In recent years, the incidence of upper third gastric cancer has gradually increased, and PG has been used more and more, whether laparoscopic surgery, open surgery or robotic surgery[2]. PG is a kind of function-preserving gastrectomy[13]. The remnant stomach has the function of storing and digesting food, which can increase food intake and improve postoperative nutritional status[14, 15]. Gastric acid and intrinsic factors secreted by the remnant stomach can preventing anemia and vitamin B12 deficiency, and the pylorus can slow down the food emptying, making blood glucose levels more stable[16–18]. When performing PG, more than half of the distal stomach needs to be preserved to ensure that the remnant stomach is functional after surgery[19]. Routine dissection of distal perigastric nodes Nos.4d, 5, 6 is not required when D1 + lymph node dissection is performed. Yura[20] et al. reported a pathological study of 202 patients with T2/T3 proximal gastric cancer who underwent TG showed that The metastatic rates at Nos.4d and Nos.12a were very low (0.99% and 0.006%, respectively), and those at Nos.5 and Nos.6 were zero. Therefore, PG with D1 + lymph node dissection for proximal early gastric cancer is an oncologically safe option[21]. There is no absolute requirement for preservation of the vagus nerve[22], but when the vagus nerve around the esophagus is cut, it may cause pyloric spasm and dysfunction[14], and preservation of the hepatic branch of the vagus nerve helps to reduce the occurrence of delayed gastric emptying[23]. Common reconstruction methods after PG include esophagogastrostomy(EG), jejunal interposition(JI) and double-tract(DT)[11, 24]. EG lacks anti-reflux structure, so postoperative esophageal reflux is very high. Miyauchi[4] et al. showed that the reflux rate of EG was as high as 74% (n = 23). Therefore, EG is usually accompanied by anti-reflux procedures, including gastric tube reconstruction, double-flap, side overlap, fundoplication, and rebuild of fundus and HIS Angle. Gastric tube reconstruction reduces the amount of acid-secreting gastric tissue, and an artificial fundus structure above the anastomosis can cushion acid reflux[25]. A comparative study by Chen showed a reflux rate of 11.8% (n = 34) with gastric tube reconstruction[26]. The anti-reflux effect of JI and DT is achieved by inserting a segment of jejunum between the esophagus and the remnant stomach[27, 28], and DT has remnant stomach passage and jejunum passage, which can increase food intake and improve postoperative nutritional status[29]. Tanaka[30] et al. thought that when most food intake pass through the jejunum passage rather than remnant stomach passage,the effect of DT will be similar to TG. Kamiya[31] et al. used the Post-Gastrectomy Syndrome Assessment Scale-45 (PGSAS-45) to investigate 172 patients after DT and found that patients with a larger remnant stomach, a shorter interposition jejunal length (≤ 10 cm), and a bigger size of gastrojejunostomy (> 6 cm) had higher postoperative QOL scores. In this study, we used a Propensity Score Matching to reduce any treatment selection bias and potential confounding[32], we found that DT had a longer operation time than Tube, but there was no increase rate in postoperative complications, It shows that DT is a safe operation, while for some patients who cannot tolerate long surgery, we can try to choose Tube. There was no significant difference in nutritional status between DT and Tube groups. Nomura[16] et al. showed that the weight loss in 12 month after TG was 16.2%, while in our study, the weight loss in 12 month after DT and Tube was 7.1%-7.8%, indicating that the preservation of the stomach has certain advantages over TG. In terms of anti-reflux, although gastric tube reconstruction added anti-reflux structures, compared with DT, endoscopic reflux esophagitis and visick score were higher, and the incidence of reflux ≥ grade B was 40.8% vs 13.5% (p = 0.001), visick score ≥ 3 was 36.3% vs 13.6% (p = 0.012). The effects of EG with other antireflux procedures also need to be studied furthermore.
This study had some limitations. First, this is a single-center retrospective study, and some patients with DT or gastric tube reconstruction after PG were not included in this study due to lack of follow-up data, which will lead to a choosing bias on our propensity score matching. Secondly, the majority of surgeons are not as skilled with DT as with gastric tube reconstruction, thus prolonging his operation time in DT.