Prospective study of the effect of ERAS on postoperative recovery and complications in patients with gastric cancer

Patients and randomization

This prospective study was conducted in the advanced ward of Tianjin Medical University Cancer Institute & Hospital from May 2016 to March 2018. Only those patients undergoing radical resection of gastric cancer in the same treatment group were enrolled in the study. After obtaining written consent from all patients, we randomly assigned them into the ERAS (intervention) and non-ERAS (control) groups using a completely randomized design with a 1:1 allocation (gender, age, and TNM stage were used for the allocation). The Central Randomization System (CRS) was used by our staff to screen and randomize the patients. When participants were enrolled, after entering the screen number and the individual’s information, the computer-based randomized number was determined using the Internet, and the allocated group was retrieved on the website of the CRS. The investigators who were responsible for enrollment and intervention, and the responsible physicians, did not participate in the randomization process.

Inclusion criteria

Inclusion criteria were as follows: (1) patients who received radical gastrectomy; (2) ages of 20–75 years, without gender limits; (2) patients who had complete clinical data; (3) patients who objectively described their symptoms and actively cooperated; and (4) written informed consent from each patient.

Exclusion criteria

Exclusion criteria were as follows: (1) refusal to sign the consent form; (2) pregnancy or lactating female patients; (3) known allergy to propofol, desflurane, or any other anesthetic agent; (4) patients who received upper abdominal surgery; (5) patients with a severe mental disorder; (6) patients with a history of previous neoadjuvant chemotherapy, radiotherapy, or clinical trial treatment within 3 months; (7) patients who actively participated in another trial where the primary endpoint follow-up was ongoing; (8) patients with complications (bleeding, perforation, or obstruction) caused by gastric cancer; (9) patients with other major medical illnesses of the cardiovascular, respiratory, or immune systems; and (10) patients with severe liver and renal dysfunctions (Child-Pugh ≥ 10; creatinine clearance < 25 mL/min).

Interventions

During the perioperative period, the 40 patients in the ERAS group were managed in accordance with the ERAS protocol that we optimized¹⁵, which included preoperative counseling and education about the ERAS program, no smoking or drinking, no bowel preparation, a normal diet until 6 h before surgery, liquid intake until 2 h before surgery, preoperative carbohydrate loading before surgery (100 g glucose/1,000 mL water taken orally at 10 PM on the evening before the surgery and 50 g glucose/500 mL water taken 2–3 h preoperatively). Anesthesia consisted of a combination of epidural analgesia and general anesthesia. To prevent hypothermia, a blanket warming system and warming set for intravenous infusions were used. To prevent postoperative pain, a continuous thoracic epidural infusion of analgesics was given until 2 days after surgery. A non-steroidal anti-inflammatory drug was regularly used to prevent wound pain. Patients started progressive oral feeding and goal-oriented ambulation in postoperative days 1–4. Patients measured the distance of ambulation by the marker in the ward and reported it to the study nurse. If the urinary catheter was not obstructed, it was removed after completion of bladder training. Patients were then encouraged to continue and prolong out-of-bed activities. The remaining 40 patients (non-ERAS group) were intervened by conventional perioperative management as controls. The general interventions in both groups are listed in Table 1.

Discharge criteria

Discharge was recommended when the patient met the following criteria: (1) restoring a semi-liquid diet or oral supplementary nutrition; (2) no need for intravenous fluid therapy; (3) no pain or the pain could be controlled with oral analgesics; (4) normal body temperature in the previous 24 h; (5) the patient could freely perform out-of-bed activities; (6) no wound problems; and (7) the patient agreed to be discharged¹⁶.

Subgroup analysis

Subgroup analysis was assigned based on the BMI level. China has recommended that BMI ≥ 28 kg/m² is considered as obesity. Moreover, our study purpose was to evaluate the efficacy of ERAS between obese and non-obese patients undergoing gastric cancer surgery, so we set 28 kg/m² as the cut-off value for subgroup analysis. The 40 patients in the ERAS group were divided into 2 subgroups according to their BMI levels, namely group A (BMI < 28 kg/m², n = 16) and group B (BMI ≥ 28 kg/m², n = 24). Patients in the non-ERAS group were divided in the same way, involving group C (BMI < 28 kg/m², n = 18) and group D (BMI ≥ 28 kg/m², n = 22), respectively.

Visual Analogue Scale (VAS) score

The VAS is commonly used to measure panic, depression, fatigue, and pain¹⁷, and is usually 100 mm in length with anchor descriptors such as “no pain” and “worst pain imaginable.” In the present study, the VAS was reported in centimeters, i.e., on the scale of 0–10 to evaluate postoperative pain. The evaluation was conducted 12 h after surgery.

Statistical analysis

Statistical analysis was performed using SPSS statistical software for Windows, version 23.0 (SPSS, Chicago, IL, USA). Measurement data were compared between groups using Pearson’s chi-squared test or Fisher’s exact test. Enumeration data were compared between groups using the independent Student’s t-test. The normality of data was analyzed using the Kolmogorov-Smirnov test. Statistical significance was set at P < 0.05.

Baseline characteristics

A total of 80 eligible patients were recruited. The ERAS and control groups comprised 25 and 26 males (P = 0.816), respectively. There was no significant difference in TNM stages (I/II/III/IV,7/13/20/0 vs. 8/13/19/0, P = 0.955) and type of surgery (distal gastrectomy/total gastrectomy/proximal gastrectomy, 30/7/3 vs. 29/30/2, P = 0.792) on enrollment. Smoking (P = 0.502) and drinking (P = 0.499) were similar between groups. The mean BMI of the ERAS and non-ERAS groups were 30.7 kg/m² and 30.6 kg/m², respectively (P = 0.651). Baseline characteristics of recruited patients are shown in Table 2.

Postoperative recovery of patients in the ERAS and non-ERAS groups

We analyzed the differences in postoperative recoveries between the ERAS and non-ERAS groups. After determining the normality of the data (P > 0.05), the Student’s t-test was used to compare the data. Patients in the ERAS group had a shorter length of stay, faster postoperative exhaustion and defecation, faster recovery of leukocyte counts, shorter duration of intravenous nutrition, and lower VAS score on the postoperative first day, when compared with those in the non-ERAS group (all, P < 0.05); only the time to stopping decline of albumin had no significant difference between the groups (P = 0.346). The results showed that use of ERAS during the perioperative period significantly triggered postoperative recovery of patients undergoing gastric cancer surgery (Table 3).

Subgroup analyses of the BMI

Patients were allocated into 4 groups depending on their BMI levels; in the ERAS group: group A (BMI < 28 kg/m², n = 16), group B (BMI ≥ 28 kg/m², n = 24), and in the non-ERAS group: group C (BMI < 28 kg/m², n = 18), group D (BMI ≥ 28 kg/m², n = 22). We analyzed the 4 groups in pairs, with the results summarized in Tables 4 and 5. Compared with the patients in group C, perioperative intervention of ERAS in patients whose BMI < 28 kg/m² (group A) achieved better postoperative recovery, showed a shorter length of stay, faster postoperative exhaustion and defecation, faster recovery of leukocyte counts, shorter duration of intravenous nutrition, and a lower VAS score on the postoperative first day (all, P < 0.05).

Compared with patients in group D, perioperative intervention of ERAS in patients whose BMI < 28 kg/m² (group B) showed better postoperative recovery, a shorter length of stay, faster postoperative exhaustion and defecation, faster recovery of leukocyte counts, shorter duration of intravenous nutrition, and lower VAS score on the postoperative first day (all, P < 0.05).

The comparison of group A vs. group B showed that group A had a shorter length of stay and faster postoperative defecation and lower VAS score on the postoperative first day than those of group B. In a similar manner, patients in the non-ERAS group with BMI < 28 kg/m² (group C) showed faster recovery of leukocyte counts and shorter duration of intravenous nutrition than those of group D (P < 0.05).

Postoperative complications

There were 13 cases of intermuscular deep vein thrombosis of the lower extremities, 1 case in group A, 2 cases in both groups B and C, and 8 cases in group D. Two patients developed postoperative lung infections in group D. The incidence of intermuscular deep vein thrombosis of lower extremities was significantly higher in group D than in group B (χ² = 4.800, P = 0.028), suggesting that perioperative intervention of ERAS in patients with high BMI (BMI ≥ 28 kg/m²) significantly decreased the incidence of intermuscular deep vein thrombosis of the lower extremities. A higher incidence of postoperative intermuscular deep vein thrombosis of lower extremities and lung infection were found in group D, when compared with those of other groups. Other complications such intestinal obstruction, leakage, and anastomosis failure were not observed in any groups (Table 6), indicating that implementation of ERAS in patients with high BMI had advantages in reducing postoperative complications.