The Prognostic Value of Postoperative Lymph Node Ratio in Gastric Adenocarcinoma Patients Treated With Neoadjuvant Chemotherapy

Objective In this study, we aimed to investigate the prognostic value of postoperative lymph node ratio (LNR)in locally advanced gastric cancer (GC) patients receiving neoadjuvant chemotherapy (NACT). Methods LNR was calculated as the ratio of positive LNs to the total LNs removed. The receiver operating characteristic (ROC) curve was plotted to estimate the cut-off value of LNR for recurrence. The area under the curve of LNR was 0.714 (95% CI: 0.604-0.825, p<0.001) with 60% sensitivity and >0.255 with 76% specificity. Patients were grouped as group I (≤0.255) and group II (>0.255). Results In this study, 157 GC patients were included (39.5% female and 60.5% male). Of the patients, 97 (61.8%) were in group I and 60 (38.2%) were in group II. Disease‑free survival (DFS) was not reached in group I, and it was 16 months in group II (p<0.001). Overall survival (OS) was 58 months in group I and 28 months in group II (p>0.001). In multivariate analysis, lymphovascular invasion, neoadjuvant response, adjuvant treatment, and LNR were found to be the factors associated with DFS and OS (p<0.05). Conclusion In our study, it was observed that LNR can predict survival rates better than LN staging.


Introduction
The incidence of gastric cancer (GC) has been decreasing since the 1930s; however, it still remains a major cause of cancer-related deaths globally. Most GC patients are symptomatic and have a locoregional or advanced-stage disease at the time of diagnosis, but despite the advances in treatment modalities, only half of those patients with locoregional tumors are able to undergo potentially curative resection [1][2][3].
Prospective randomized trials and meta-analyses have indicated improved survival with multimodality approaches, such as adjuvant chemoradiotherapy, adjuvant chemotherapy (ACT), and neoadjuvant chemotherapy (NACT) compared to surgery alone. The positive effects of these therapies on survival outcomes in GC patients have become prominent over time, although there is no consensus as to which is the best approach [4][5][6][7][8].
There are two major staging systems related to GC, which are as follows: (i) the Japanese classification based on anatomical location, especially of the lymph node (LN) stations, and (ii) the tumor, node, and metastasis (TNM) staging system developed jointly by the American Joint Committee on Cancer (AJCC) and the Union for International Cancer Control (UICC). The recent AJCC/UICC TNM staging classification (eighth edition, 2017) includes another prognostic stage group for clinical and pathological staging following NACT [9,10].
In the TNM classification system, LN staging is based on the total number of metastatic LNs and does not take into account the number of total LNs removed. In addition, it may cause a limitation in both tumor staging and prediction of survival, especially in patients with inadequate LN dissection [11]. 1 1 2 3 4 5 6 1 In Japan, D2 dissection has been recommended as standard practice since the 1960s. East Asian surgeons, especially Japanese and Korean surgeons, have routinely performed gastrectomy with D2 dissection. However, most Western surgeons perform gastrectomy with only D1 dissection, because D1 was reportedly associated with less mortality and morbidity than D2 in prospective randomized trials performed in the Netherlands and the UK, which concluded that there was no survival benefit for D2 over D1 LN dissection [12,13].
The debate on lymphadenectomy and the extent of the total number of LNs removed began in the 1980s. The National Comprehensive Cancer Network (NCCN) guidelines recommend the examination of 16 or more regional LNs to determine the N status. Several studies have demonstrated a robust association between the number of LNs removed and improved survival rates [14,15].
LN metastasis was the only independent predictor of survival in patients treated with NACT in the Medical Research Council Adjuvant Gastric Infusional Chemotherapy (MAGIC) trial, and pathological response to chemotherapy was not associated with survival [16]. The LN ratio (LNR) is calculated as the ratio of positive LNs to the total LNs removed. Previous studies and meta-analyses have confirmed that LNR is an independent prognostic factor for overall survival (OS) in GC patients who undergo surgery without NACT [17][18][19][20]. In the current study, we aimed to investigate the prognostic value of postoperative LN ratio (ypLNR) in locally advanced GC patients receiving NACT.

Study population
The patients who were followed up at the Yüzüncü Yıl University Faculty of Medicine and Prof. Dr. Cemil Taşcıoğlu Istanbul City Training and Research Hospital between 2010 and 2019 were included in the study.
Patients who met any of the following criteria were excluded: metastatic stage, unoperated patients, patients with disease progression during NACT, those with complete response to NACT, surgical margin-positive patients, those who were <18 years of age, patients operated on without NACT, patients who were LN-negative after NACT, those with a history of a second primary cancer, and those with missing data. A total of 653 patient files were reviewed, and 157 patients were ultimately recruited for the study (Figure 1). Patients were restaged according to the AJCC Cancer Staging Manual, 8th edition.

Ethics committee approval
This study was conducted in accordance with the Declaration of Helsinki and reviewed and approved by the Ethics Committee of the University of Health Sciences, Prof. Dr. Cemil Taşcıoğlu Istanbul City Training and Research Hospital (46870771-514.10-12.10.2020).

Statistical analysis
SPSS Statistics 22.0 for Windows software (IBM Corp., Armonk, NY) was used for the statistical analysis. Descriptive analyses were described as mean, standard deviation, and minimum and maximum values for numerical variables while categorical variables were presented as numbers and percentages. The student's t-test was used when numerical variables had normal distribution in two independent groups, and the Mann-Whitney U test was used in the absence of normal distribution. Chi-square analysis was used to compare the ratios in the groups. Monte Carlo simulation was applied when the conditions were not met. Survival analyses were performed by the Kaplan-Meier method. The determinant factors were examined by Cox regression analysis. The ENTER model was used for the factors with p<0.05 as determined in the univariate analysis. The receiver operating characteristic (ROC) curve was plotted to estimate the optimal cut-off value of ypLNR for recurrence. Area under the curve of ypLNR was 0.714 (95% CI: 0.604-0.825, p<0.001) with 60% sensitivity and >0.255 with 76% specificity ( Figure 2). Patients were grouped according to ypLNR as group 1 (ypLNR of ≤0.255) and group 2 (ypLNR of >0.255). Disease-free survival (DFS) was calculated as the time from the initiation of treatment to progression. Overall survival (OS) was calculated as the time from the date of diagnosis to the date of death or last follow-up. Cut-off values were determined according to the ROC curve analysis. An overall 5% alpha error level was used to infer statistical significance. A p-value of <0.05 was considered statistically significant.

Results
The study population consisted of 157 eligible patients; 62 (39.5%) of them were female and 95 (60.5%) were male.  Table 1.

Discussion
The present study investigated the effect of ypLNR on prognosis in patients with locally advanced GC who were operated on after NACT and were LN-positive postoperatively. The cut-off value for ypLNR was identified as 0.255 with 60% sensitivity and 76% specificity. Both DFS and OS were significantly better in patients with low ypLNR rates. Furthermore, ypLNR higher than 0.255 was observed to increase the risk of recurrence by 2.4 times and the risk of mortality by 2.26 times.
Previous studies have investigated the association between LNR and survival in several types of solid tumors [21][22][23]. In GC, the association between LNR and survival has been investigated mostly in patients not treated with neoadjuvant therapy [24]. Studies have demonstrated that LNR may be used as a more convenient and reliable parameter than TNM classification in operated patients with locally advanced GC. Moreover, LNR has been shown to be potentially prognostic for liver and peritoneal metastasis in this patient group [25][26][27].
In the literature, the only study conducted with patients treated with NACT appears to be a retrospective single-center study by Rawicz-Pruszyński et al. involving 95 patients. Their study looked at the effect of NACT on ypLNR in GC patients. The authors reported that tumor diameter of >3.5 cm, Lauren intestinal subtype, a lack of response to NACT, serosal infiltration, LN metastases, and distant metastases were significantly associated with higher ypLNR [18]. Eren et al. showed that higher ypLNR was associated with worse survival in their study in which patients received modified docetaxel, cisplatin, and fluorouracil (mDCF) preoperatively [28]. Higher ypLNR was associated with significantly shorter DFS and OS in our study. ypLNR was determined as a factor that affects survival both in the univariate analysis and multivariate analysis. There was no difference between ypLNR groups in terms of recurrence localization.
Studies have shown the independent effect of pT stage and tumor size on survival in patients operated on without receiving neoadjuvant treatment [29][30][31]. On the other hand, studies in patients receiving NACT revealed an effect on survival with ypN rather than the ypT stage [32,33]. In our study, the multivariate analysis did not show any effect of the ypT stage on survival. While the ypN stage appeared to affect survival in the univariate analysis, ypN was observed to lose its importance when added together with ypLNR in the multivariate analysis.
Conflicting study results have been reported concerning the effect of NACT response on survival. Most studies have shown increased survival with increased tumor response [32,34,35]. However, Smyth et al. did not find a correlation between tumor response and survival [16]. Our study, unlike other studies, included only LN-positive patients, and in line with most of the other studies, we observed increased survival with increased response to NACT. In addition, we observed significantly increased DFS and OS associated with NACT completion.
Patients with positive surgical margins were not included in the present study to avoid a potential effect on results. This study has some limitations, including the retrospective design and enrollment from only two centers. Also, our study population consisted only of Turkish patients.
Increased tumor response in GC patients operated on after NACT, NACT completion, and ypLNR rates under 0.255 were observed to significantly improve both DFS and OS in the present study. In conclusion, ypLNR has been identified as a more significant prognostic marker than the ypN stage in patients who remain LN-positive after NACT. We believe ypLNR may be used as a conveniently estimated prognostic marker in this group of patients. Our study results warrant confirmation via larger studies across different populations.

Conclusions
Based on our findings, ypLNR is a more significant prognostic marker than the ypN stage in patients who remain LN-positive after NACT. We believe ypLNR may be used as a conveniently estimated prognostic marker in this group of patients. Larger studies across different populations need to be conducted in order to confirm our findings.

Additional Information
Disclosures