Effects of a stepwise, local patient-specific early oral feeding schedule after gastric cancer surgery: a single-center retrospective study from China

Nowadays, early oral feeding after gastrectomy has been gradually accepted and applied in the clinical practice, but there is still no specific uniform feeding regimen available which works best for patients in different regions with different races and eating habits. Aiming to establish an early oral feeding schedule suitable for local Chinese patients after gastric surgery, from May 2014 to May 2018, 87 gastric cancer patients undergoing various types of gastric resections were enrolled in an early feeding protocol and their clinical course was reviewed retrospectively. A stepwise, local patient-specific, early oral feeding schedule was proposed, implemented within an early recovery after surgery (ERAS) protocol and accessed in terms of its safety and tolerability. The primary surgical outcomes included: a median (interquartile range; IQR) postoperative hospital stay of 6 (3) days; 67 (77%) patients were well tolerant of this schedule from postoperative day (POD) 1 to POD 4; 20 (23%) patients had mild I/II grade complications (Clavien-Dindo classification); 3 (3%) patients had IIIB complications, zero cases of hospital mortality. Compared to similar studies in the past, our early oral feeding program is also safe and beneficial, and it can shorten the postoperative hospital stay without causing any increase in postoperative complications. In summary, our work herein reported the establishment of a detailed early oral feeding schedule embedded within an ERAS protocol which was found to be suitable for local Chinese patients after gastric surgery. Accordingly, this early oral feeding schedule is worth further research and promotion.

1. Preoperative consultation and patient's qualification survey 2. Preoperative carbohydrate loading (800 ml of Maltodextrin drink was given 2-12 h prior to surgery) 3. Antibiotic prophylaxis (1.5 g Cefuroxime was given 30 min prior to surgery by iv) 4. Operation approach (laparoscopic gastrectomy is preferred to open gastrectomy) 5. Balanced intravenous fluid therapy (<2500 ml intravenous fluid during the day of surgery, postoperative rehydration includes intravenous infusion of amino acid and fat emulsion solution) 6. No routine gastric tube or pull the gastric tube as soon as possible after surgery 7. TAP block and standard anesthesia protocol 8. Postoperative analgesia avoiding opioids (Parecoxib sodium 40 mg Bid. iv) 9. Postoperative oxygenation therapy (4-6 L/min) 10. Postoperative antiemetic, tropisetron hydrochloride injection (4 mg, iv, qd.) when needed 11. Early oral feeding (initiate on POD 1) 12. Urinary catheter removal on POD 1 13. Abdominal drainage tube removal on POD 4 14. Full ambulation on POD 1 (walking along the corridor, at least 2 hours out of bed)  Table 2. Advanced schedules of oral intake and intravenous fluid infusion (standard = 60 kg adult). Note: Oral diet were supplemented with parenteral nutrition up to the calculated total calories.  www.nature.com/scientificreports www.nature.com/scientificreports/ on the first postoperative day (POD). If there is no obvious gastrointestinal discomfort, such as repeated nausea, vomiting, persistent cramping pain, then give a clear liquid diet (six times a day) to patients on POD 2, then take a liquid diet on POD 3, and take a semi-liquid diet from POD 4 until hospital discharge (Table 2). Tables 3-6 gives adetailed dietary plan and calorie intake for this early oral feeding schedule. According to routine recommendations, the patient's first day of calorie supply is about 20 kcal/kg, and about 30 kcal/kg from the second day after surgery. A joint parental nutrition program was also implemented to supplement the patient's calorie needs as to the total calories needed per day. During the first four PODs all patients received intravenous fluids in the form of a fluid restriction regimen (Table 2). Starting on POD1, they were also encouraged to ambulate actively. Once the following objective criteria are met, all patients should be discharged on POD 5 or 6: no signs of postoperative symptoms, can ambulate without help, can take ≧60% of a meal, pain can be tolerated on oral analgesia, and a self-willingness to go home.
Assessment of surgical outcomes. The primary outcomes were postoperative complication (graded according to the Clavien-Dindo classification) 27,28 , the duration of postoperative hospital stay (DPOS), the time of first flatus (gas passage), the 30-day post-discharge readmission rates and hospital mortality. The secondary outcome was the tolerance of early oral feeding (capable of ingesting 60% or more of a given meal means tolerated, otherwise it is intolerable) and adherence to other provisions of the ERAS protocol.
Statistical analysis. Since the sample size in this study was small (n = 87), an exploratory study was performed on all patients. Descriptive statistics were calculated and in accordance with general principles. If the continuous variables are normally distributed, they are presented as mean ± sd. Otherwise, they will be displayed by median (interquartile range; IQR) values and compared by a non-parametric test (Mann-Whitney), while the count and frequency are used for categorical variables. To explore the relationship between the possible influencing factors and the likehood of a responsive event-Early Oral Feeding Intolerance, univariate Logistic regression analysis was performed firstly and followed by a stepwise regression method. OR (95% CI) is calculated for each predictor, and OR > 1 indicates a high risk of intolerance, and OR < 1 indicates a low risk of intolerance. SPSS ver.     www.nature.com/scientificreports www.nature.com/scientificreports/ 22.0 (SPSS Inc., Chicago, IL, USA) was utilized throughout, and statistical significance was accepted for P-value of < 0.05. Ethical approval. This study was approved by the local Ethics Review Committee of Tongren Hospital, Shanghai Jiao Tong University School of Medicine. The study was carried out in accordance with the Declaration of Helsinki in 1964 and its later amendments. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and are consistent with the Helsinki declaration of 1964 and its subsequent amendments or similar ethical standards.
Informed consent was obtained from all individual participants included in the study.
Postoperative complications occurred in 23 (26%) patients after surgical resection and recovery. Of these, 13 (15%) were mild (Clavien-Dindo grade 1), of which 9 (10%) had just postoperative nausea and vomiting, and 3 (3%) were severe (Clavien-Dindo grade 3-5). Table 9 shows a detailed description of these complications. Two patients were readmitted due to gastric retention and intestinal ileus within 30 days after discharge, and were cured by conservative medical treatments. No single in-hospital mortality was observed. The tolerability to early oral feeding. Sixty-seven patients (77%) successfully started water intake on POD 1 and successfully started a half-amount semi-liquid diet on POD 4 (Table 8). Twenty (23%) patients were unable to tolerate the postoperative early oral feeding schedule. Of these 20 patients, 14 patients delayed the oral feeding schedule due to gastrointestinal symptoms, such as recurrent nausea, vomiting, and persistent abdominal discomfort or pain, and 6 patients were discontinued on oral intake completely due to postoperative complications (3 cases of paralytic ileus, 2 cases of luminal bleedings, and 1 case of anastomotic leakage) ( Table 9). In addition, according to Fig. 1, it can be seen that in all items of the ERAS protocol, patients had the lowest compliance with early oral feeding (77%), followed by the other two lower: Balanced intravenous fluid therapy (80%) and Full ambulation on POD 1 (82%). Possible predictors of early oral feeding intolerance after gastric resection. Twenty (23%) patients failed to follow the early postoperative oral feeding schedule after the start of water intake or semi-liquid diet. To reveal variables that indicate early postoperative oral feeding intolerance, we compared success and failure in terms of patient's age, sex, medical comorbidity, ASA grade, BMI, resection type, operative approach, and TNM    (Table 10) demonstrate that the operative approach may be a relevant factor indicating patients' low compliance with early oral feeding schedule in the ERAS protocol, and it is the only factor in the model that uses a stepwise approach. The TNM stage III vs. I (P = 0.06) also showed a marginal significance associated with the schedule deviation.

Discussion
Gastric cancer is one of the most common malignancies, and resection is the only way to prolong the survival of patients with gastric cancer 1,29 . In recent years, with the increasing diagnosis rate of early gastric cancer, the continuous improvement of laparoscopic gastrectomy, the overall survival rate of patients after gastric cancer surgery is also increased [30][31][32][33][34] . Therefore, how to improve the quality of postoperative life and faster recovery of patients with gastric cancer has become the focus of medical workers.
Preoperative care programs based on the ERAS protocol have been proposed to maintain physiological function and accelerate recovery after various types of surgery 7,14,18 . Dietary care is an important factor affecting postoperative recovery in patients with gastric cancer 7,8 . In the past, patients who underwent gastric cancer resection were routinely fasted for 3 to 4 days, and gastrointestinal decompression and enteral feeding were performed through the nasogastric tube until the patient's intestinal function recovered. However, in recent years, early postoperative oral feeding as an important component of accelerated rehabilitation surgery has been extensively used in clinical practice abroad, especially in colorectal surgery 18,20,35,36 . A growing body of clinical evidence suggests that early oral feeding in patients with colorectal cancer is safe and tolerable. Compared with traditional postoperative diet, oral feeding within 24 hours after colorectal surgery can speed up the recovery of intestinal function and promote early recovery of patients without increasing postoperative complications. However, compared with colorectal cancer, the safety and reliability of early oral feeding after gastric cancer surgery has been lacking high-level therapeutic evidence. Therefore, this method has not been adopted in gastric surgery for a long time because of fear that early food intake may cause anastomotic leakage due to direct stimulation of anastomotic sites and increased intraluminal pressure. Until 2004, it was reported for the first time by Suehiro et al. that patients undergoing gastrectomy can accelerate their recovery after postoperative early oral intake 7 . In the following decade, more and more studies have shown that early oral feeding after gastrectomy is safe and feasible [8][9][10][11] . In July 2014, the European Association of Accelerated Rehabilitation Surgery first developed and released the guide to Accelerated Rehabilitation for Gastric Resection Surgery (hereinafter referred to as the Guide), which states that patients who underwent total gastrectomy can be administered the postoperative oral diet according to their tolerance as of POD 1 13 .
In this case,due to the lack of similar studies in China, we conducted a retrospective study to assess the safety and efficacy of early oral feeding and proposed a versatile ERAS protocol for accelerated recovery of Chinese patients after gastric surgery. Overall, this study provides satisfactory short-term clinical outcomes and we strongly believe that patients undergoing gastrectomy should be encouraged to begin the early oral feeding carefully and adjust the feeding schedule based on their tolerability. In our study, the primary outcome was a postoperative complication rate of 26%, and a severe 3% (Clavien- . It is comparable to the rates in other reports 8,11,37,38 . It is worth noting that Clavien-Dindo classification of surgical complications can compensate for other reports that are graded according to different criteria. These reports only identify severe complications, and  Table 10. Relationship between potential factors and early oral feeding intolerance -univariate logistic regression analysis. CI, confidence interval; OR, odds ratio; if the p-value is less than 0.05, it is in bold and flagged with two stars (**). (2019) 9:16539 | https://doi.org/10.1038/s41598-019-52629-0 www.nature.com/scientificreports www.nature.com/scientificreports/ often ignore complications that do not require intervention 27 . In addition, in our patient group, 68% of patients were discharged on POD 6 and all patients had a median postoperative hospital stay (IQR) of 6 (3) days. Similarly, according to the study by Hur and Colleagues, that was 8.03 ± 1.43 days in the early oral feeding group and 9.9 ± 2 days in the control group 39 . In addition, Jeong and colleagues reported that in their study, hospital stay was 7.4 days in the early feeding group and 8.9 days in the control group 11 . Although there is no comparison done between the early oral feeding group and a control group in our study, the duration of postoperative stay was shortened compared to the results from the above-mentioned two studies and other studies as well 21,37,38 . All of these results indicate that patients who underwent gastrectomy had shorter postoperative hospital stays without increasing postoperative complications 11 .
Although many studies have applied early oral feeding for patients after gastric surgery, and reported improvements in postoperative outcomes and accelerated recovery, the actual implementation of early oral feeding varied from one to the other. To date, there is still no consensus on appropriatetiming, composition, frequency and volume or amount of early oral feeding. We performed an early oral feeding of 204 patients undergoing colorectal surgey (adjusted according to the dietary habits of local Chinese residents), and more than 90% of them were tolerated (unpublished data). Our previous data also found that patients who underwent gastrectomy were less tolerant to oral feeding than patients undergoing colorectal surgery. Based on our past experience with early oral feeding programs for patients undergoing colorectal surgery, combined with other early oral feeding schedules found in recent studies, we developed a stepwise, Chinese patient-specific early oral feeding plan from water to other liquids, followed by semi-fluids, and finally have a normal diet. We initiated the oral intake at the target volume within 24 hours of surgery. According to our findings, 77% of the patients were able to adhere to the oral feeding schedule without causing any adverse events, and this is moderate in the tolerability (57-93%) after gastric cancer surgery reported in previous studies 10,11,21,38 . In addition, according to the dietary habits of our local patients, we recommend the following dietary guidelines for discharged patients: from the discharge time to 3 weeks after surgery, eat 6 to 8 semi-liquid diets a day, and each time the calorie is 200~250 kcal; 3~4 weeks after the operation, gradually transit to a normal diet, eat 6 times a day with the calorie of each meal between 250 and 300 kcal.
In theory, it is assumed that the age of the patient, the extent of tumor resection, the surgical approach, and the stage of the tumor allaffect early oral feeding tolerance in postoperative gastric cancer patients. For instance, a recent study by Shimizu et al. supported that compared to the distal gastrectomy (DG) groups, the postoperative stay was significantly shorter in the total gastrectomy (TG) groups, and the incidence of postoperative complications after early oral feeding supplementation was lower 40 . However, no differences were found in the duration of postoperative stay between early oral feeding and normal groups of the patients receiving distal gastrectomy (DG). But a high incidence of postoperative complications was found in the early oral feeding groups of patients with DG 40 . This report seems to indicate that the extent of resection (rather than surgical approach) may shorten the postoperative stay and reduce postoperative complications as beneficial results of early oral feeding. Nevertheless, as the authors say, the above conclusions are far from confirmed due to insufficient sample size in the TG group, which requires further researches.
Interestingly, our results demonstrated that laparoscopic gastrectomy can reduce the chance of early oral feeding intolerability compared with patients undergoing an open operation, with an Odds Ratio of 0.30. However, due to the small sample size this estimated OR is not precise enough, reflected by a much wider range of 95% CI (0.11-0.85). Also because of that, we realized that there is no point in conducting a further stratified comparison analysis between open or laparoscopic operation groups because of the limited number of cases of sub-or total gastric resection within each group. Therefore, caution should be exercised when expanding this conclusion. A complete model with all potential factors considered was fitted but it failed to reliably estimate the parameters and thus it won't be presented in this paper. We propose to conduct a multi-center case-control study with a much larger sample size in the future and further corroborate our findings. In addition, it is worth pointing out that both postoperative fluid restriction regimen and postoperative mobilization protocol may also affect early oral feeding. Thus, it is necessary to conduct further research on this to better address the above-discussed factors affecting early feeding tolerance before we can improve our early oral feeding schedule to better promote the rehabilitation of patients. Additional analysis of the immunological functions of postoperative gastric cancer patients will also be needed to optimize the early oral feeding schedules.

Conclusion
In conclusion, this retrospective study concluded that our ERAS protocol incorporating a stepwise, Chinese patient-specific early oral feeding schedule was also safe and beneficial compared to similar overseas studies in the past. It can shorten the postoperative hospital stays without increasing the incidence of postoperative complications. In addition, as far as we know, this is the first domestical report on detailed and most importantly suitable early feeding plan for Chinese patients after gastric surgery. Therefore, although the study has some limitations, such as small sample size and lack of control studies, it still deserves further research and promotion. Finally, we recommend that a customized, optimized oral feeding plan should be developed for local Chinese patients with their dietary habits, races, and other factors considered. Together with fine-tuning of other essential elements of the current ERAS protocol, it is foreseeable that a more effective ERAS protocol will be offered in the near future, includinging an optimized early oral feeding schedule for gastric cancer surgery.

Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable requests.