperative treatment; however, whether different neoadjuvant regimens can generate different levels of TRG combined with ypN status have not been determined.
In this study, we aimed to evaluate the clinical outcomes of patients receiving NAT, so as to provide surrogate endpoints for NAT in clinical trials.We focused on creating a novel prognostic factor and highlighted its predictive value.Additionally, we opted to expose whether preoperative immunotherapy could improve TRG and ypN status.

| Patients
Patients diagnosed with locally advanced GEJ or GC and administered NAT at Peking University Cancer Hospital between January 1, 2003 and June 30, 2021 were selected for this study.The inclusion criteria were: (1) patients pathologically diagnosed with GEJ or gastric adenocarcinoma before NAT; (2) patients with locally advanced stage of the disease (8th AJCC clinical stage III-IVA); (3) patients administered at least two cycles of preoperative treatment, including chemotherapy, anti-HER-2 therapy, and ICI therapy; and (4) patients who underwent D2 gastrectomy surgery and R0 resection.The exclusion criteria were: (1) patients with concurrent malignant tumors; (2) patients treated with radiotherapy; (3) patients subjected to D1 gastrectomy or R1/R2 surgical resection; and (4) patients whose TRG was unavailable.Finally, 376 of the 723 patients were enrolled in the study.

| Neoadjuvant therapy regimens
Patients received the following NAT regimens: immunotherapy (69 patients; the patients did not participate in clinical trials); platin-based doublet regimens (193 patients); taxol-based or taxol-platin-based triplet regimens (94 patients); and anti-HER-2 therapy without immune checkpoint inhibitors (ICI) (20 patients).Written informed consent was obtained from patients or their legal guardians before treatment initiation.

| TRG assessment and histopathological characters
TRG was assigned according to the NCCN standard.NCCN TRG was categorized into four groups as follows: TRG 0 (complete response, including lymph nodes): absence of viable cancer cells; TRG 1 (near-complete response): presence of single cells or few small groups of cancer cells; TRG 2 (partial response): presence of residual cancer cells with evident tumor regression but a larger number of single cells or groups of cancer cells; and TRG 3 (poor or no response): presence of extensive residual cancer without evident tumor regression. 8TRG of the 376 participants were reviewed by two experienced gastrointestinal pathologists, respectively, and the numbers of regional lymph node metastases were determined by them.Histological types, degrees of cell differentiation, tumor sizes, post-therapy infiltration depth (ypT), vascular or lymphatic invasion (LVI), and venous invasion (VI) were recorded in the patient's pathological report.Expression levels of human epidermal growth factor receptor 2 (HER-2), epidermal growth factor receptor (EGFR), Epstein-Barr encoding region (EBER) and MMR status were detected via immunohistochemistry (IHC).

| Follow-up
Follow-up was performed every 3 months during the first 2 years, every 6 months from the third year to the fifth year after surgery, and once per year thereafter.If patients had symptoms or signs of recurrence or metastasis, follow-up visits were scheduled.

| Statistical methods
Disease free survival (DFS) was calculated from the date of surgery to first tumor recurrence or death.OS was defined as the time from initial NAT to death from any cause or the last date of follow-up.We defined different groups based on different TRG categories and ypN status.Differences in DFS and OS stratified by different combinations were evaluated implementing the Kaplan-Meier method and compared using the log-rank test.A two-tailed p < 0.05 was considered statistically significant.The four-group combination was proven to be a favorable evaluation system.Univariate and multivariate analyses were performed using Cox regression and proportional hazard models.Univariate analysis comprising age, sex, completeness of local resection, TRG combined with ypN status, ypT stage, LVI, and VI was performed to determine the variable that was statistically correlated with prognosis.Statistically significant variables (p < 0.05) in univariate analysis were incorporated into the multivariate Cox regression models to analyze the survival difference among different groups.Neoadjuvant regimens among the four groups were compared using the chi-squared test.SPSS software (version 26) was used to perform statistical analysis.

| TRG combined with ypN status is associated with DFS and OS
The median follow-up time was 38.50 months (95% CI, 37.49-39.51months).The median DFS (mDFS) and median OS (mOS) for the entire population were not achieved.However, 58.1% of the population achieved the 3-year DFS while 71.4% achieved the 3-year OS.
Third, we dichotomized patients into the following two groups for the subsequent study: with or without pCR.We found that patients who achieved pCR demonstrated significantly better survival rates in terms of both DFS (p < 0.001) and OS (p < 0.001), compared to those who did not achieve pCR.(Figure S3A,B).The pCR group had not achieved the mDFS and mOS.In contrast, the non-pCR group had an mDFS of 50.63 months (95% CI, 26.72-74.54months) and mOS of 73.80 months (95% CI, not reached).The 3-year DFS rates of the pCR group versus non-pCR group were 92.5% versus 53.8% and the 3-year OS rates were 97.6% versus 68.2%, respectively.

| TRG combined with ypN status is an independent prognostic factor based on multivariate analysis
We conducted univariate analysis to evaluate the prognostic value of age, sex, location of tumor, Lauren classification, ypT stage, LVI, and VI, TRG combined with ypN status, and neoadjuvant regimens on DFS and OS (Table 2).TRG combined with ypN status, ypT stage, LVI, and VI were all considered as prognostic predictors of DFS (p < 0.001,  respectively) and OS (p < 0.001, respectively).Location of tumor and neoadjuvant regimens were only significantly associated with DFS (p = 0.019, p = 0.007, respectively), and not with OS (p = 0.12, p = 0.12, respectively).The statistically significant factors in the univariate analysis were included in the multivariate analysis (Table 3).TRG  Abbreviations: CI, confidence interval; DFS, disease free survival; HER-2, human epidermal growth factor receptor 2; HR, hazard ratio; ICI, immune checkpoint inhibitors; OS, overall survival; TRG, tumor regression grade; ypT, post-therapy infiltration depth.
regimens was 59.4%; this percentage was obviously higher than that in patients undergoing platin-based doublet regimen (42.0%) and taxol-based doublet or triplet regimens (35.1%).Due to the small number of patients, the results of this section should be interpreted with caution.

| DISCUSSION
Our study aimed to assess the predictive value of combining TRG and ypN status as a prognostic indicator for patients with resectable, locally advanced GEJ, and GC who received NAT.After NAT, patients in the TRG0N0 group had the longest survival, followed by patients in the TRG1-3N0, and TRG0-3N+ groups.Among the patients in the TRG0-3N+ group, the prognosis of those with TRG 0-1 was better than that of those with TRG 2-3.To the best of our knowledge, this is the first study to assess the combination of TRG and ypN status for GEJ and GC.The study is anticipated to shed light on a novel option for defining the endpoints in clinical trials.
Patients achieving pCR had a superior survival benefit, which is consistent with the finding of Li Ziyu for locally advanced GEJ and GC in a meta-analysis. 17However, the proportion of patients achieving pCR was relatively small.There is still no consensus on how to perform postoperative adjuvant therapy for patients with different effects of NAT.Patients who respond well to preoperative treatment generally continue postoperative adjuvant chemotherapy according to the perioperative treatment mode.However, patients who derive limited benefits from NAT are recommended to receive multidisciplinary team (MDT) consultation.Future randomized trial can determine whether changing or intensifying treatment of non-pCR will result in improvements in OS for these patients. 10Moreover, we need to analyze in-depth the molecular biology characteristics of this group of people, hoping to find suitable treatment methods.More patients who could benefit from NAT were selected based on the evaluation of TRG and ypN status.TRG was first used to determine the efficacy of esophageal cancer after concurrent chemoradiotherapy. 18Thereafter, TRG was gradually applied to other malignant tumors including GC.Several TRG systems have been established by different investigators, but no agreement was reached on which one is the most representative for assessment of prognosis.Becker TRG system and Mandard system are the commonly used systems in clinical practice and have been evaluated in many studies. 10,19The CSCO guidelines recommend TRG which is proposed by the 8th AJCC TNM classification or the NCCN guidelines. 8TRG0 using the NCCN definition would be TRG0-ypN0.The predictive value of the NCCN TRG system on GEJ and GC had been recently proven.Sinnamon et al. have found that TRG using the contemporary NCCN definition is associated with OS in locoregional GC. 20 In our study, not only the NCCN TRG system was evaluated, but also the relationship between NCCN TRG combined with ypN status and prognosis on GEJ and GC was revealed for the first time.Based on our results, TRG combined with ypN status was identified as an independent prognostic factor for DFS and OS.Furthermore, patients were subdivided into four groups, including pCR and TRG0N+, enabling a more comprehensive analysis than the simple division of patients into pCR and non-pCR groups.
Patients in the TRG0N0 group achieved the longest survival, and patients with ypN+ had poorer outcomes than those with ypN0 for the rest patients.This result is consistent with those of previous studies.Smyth found that both TRG and ypN status were related to survival.However, only ypN status was revealed to function as an independent predictive factor of OS for patients with chemotherapy plus resection in the MAGIC trial.TRG was not identified as an independent marker for prognosis in the trial. 10Xu Xing drew a similar conclusion that TRG was not an independent prognostic predictor of GC. 21onohoe Claire L. also put forward that none of the existing TRG measures had independent significance for the prognosis of esophageal and GEJ cancers. 22However, these researchers only concentrated on the prognostic significance of TRG or ypN status separately.According to Smyth, the mOS for patients with Mandard TRG 1-2 was 17.3 months in the ypN+ group, which was longer than the 15.5 months found for patients with Mandard TRG 3-5. 10 Therefore, we proposed a novel evaluation approach of combining TRG with ypN status.A few studies reported that the combination of TRG and ypN status had a significant association with survival for esophageal carcinoma. 19,23,24But no such studies had been conducted for GEJ and GC patients.In the present study, the combination of TRG and ypN status was recognized to be of great prognostic value for GEJ and GC.
Both TRG and ypN status have been identified as short-term efficacy evaluation markers of NAT; however, the association between them had not been demonstrated.At the end of our research, we made a supplement to this issue: patients undergoing ICI therapy had the greatest superiority in achieving TRG0N0; and the TRG0N0 rate plus TRG1-3N0 rate in patients treated by ICI-containing regimens was higher than that in patients undergoing chemotherapy only.7][8] In recent years, immunotherapy has become the standard first-line therapy for GEJ and GC.The CheckMate 649 trial suggested that nivolumab with chemotherapy as the first-line treatment significantly prolonged OS and DFS compared with chemotherapy in PD-L1 CPS ≥5 patients. 25The ATTRACTION-4 trial provided evidence that nivolumab combined with chemotherapy might potentially be a new approach for untreated, advanced GEJ, or GC. 26 Besides, in the Check-Mate 577 study, the adjuvant therapy of nivolumab was found to improve the mDFS from 11.0 to 22.4 months in comparison with the placebo, whose participants were esophageal or GEJ cancer without pCR after neoadjuvant chemoradiation. 279][30] Neoadjuvant immunotherapy for esophageal cancer has shown promising initial results.A retrospective analysis revealed that neoadjuvant immunotherapy plus chemotherapy for patients with locally advanced esophageal squamous cell carcinoma led to an advantage in pathological response, and could prolong DFS compared with chemotherapy alone. 31Recently, two single-arm clinical trials with small sample sizes suggested that neoadjuvant immunotherapy could improve the pCR rate of GC patients. 32,33fter signing written informed consents by some GC patients, who might potentially benefit from ICI, or their legal guardians, we made exploratory attempts to treat these patients with ICI before operation in clinical practice.Encouragingly, our retrospective study revealed that this modality had the highest TRG0N0 (pCR) rate, symbolizing superior survivals.This study was the first time to illustrate the short-term therapeutic effect of ICI versus chemotherapy alone as NAT for GEJ and GC, confirming the confidence that NAT containing ICI would lead to successful outcomes.The results also highlight the potential of TRG combined with ypN status as a surrogate endpoint for neoadjuvant clinical trials.
Despite of the promising results, our study had some limitations.This study was a single-center retrospective exploratory analysis that has not been validated.Furthermore, although patients meeting the TRG0N+ criteria were eligible for inclusion in this study, only three such patients were enrolled.As a result, the conclusion was underpowered for application to the subgroup.Finally, the follow-up time for patients receiving neoadjuvant immunotherapy was short, warranting long-term survival validation.

| CONCLUSIONS
Based on our results, TRG combined with ypN status is a novel independent predictor of both DFS and OS in resectable, locally advanced GEJ, and GC.The application of neoadjuvant immunotherapy increased the TRG0N0 and TRG1-3N0 rates, which indicated a good prognosis in our findings.However, prospective and multicenter studies with a larger number of participants are required to verify the results.
with the Declaration of Helsinki and approved by the Institutional Review Board of Peking University Cancer Hospital (protocol code 2020KT08).

T A B L E 1
(Continued) F I G U R E 1 Differences of disease free survival (A) and overall survival (B) for patients separated into four groups by tumor regression grade (TRG) and post-therapy lymph node (ypN) status: TRG0N0, TRG1-3N0, TRG0-1N+, and TRG2-3N+ groups.
Patient characteristics.
T A B L E 1 90.8% compared with 78.8% for TRG 1, 51.5% for TRG 2, and 46.3% for TRG 3. Further, the 3-year OS for patients with TRG 0 was 95.4% versus 87.5%, 66.2%, and 63.7% for TRG 1-3, respectively.[Correction added on October 18, 2023 after first online publication.The values in the previous sentence have been updated in this version.] Univariate analysis of the factors affecting the DFS and OS of patients with gastroesophageal junction and gastric cancer and treated with neoadjuvant therapy.

3.4 | TRG combined with ypN status is significantly associated with different preoperative treatments
Multivariate analysis of the factors affecting the DFS and OS of patients with gastroesophageal junction and gastric cancer and treated with neoadjuvant therapy.
T A B L E 3