Introduction

Gastric cancer is the fourth most common cancer and the second most common cause of cancer-related death globally [1]. In Japan, gastric cancer is the third most common cause of cancer-related death, following lung cancer and colorectal cancer. Early tumor detection; curative surgical resection, including extended lymph node dissection (D2 or D3); and appropriate adjuvant therapy have improved the survival rate of patients with primary gastric cancer. However, surgery for gastric cancer is still a high-risk procedure involving clinically significant postoperative stress, complications, and sequelae. In addition, radical gastrectomy–related morbidity and mortality range from 10 to 46% and from 0 to 13%, respectively [2,3,4,5,6,7,8].

A postoperative nutritional status in the patient undergoing a major abdominal surgery is a major factor determining the patient’s surgical outcome and might be improved by early postoperative enteral feeding [9, 10]. In addition, appropriate perioperative management can help decrease the morbidity rate. In the 1990s, to enhance postoperative recovery and reduce morbidity in the field of elective colorectal surgery, Henrik Kehlet proposed the concept of multimodal perioperative care, such as enhanced recovery after surgery (ERAS) and fast-track surgery (FTS) [11, 12]. ERAS and FTS rapidly gained popularity among surgeons globally.

ERAS programs comprise many elements, such as (i) preoperative education, (ii) preoperative carbohydrate loading, (iii) omission of bowel preparation, (iv) epidural analgesia without opioids, (v) early postoperative enteral feeding, (vi) early mobilization of patients, and (vii) thromboprophylaxis. All these elements are independent but directed toward the same two goals: reducing surgical stress and optimizing recovery [13].

Early oral feeding (EOF) is the most important part of ERAS. Several clinical randomized controlled trials (RCTs) have demonstrated the advantages of EOF after elective colorectal cancer surgery, such as reduced length of postoperative hospital stay and reduced postoperative morbidity and mortality compared with traditional postoperative oral feeding [9, 14,15,16]. However, the feasibility of EOF for gastric cancer surgery patients remains unclear because traditionally EOF is avoided to minimize strain to the anastomoses and reduce the inherent risks of postoperatively impaired gastrointestinal motility. As a result, many institutions in Japan still implement a period of 3–4 days of postoperative fasting, followed by subsequent stepwise slow dietary rehabilitation.

Recently, the Japanese conventional fixed dietary schedule after radical gastrectomy was reexamined, and many institutions in Japan are now adopting early postoperative oral intake. However, little data support the feasibility and safety of EOF after gastrectomy. Therefore, this study aimed to evaluate the feasibility and safety of EOF for patients with gastric cancer after radical gastrectomy. In addition, we retrospectively analyzed the risk factors for dropout from the EOF protocol.

Materials and Methods

Patients

We retrospectively reviewed the medical records of 463 consecutive patients with gastric cancer who had undergone curative gastrectomy between 2005 and 2014 at the Department of Digestive Surgery, National Hospital Organization Hokkaido Cancer Center, Japan. Patients who underwent emergency surgery for complications such as perforation or/and bleeding were excluded from the study. Patients who had poor performance status, gastric carcinoma in the remnant stomach, esophagus invasion gastric cancer, or other synchronous malignant diseases and multiple organ resections were also excluded. Finally, of these 463 patients, the EOF protocol was implemented in 397 patients undergoing Roux-en Y reconstruction after radical gastrectomy (86%). The protocol was common in 277 (277/283, 98%) patients after distal gastrectomy (DG) and 120 (120/180, 67%) patients after total gastrectomy (TG). All 397 patients were retrospectively analyzed for the incidence of postoperative complications and surgical outcomes. This study was reviewed and approved by the research ethics committee of National Hospital Organization Hokkaido Cancer Center (approval no. 30-102). Informed consent has been obtained from all patients included in this study.

Schedule of the EOF Protocol

All 397 patients were scheduled for immediate nasogastric tube removal postsurgery in the operation room. They were also scheduled to start drinking water in the morning of the first postoperative day (1POD), eat thin rice gruel with a liquid nutrition supplement on the second postoperative day (2POD), and gradually progress to regular rice porridge and solid food over a period of 4 days, as tolerated. The discharge criteria were as follows: adequate pain relief, no high fever, no inflammatory signs, ability to mobilize with self-care, and tolerance for more than half of the solid food intake, including liquid nutrition supplement.

Data Collection: End Points

All data were retrospectively retrieved from the patients’ medical records. The primary end point was defined as the incidence of postoperative complications. Complications were defined as ≥ grade 2 complications according to the Clavien–Dindo classification within 30-day postsurgery. Secondary end points were defined as the incidence of anastomotic leakage, pneumonia, the percentage of patients starting oral feeding from 2POD, completion rate of the EOF protocol, length of postoperative hospital stay, 30-day postdischarge readmission rate, and 30-day postoperative mortality rate.

Meal step-up was delayed for patients who ate < 30% of meals for 2 days. Dropout from the EOF protocol was defined as stopping of meals for > 2 days because of postoperative observation of the patients’ condition and postoperative complications.

Pathological findings were categorized on the basis of the TNM Classification of Malignant Tumours (7th edition) of the Union for International Cancer Control. Continuous data were expressed as medians (range).

Statistical Analysis

All statistical analyses were performed using EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), a graphical user interface for R version 2.13.0 (R Foundation for Statistical Computing, Vienna, Austria). This modified version of R commander version 1.6-3 was designed to add statistical functions frequently used in biostatistics. Risk factors for dropout from the EOF protocol were identified using univariate and multivariate analyses with cross-linked tables and logistic regression models. P < 0.05 was considered statistically significant.

Results

Patients’ Characteristics

Table 1 summarizes the preoperative clinical characteristics of all 397 patients (251 men and 146 women). The mean age of the patients was 67 years, 240 patients (60.7%) had some comorbidity, and 47 patients (11.8%) received preoperative chemotherapy. Table 2 shows our surgical and oncological findings. Laparoscopic surgery was performed on 73 patients (18.4%), 52 patients (13.1%) underwent combined organ resection excluding cholecystectomy, and 245 patients (61.7%) received extended lymph node dissection (D2 or D3). The mean operating time was 255 min. The final pathologic examination showed that 198 patients (78.6%) were TNM stage I, 63 (15.9%) were TNM stage II, 70 (17.6%) were TNM stage III, and 66 (16.6%) were TNM stage IV. In addition, 52 patients (13.1%) had a noncurative resection.

Table 1 Baseline clinical characteristics of the patients
Full size table
Table 2 Surgical and oncological factors of the patients
Full size table

Postoperative Outcomes

Tables 3 and 4 show postoperative outcomes and details of postoperative complications, respectively. The nasogastric tube could be removed for all patients, and all patients started drinking water from 1POD, while 26 patients (15 DG patients [5.4%] and 11 TG patients [9.2%]) were unable to start oral intake because of abdominal distention (DG11, TG3), hemorrhage (DG3, TG2), poor activities of daily living (ADLs) (TG2), or fever (TG1). The EOF protocol was implemented in 371 patients (93%), of which 48 patients (DG35, TG13) stopped eating because of remnant stomach dilatation (DG13), abdominal distention (DG6, TG2), intra-abdominal abscess (DG8, TG2), hemorrhage (DG3, TG3), aspiration pneumonia (DG2, TG2), or fever (DG1, TG1). Overall, 227 patients (87%) after DG and 96 patients (88%) after TG were able to follow the EOF protocol perfectly, and the EOF protocol completion rate was 81.4%. The incidence of postoperative complications (≥ Clavien–Dindo grade 2 complications), including anastomotic leakage (n = 0), ileus (n = 22), reoperation (n = 3), relaparotomy hemostasis (n = 5), endoscopic hemostasis (n = 3), intra-abdominal abscess (n = 19), and pneumonia (n = 11), was 12.3%. In addition, the incidence of clinically significant events (≥ grade 3 complications) was only 4.3% despite a high adaptation percentage of the EOF protocol. We performed reoperation in 10 patients (2.5%). The median length of postoperative hospital stay was 20 days, and 6 patients were readmitted to the hospital within 30 days postdischarge. The 30-day postoperative mortality rate was 0%.

Table 3 Postoperative complications
Full size table
Table 4 Postoperative outcomes
Full size table

Risk Factors for Dropout from the EOF Protocol

In total, 74 patients (18.6%) dropped out from the EOF protocol. Table 5 shows the results of univariate analyses performed to determine which variables were associated with dropout from the EOF protocol. We found that the rate of dropout from the EOF protocol significantly increased in male patients, the elderly, and patients with any comorbidity. No significant differences were observed between the two patient groups in terms of preoperative chemotherapy and the body mass index. In addition, with regard to surgical factors, we found no significant differences in terms of surgery time, curability, type of approach (laparoscopy or open gastrectomy), extent of gastric resection (DG or TG), and extent of lymph node dissection (D2 or D3). However, pathologic stages and the amount of intra-operative bleeding significantly differed between the two groups.

Table 5 Univariate and multivariate analyses to determine clinicopathologic factors which effected on the drop from early oral feeding clinical pathway
Full size table

Multivariate analysis showed that the male gender (P = 0.025; odds ratio [OR] = 0.501), comorbidities (P = 0.033; OR = 0.523), and intra-operative bleeding (P = 0.041; OR = 0.565) were independent risk factors for dropout from the EOF protocol.

Discussion

The aim of the present study was to evaluate the safety, efficacy, and outcome of the EOF protocol employed in the perioperative treatment of radical gastrectomy for gastric cancer. This study was a retrospective, one-arm, nonrandomized analysis. However, the subjects included the elderly and also those who underwent preoperative chemotherapy. On the other hand, most previous RCTs were performed on selected patients who had good performance status and patients were excluded if they had advanced gastric cancer, malnutrition, or any important vital organ comorbidity.

In this study, the EOF protocol implemented in patients with gastric cancer after radical gastrectomy was common regardless of the extent of stomach resection, and the rate of use of the EOF protocol was high in 87% of all patients; especially, the rate was 98% in DG patients. Therefore, a common EOF protocol can be implemented in patients undergoing DG and TG routinely.

As mentioned before, previous studies have reported that morbidity and mortality from radical gastrectomy with conventional perioperative care range from 10 to 46% and from 0 to 13%, respectively [2,3,4,5,6,7,8]. Our data was good compared with these reported complication rates of conventional perioperative care study without the ERAS program. Some studies have advocated the benefits and feasibility of the EOF protocol after radical gastrectomy. In 2004, Suehiro et al. [17] first reported that implementing the EOF protocol within 48 h of radical gastrectomy is safe, with no evidence of increased postoperative morbidity and mortality, including nausea, vomiting, and anastomotic leakage, compared with conventional postoperative fasting. Recently, Hur et al. [18] reported in a small randomized trial that the EOF protocol on 2POD significantly enhances bowel recovery, reduces the length of postoperative hospital stay, and improves some elements of the patients’ quality of life. Wang et al. [19] introduced the EOF protocol as part of the FTS program in patients with gastric cancer and found that the length of postoperative hospital stay and medical costs reduced and the patients’ quality of life improved. Other studies have also reported the usefulness of ERAS programs for radical gastrectomy [20,21,22,23]. The incidences of postoperative complications in these reports (morbidity, 8–20% and mortality, 0%) were similar to our results and were acceptable. Feng et al. [24] and Sierzega et al. [25] analyzed the potential applicability of the EOF protocol in TG patients. Both studies concluded that the EOF protocol reduces the incidence of complications compared with the conventional late oral feeding (morbidity, 10% vs. 28%; mortality, 20% vs. 36%). Our results and the findings of these studies showed that implementing the EOF protocol after radical gastrectomy for gastric cancer is feasible in terms of safety.

Historically, surgeons believed that decompression with a nasogastric tube and a fasting period of 2–5 days after gastrointestinal surgery was required to prevent anastomotic leakage and aspiration pneumonia. This fasting period was deemed necessary at the time of transient ileus postsurgery and to ensure protection of the anastomotic site [26]. Therefore, in this study, we set the incidence of anastomotic leakage and pneumonia as secondary end points. As reported, the incidence of anastomotic leakage was none-to-low, similar to previous studies (0.8–1.9%) based on conventional perioperative management [5, 7, 8]. Some studies evaluated the value of ERAS programs in a Japanese high-volume center and reported low incidence of anastomotic leakage (0.8–1.5%) [21, 27], same as in this study.

Another potential disadvantage of the EOF protocol is aspiration pneumonia. Sunil et al. [28] reported a 4.6% incidence of respiratory events following D2 gastrectomy without the ERAS program, but in this study, we observed respiratory events in only 3% of the patients (12/397, including 4 patients with aspiration pneumonia). Previous studies on ERAS protocols for colorectal cancer have shown a 7.8% incidence of respiratory complications [29]. Similarly, Feng et al. [24] reported pneumonia in 8.5% of TG patients who received the EOF protocol. However, other studies have reported no pneumonia after initiation of the EOF protocol after radical gastrectomy [18, 19, 21, 27]. Therefore, we concluded that the incidence of aspiration pneumonia does not increase because of the EOF protocol after radical gastrectomy.

Traditionally, many surgeons often left nasogastric tubes in for several days until the first flatus after gastric resection. Recent studies comparing nasogastric decompression versus no decompression demonstrated that the nasogastric tube might induce pulmonary complications after gastric cancer surgery and prolong the time to first flatus with no difference in the anastomotic leakage rate [30–31]. In this study, we could immediately remove the nasogastric tube postsurgery, and only 18 patients (4.5%) required insertion of a nasogastric tube. We recommend the practice of not routinely using nasogastric tubes.

In this study, we determined several risk factors for dropout from the EOF protocol. Suni et al. [28] reported that the male gender is a risk factor due to associated smoking, which can predispose the patient to respiratory events. In this study, there were more male patients with some comorbidities than female patients (89/251 male patients, or 64.5%). Several studies have reported that age is a risk factor for dropout of the EOF protocol [22, 32]. In addition, Choi et al. [32] suggested that the extent of gastric resection, operative approach, and combined resection could be significant risk factors for dropout of the EOF protocol. These results were not different from our analysis. Therefore, we could use the EOF protocol regardless of the extent of gastric resection and surgical approach.

Conclusions

EOF can be safely implemented in patients with gastric cancer after radical gastrectomy regardless of the extent of stomach resection. The male gender, comorbidities, and intra-operative bleeding are independent risk factors for dropout from the EOF protocol.