Prognostic factors associated with locally advanced gastric cancer patients treated with neoadjuvant chemotherapy followed by surgical resection

In this retrospective study, we analyzed prognostic factors associated with survival outcomes in 73 locally advanced gastric cancer patients treated with neoadjuvant chemotherapy (NAC) followed by surgical resection. Median disease-free survival (DFS) for 64 patients that received R0 resection was 685 days, whereas median overall survival (OS) for 73 patients was 930 days. Multivariate analysis demonstrated that post-treatment nodal stages (P = 0.002), nervous invasion (P = 0.0492) and serum CA199 levels (P = 0.0398) were independent prognostic factors for DFS. Nodal stages (P = 0.0007), presence of nervous invasion (P = 0.0259) and non-radical resection (P = 0.0165) were independent prognostic factors for OS. These results indicate that post-treatment nodal stages, neural invasion and serum CA199 levels are all associated with poor DFS. Moreover, post-treatment nodal stage, resection type and neural invasion status are independent prognostic factors for OS.


INTRODUCTION
Gastric cancer ranks fifth among most malignant cancers and third among cancer related deaths worldwide [1]. China has the highest incidence of gastric cancer accounting for 35% of total gastric cancer cases worldwide with a high mortality rate of 25.16 cases per 100,000 [2]. Prognosis of locally advanced gastric cancer is poor with a 5-year overall survival (OS) rate of20-30% for surgery-only patients [3]. Neoadjuvant chemotherapy (NAC) is preferred for locally advanced gastric cancer patients since the release of the MRC Adjuvant Gastric Infusional Chemotherapy (MAGIC) trial results [4]. Many studies demonstrate that NAC reduces the size of gastric cancer lesions thereby decreasing tumor staging and increasing the chances for radical resection and survival while decreasing post-operative complications [5][6][7][8][9]. However, standard regimen and courses of NAC are not yet established and prognostic factors associated with survival outcomes for patients treated with NAC followed by surgery remain unclear. Therefore, we analyzed the relevant prognostic factors associated with survival by reviewing medical records of 73 locally advanced gastric cancer patients that were treated with neoadjuvant chemotherapy followed by curative-intent surgery. Our analysis included the prognostic status of pathological response to NAC and post-therapy node status (Nodal stage or lymph node metastasis ratio).

Prognostic factors associated with disease-free survival
Median DFS for 64 patients that received R0 resection was 685 days (Figure 1). Univariate analysis showed that nodal stages, lymph node metastasis ratio and nervous invasion were prognostic factors associated with DFS ( * Tumor response to NAC was not evaluated for 3 patients; 2 of them developed ileus during treatment, and the other one missed assessment for personal reasons. ‡/ § vascular invasion and nervous invasion status were not reported for one patient. ║51 patients were administered post-surgery chemotherapy and had medical records in our hospital;The post-surgery chemotherapy records of 22 patients was not known. ¶ 8 patients received post-operative radiotherapy.

Prognostic factors associated with overall survival
Kaplan-Meier survival analysis showed a median OS of 930 days (Figure 2). Univariate analysis for OS showed that nodal stage (N), lymph node metastasis ratio (NR), nervous invasion, resection type and serum CEA/ CA199ratiowere prognostic indicators of OS, whereas, graded pathologic response and presence of vascular invasion showed marginal association with OS (Table 2).  (Table 4).

DISCUSSION
Patients with potentially resectable gastric cancer are treated with neoadjuvant chemotherapy to improve survival. However, a standard regimen is not clear. Several investigations have shown that patients that received NAC followed by surgery, age at diagnosis, post-treatment nodal status, diffuse-type histology, perineural invasion/vascular invasion and salvage surgery are associated with OS [10][11][12][13]. Our study demonstrated that clinical response (SD or PR) was not a significant prognosis factor for DFS or OS. One possible reason is that we enrolled only limited number of patients that underwent curative-intent surgery after NAC.
The prognostic value of pathologic response to neoadjuvant chemotherapy remains controversial although it has been investigated for various malignancies. Kurokawa Y et al. suggested that pathological response was a better surrogate endpoint than RECIST in neoadjuvant chemotherapy for gastric cancer [14]. In many studies, univariate analysis showed that pathologic response was a predictor of survival in GC patients after receiving neoadjuvant chemotherapy; but, multivariate analysis showed that it was not an independent predictor of OS [10,[15][16][17][18]. In this study, univariate analysis showed that pathologic response was marginally associated with OS, but not associated with DFS; multivariate analysis showed that it was not associated with both DFS and OS. Fujitaniet al. showed that pathologic response was associated with OS in the subset of patients with nodal stages N0-1 [10].Therefore, further prospective studies with larger sample size are necessary to confirm the prognostic status of pathologic response in patients that undergo curative-intent surgery after NAC. Metastatic lymph node ratio (NR) is an alternate prognostic factor instead of the number of lymph nodes (N in TNM staging) in GC because of the limited number of lymph nodes [18,19]. Persiani R et al. showed that TRM (R means metastatic lymph node ratio) staging system had better prognostic power than the TNM system by reviewing 219 patients that underwent gastrectomy  [21]. In the present study, NR was a prognostic factor for both DFS and OS in univariate analysis, but was in significant in multivariable analysis. However, posttreatment N stage showed association with DFS and OS in both univariable and multivariable analyses, consistent with the study by Fujitani et al. [10].Therefore, posttreatment N stage was a more reliable prognostic factor than metastatic lymph node ratio (NR) in locally advanced GC patients that received gastrectomy after NAC. Changes in serum CA199 levels demonstrate therapeutic efficacy with increased serum CA199 levels indicating treatment failure or recurrence. Mohri et al. showed thatCA199 was an independent prognostic factor for OS in patients with metastatic gastric cancer [22]. Schauer et al demonstrated that serum CA199 levels predicted survival in patients with diffuse type gastric cancer after surgical treatment [23]. Zhu et al showed that serum CA199 was part of the prognostic index for patients with metastatic gastric cancer that received epirubicin (EPR)-containing triplet regimen as first-line treatment [24]. In our study, univariate analysis showed that elevated serumCA199 levels were associated with both shorter DFS and OS, but multivariate analysis showed that it was associated with poor DFS. This suggested that serum CA199 levels were strong predictors of long-term survival inGC patients.
Our findings are limited because this was a retrospective study conducted in a single institution with few select patients (73 study subjects) with different preand post-operative chemotherapeutic regimens. Larger multi-center prospective studies are necessary to confirm our findings.
In conclusion, our study demonstrates that posttreatment nodal stages, neural invasion and serum CA199 levels are associated with poor DFS. Moreover, posttreatment nodal stage, resection type and neural invasion status are independent prognostic factors for OS. This study also revealed that post-treatment N stage was a more reliable prognostic factor than metastatic lymph node ratio (NR) in locally advanced GC patients that received gastrectomy after NAC.

Patients and treatment schedule
We retrospectively analyzed the outcomes of 73 consecutive patients with locally advanced gastric cancer that were treated with neoadjuvant chemotherapy followed by surgical resection between August 2007 and July 2012 at the National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, China. The procedures followed were in accordance with the ethical standards set by the independent ethnic committee of Cancer Hospital of Chinese Academy of Medical Sciences & Peking Union Medical College on human experimentation and the Helsinki Declaration. The patients were diagnosed with resectable advanced adenocarcinoma (gastric cancer) and completed at least one course of NAC with the tumor response evaluated by Response Evaluation Criteria in Solid Tumors (RECIST, version 1.0). All patients received gastrectomy, with D2 lymph-node dissection. The posttherapy node status including number of metastatic lymph nodes (N stage) and lymph node metastasis ratio (NR)

Statistical analysis
SAS statistical software 9.3 (SAS Institute Inc., Cary, NC, USA) was used for all statistical analyses. Overall survival (OS) was defined as the time from the date of initiation of NAC until death from any cause (event) or the last follow-up date (censored). Disease-free survival (DFS) was defined as the time from the date of surgery until local or distant relapse was detected (event) or the last follow-up date (censored). Kaplan Meier survival analysis was used to determine both DFS and OS.
Multivariate analysis of prognostic factors was performed by the stepwise Cox proportional hazard model with the variables identified as significant factors in the univariate analyses and hazard ratio (HR) and their95% confidence intervals (CIs) were calculated. A two-sided P < 0.05 was considered statistically significant.