[Exploration of potential beneficial people of neoadjuvant chemotherapy based on clinical staging in gastric cancer: a single center retrospective study].

To evaluate the accuracy of the clinical staging by comparing preoperative clinical stage and pathological stage in gastric cancer patients, and to explore the potential beneficial population of neoadjuvant chemotherapy for gastric cancer. We retrospectively collected the clinical data of consecutive patients with gastric cancer who met the inclusion criteria (gastric adenocarcinoma, undergoing laparoscopic or open D2 radical operation, definite cTNM and pTNM) for admission of the Gastrointestinal Center of Peking University Cancer Hospital from July 2013 to April 2019. Patients with the number of harvested lymph nodes less than 16, history of gastric operation or preoperative radiochemotherapy were excluded. Preoperative clinical stage was obtained from abdominal and pelvic enhanced CT by radiologists, and postoperative pathological stage was derived from postoperative pathology reports. The concordance rate between preoperative clinical stage and postoperative pathological stage, and the proportion of pathological stage I in patients with specific preoperative clinical TNM stage were analyzed and compared. The potential beneficial population of neoadjuvant chemotherapy were considered as pI < 5%. Relationship between clinical features and concordance rate of stage was further analysed. A total of 459 patients were included in the analysis, including 321 males and 138 females with mean age of 60 (23 to 85) years old. The concordance rate from T1 to T4 between preoperative clinical T staging and postoperative pathological T staging was 82.5% (33/40), 31.1% (28/90), 34.4% (62/180), and 55.0% (96/149), respectively. The concordance rate from N0 to N3 between preoperative clinical N staging and postoperative pathological N staging was 58.8% (134/228), 22.1% (19/86), 23.6% (26/110), and 54.3% (19/35), respectively. The sensitivity and specificity of abdominal enhanced CT in the diagnosis of lymph node metastasis were 64.5% (171/265) and 69.1% (134/194) respectively. The clinical stage of cT3/T4 patients with pathological stage I was 9.1% (30/329), and the sensitivity of corresponding pathological stage III was 94.8% (164/173), while the cT3/4+cN1-3 patients with pathological stage I stage was 1.4% (3/218), and the sensitivity of corresponding pathological phase III was 76.9% (133/173). Tumor location was associated with the concordance of cT/pT staging [gastroesophageal junction: 64 (56.6%), upper stomach: 9 (9/17), middle stomach: 31 (40.3%), lower stomach: 97 (39.9%), whole stomach: 4(4/9), χ(2)=9.845, =0.043]; the degree of tumor differentiation was associated with the concordance of cN/pN staging [poorly differentiated: 94 (42.3%), moderated differentiated: 92 (41.1%), well differentiated: 12 (12/13), χ(2)=13.261, =0.001], whose differences were statistically significant (all <0.05). Based on a single-center retrospective data from Peking University Cancer Hospital, we think that the potential beneficial population of neoadjuvant chemotherapy for gastric cancer are those clinically staged as cT3/4+N1-3.