Abstract
Purpose
This study aimed to compare clinical, nutritional, and endoscopic outcomes among Roux-en-Y (RY), uncut Roux-en-Y (uRY), and Billroth II with Braun anastomosis (BB) reconstruction methods in gastric cancer patients after distal gastrectomy.
Methods
This retrospective study analyzed gastric cancer patients who underwent laparoscopic distal gastrectomy between January 2018 and December 2022, categorized into RY, uRY, and BB. Demographic, pathological, clinical, nutritional, and endoscopic data were collected over 24 months. Endoscopic outcomes were assessed using the RGB scoring system, evaluating residual food, gastritis, and bile reflux. Delayed gastric emptying was defined by clinical and imaging criteria. Multivariable linear regression was performed to identify relevant factors associated with 24 months RGB scores and mixed linear model was applied to assess the time interaction.
Results
A total of 221 patients were included (70 RY, 75 uRY, 76 BB). Baseline characteristics, perioperative and nutritional outcomes were comparable among groups. The uRY group showed less weight loss at 3 months (5.9%, P<0.05) but did not differ in other studied periods. The RY group had the lowest RGB scores at all time points, while BB showed the highest and progressively worsening scores. Regression model showed that BB and uRY was related to increase 24 month RGB score. BB showed increased RGB score while RY was stable in time and uRY showed intermediate change in the mixed linear model (P<0.05).
Conclusion
The nutritional clinical outcomes were comparable between the reconstructions. RY reconstruction demonstrated favorable endoscopic outcomes, while BB was associated with higher RGB scores in time.
Keywords: General surgery, Stomach neoplasms, Endoscopy
INTRODUCTION
Most gastric cancers arise in the lower third of the stomach, and distal gastrectomy is the standard treatment for gastric cancer [1–3]. However, the optimal reconstruction method following distal gastrectomy remains a matter of ongoing debate, particularly with regard to functional outcomes.
Among several reconstruction options, Roux-en-Y (RY) reconstruction is widely recommended for its efficacy in preventing alkaline reflux gastritis, esophagitis, and dumping syndrome [4]. Nonetheless, disruption of the myoneural continuity between the duodenal pacemaker and the Roux limb may lead to Roux stasis syndrome, characterized by nausea, vomiting, postprandial abdominal discomfort, and, in severe cases, delayed gastric emptying (DGE) [5].
Billroth II with Braun anastomosis (BB) reconstruction is technically straightforward and may be favored due to its minimal neural disruption and potential to preserve normal pacemaker-mediated conduction. The added jejunojejunostomy may facilitate bile diversion; however, its clinical benefit remains unclear, and bile reflux can still occur despite this anatomical modification.
Uncut Roux-en-Y (uRY) reconstruction has been proposed as a hybrid technique that preserves bowel and mesenteric continuity while functionally occluding the afferent limb—mimicking the configuration of BB with additional bile diversion through a distal jejunal bypass [6,7]. Previous studies have shown that uRY is technically feasible, safe, and applicable in minimally invasive settings [8,9]. However, its long-term efficacy has not been fully established.
In this study, we aimed to compare three reconstruction methods; RY, uRY, or BB reconstruction following distal gastrectomy regarding clinical, nutritional and endoscopic findings including residual food, gastritis, and bile reflux (RGB), and to determine which method offers better outcomes.
METHODS
Study design and patients
This retrospective study included patients diagnosed with early gastric cancer who underwent laparoscopic distal gastrectomy at a regional cancer center of a tertiary hospital between January 2018 and December 2022. Clinical data including demographic information, operative details, pathological findings, and postoperative follow-up (abdominal computed tomography and esophagogastroduodenoscopy) were reviewed. Among 291 patients who underwent distal gastrectomy, we excluded those who underwent other reconstructions (n=41), combined organ resection (n=10), or were lost to follow-up or had missing endoscopic data (n=19). Patients were classified into three groups according to the reconstruction method: RY (n=70), uRY(n=75), and BB (n=76) (Fig. 1).
Fig. 1.
Flow diagram of patient selection. A total of 221 patients were included and classified into three groups according to the reconstruction method: conventional Roux-en-Y (RY; n=70), uncut Roux-en-Y (uRY; n=75), and Billroth II with Braun anastomosis (BB; n=76).
Operative procedures
Three surgeons who had completed their gastric cancer specialist certification participated in this study. None of the surgeons were restricted to a specific reconstruction method and all were capable of performing all three procedures. All procedures were performed laparoscopically under general anesthesia in the reverse Trendelenburg position. After gastric transection and lymph node dissection using a laparoscopic linear stapler, the resected specimen was extracted through a 5-cm umbilical incision. The extent of lymph node dissection (D1+ or D2) followed the national guidelines, based on tumor characteristics and intraoperative findings [2,3,10]. Three reconstruction techniques were employed as follows: (1) BB reconstruction: a side-to-side jejunojejunostomy (60 mm) was created using the jejunum 20–30 cm distal to the ligament of Treitz. A gastrojejunostomy (GJ) was performed 15 cm distal to the JJ. All reconstructions were antecolic and isoperistaltic; (2) RY reconstruction: the jejunum was transected 20–30 cm distal to the ligament of Treitz. The distal limb was anastomosed to the remnant stomach (GJ), and a jejunojejunostomy was created 20–40 cm distal to the GJ. The Roux limb length was approximately 20 cm. Petersen’s and mesenteric defects were routinely closed; or (3) uRY reconstruction: a GJ was constructed 50 cm distal to the ligament of Treitz. Jejunojejunostomy was made between the proximal jejunum (20 cm distal from Treitz) and the jejunum 20 cm distal to the GJ. The afferent limb was occluded using a 45-mm 6-row stapler placed 10 cm proximal to the GJ. The choice of reconstruction was at the discretion of the operating surgeon based on preference and experience.
Data collection and outcome measures
Patient demographic variables, including age, sex, body mass index (BMI), and Charlson Comorbidity Index, were collected. Clinicopathologic outcomes such as extent of lymphadenectomy, number of retrieved lymph nodes, pathological stage and early postoperative courses including postoperative hospital stay and time to first flatus. Complication data within 30 days after surgery were collected prospectively and was defined according to the Clavien-Dindo grading system.
Postoperative outcomes included hospital stay, time to first flatus, DGE, and complications. Nutritional status was assessed using body weight, hemoglobin, platelet count, serum albumin, total protein, and cholesterol at 3, 6, 12, and 24 months postoperatively.
DGE was defined as the presence of abdominal pain, bloating, or early satiety, with radiologic evidence of gastric distension and air-fluid levels. Endoscopic findings confirmed retained food without anatomical stenosis or angulation.
Endoscopic findings were graded using the RGB scoring system which comprises three components: residual food (grade 0–IV), gastritis (grade 0–IV), and bile reflux (grade 0–I) (Supplementary Table 1) [11]. The total RGB score ranged from 0 to 9. Evaluations were performed by experienced endoscopists at 6, 12, and 24 months to assess temporal progression and intergroup differences.
For the uRY group, recanalization was assessed through serial endoscopic evaluation. Partial recanalization was defined as endoscopic evidence of luminal continuity between afferent and efferent limbs.
Statistical analysis
Continuous variables were expressed as mean±standard deviation and compared using one-way analysis of variance or Kruskal–Wallis tests. Categorical variables were analyzed using chi-square or Fisher exact tests. To identify clinical factors associated with the 24-month RGB score, we performed a multivariable linear regression analysis. The dependent variable was the 24-month RGB score, and the independent variables included age, sex, BMI, operation time, tumor size, extent of lymph node dissection (D1+ vs. D2), and reconstruction method (RY, uRY, and BB). Regression coefficients are presented as both unstandardized coefficients (B) and standardized coefficients (β). The 95% confidence intervals (CI) refer to the unstandardized coefficients (B). Variables with P<0.05 were considered statistically significant. Model fit was evaluated using the coefficient of determination (R2) and F-statistics. To further assess the effect of time, reconstruction type and their interaction on RGB score, a mixed linear model was applied. Reconstruction group, time, and the interaction term were included as fixed effects, and subject was treated as a random effect. Estimated marginal means and 95% confidence intervals were reported. The trajectory was depicted in a figure to show temporal effect. All statistical analyses and trajectory figure were performed using SPSS version 28 (IBM Corp.). A P-value <0.05 was considered statistically significant.
RESULTS
Clinicopathological outcomes
Baseline characteristics were comparable among the three groups: RY (n=70), uRY (n=75), and BB (n=76). Mean age was 60.9±13.3 years in the RY group, 56.9±12.1 years in the uRY group, and 61.9±11.6 years in the BB group (P=0.109). The male-to-female ratio was similar in all groups (approximately 2:1; P=0.853). BMI and the Charlson Comorbidity Index also showed no difference between the groups.
Pathological variables showed no difference between the groups; tumor location, tumor size, number of retrieved lymph nodes, extent of lymph node dissection, and pathological staging showed no significant difference among the three groups (P=0.606). Most tumors were located in the lower part of the stomach, with a mean size of 2.4–2.9 cm (Table 1). Mean operation time did not differ among groups (213.7±99.9 minutes in RY, 210.0±54.3 minutes in uRY, and 220.4±61.2 minutes in BB; P=0.247). The length of hospital stay and (10.3±8.5 days in RY, 9.1±5.3 days in uRY, and 9.7±6.5 days in BB; P=0.131) the time to first flatus did not show difference (4.2±5.1 days in RY, 3.5±3.1 days in uRY, and 3.0±0.7 days in BB; P=0.744). The overall 30-day postoperative complication rates (15.7% in RY, 14.7% in uRY, and 19.7% in BB; P=0.679) and individual complications (P>0.05) did not show differences. There was no difference in DGE between the groups (P>0.05) (Table 2). The patient distribution of participating surgeons did not show a difference between the groups.
Table 1.
Demography and patient characteristics
| Characteristics | RY (n=70) | uRY (n=75) | BB (n=76) | P-value |
|---|---|---|---|---|
| Age (yr) | 60.9±13.3 | 56.9±12.1 | 61.9±11.6 | 0.109 |
|
| ||||
| Ratio of sex (M:F) | 2.0:1 | 2.5:1 | 2.2:1 | 0.853 |
|
| ||||
| Charlson Comorbidity Index | 0.107 | |||
| 0 | 57 (81.4) | 66 (88.0) | 56 (73.7) | |
| 1 | 13 (18.6) | 9 (12.0) | 19 (21.1) | |
| ≥2 | 0 | 0 | 1 (1.3) | |
|
| ||||
| Body mass index (kg/m2) | 23.4±2.8 | 23.0±3.1 | 22.9±3.5 | 0.227 |
|
| ||||
| Operation time (min) | 213.7±99.9 | 210.0±54.3 | 220.4±61.2 | 0.247 |
|
| ||||
| Tumor location (%) | 0.264 | |||
| Upper 1/3 | 2 (2.9) | 0 | 0 | |
| Middle 1/3 | 7 (10) | 5 (6.7) | 5 (6.6) | |
| Lower 1/3 | 61 (87) | 70 (93.3) | 71 (93.4) | |
|
| ||||
| Tumor size (cm) | 2.9±2.2 | 2.5±1.5 | 2.4±1.7 | 0.266 |
|
| ||||
| Retrieved LN number | 31.8±14.2 | 33.8±11.4 | 33.4±9.7 | 0.391 |
|
| ||||
| Extent of LN dissection (%) | 0.858 | |||
| D1+ | 55 (78.6) | 60 (80.0) | 58 (76.3) | |
| D2 | 15 (21.4) | 15 (20.0) | 18 (23.7) | |
|
| ||||
| pT stage (%) | 0.211 | |||
| T1a | 32 (45.7) | 34 (45.3) | 37 (48.7) | |
| T1b | 34 (48.6) | 30 (40.0) | 34 (44.7) | |
| T2 | 2 (2.9) | 10 (13.3) | 3 (3.9) | |
| T3 | 2 (2.9) | 1 (1.3) | 2 (2.6) | |
|
| ||||
| pN stage (%) | 0.595 | |||
| N0 | 68 (97.1) | 72 (96.0) | 75 (98.7) | |
| N1 | 2 (2.9) | 3 (4.0) | 1 (1.3) | |
|
| ||||
| Stage, AJCC 8th (%) | 0.606 | |||
| I | 69 (98.6) | 72 (96.0) | 75 (98.7) | |
| II | 1 (1.4) | 3 (4.0) | 1 (1.3) | |
| III | 0 | 0 | 0 | |
Values are presented as mean±SD or number (%).
RY, Roux-en-Y; uRY, uncut Roux-en-Y; BB, Billroth II with Braun anastomosis; LN, lymph node; AJCC, American Joint Committee on Cancer; SD, standard deviation.
Table 2.
Hospital course and Postoperative complications
| Variable | RY (n=70) | uRY (n=75) | BB (n=76) | P-value |
|---|---|---|---|---|
| Hospital stay (day) | 10.3±8.5 | 9.1±5.3 | 9.7±6.5 | 0.131 |
|
| ||||
| First flatus (day) | 4.2±5.1 | 3.5±3.1 | 3.0±0.7 | 0.744 |
|
| ||||
| Complication rate within 30 day | 11 (15.7) | 11 (14.7) | 15 (19.7) | 0.679 |
|
| ||||
| Pulmonary | 5 (7.1) | 6 (8.0) | 5 (6.6) | 0.944 |
|
| ||||
| Postop leakage | 3 (4.3) | 0 | 2 (2.6) | 0.215 |
|
| ||||
| Fluid collection | 2 (2.9) | 2 (2.7) | 0 | 0.343 |
|
| ||||
| Paralytic ileus | 1 (1.4) | 1 (1.3) | 0 | 0.588 |
|
| ||||
| Pancreas fistula | 0 | 1 (1.3) | 0 | 0.376 |
|
| ||||
| Wound | 0 | 0 | 1 (1.3) | 0.384 |
|
| ||||
| Bleeding | 0 | 0 | 1 (1.3) | 0.384 |
|
| ||||
| Delayed gastric emptying | ||||
| Immediate postoperative | 2 (2.9) | 3 (4.0) | 6 (7.9) | 0.335 |
| 6 mo after surgery | 6 (8.6) | 8 (10.7) | 7 (9.2) | 0.907 |
| 12 mo after surgery | 7 (10.0) | 7 (9.3) | 7 (9.2) | 0.985 |
| 24 mo after surgery | 6 (8.6) | 6 (8.0) | 7 (9.2) | 0.965 |
Values are presented as mean±SD or number (%).
RY, Roux-en-Y; uRY, uncut Roux-en-Y; BB, Billroth II with Braun anastomosis; SD, standard deviation.
Nutritional outcomes
Serial nutritional outcomes were evaluated at 3, 6, 12, and 24 months postoperatively. Body weight changes showed the most decrease at 3 months and gradually recovered in time. At 3 months, the uRY group showed the least weight loss (5.9%±6.8%), which was lower compared to the RY group (7.1%±6.8%, P<0.05) and the BB group (7.2%±5.5%, P<0.05). However, this difference reduced over time and no significant differences in weight loss was observed at 6, 12, or 24 months. Hemoglobin levels decreased postoperatively across all groups but gradually improved by 12 months. Platelet counts and changes in serum protein, albumin, and cholesterol showed fluctuations over time but no meaningful intergroup differences were shown (Table 3).
Table 3.
Nutritional outcome between reconstructions
| Variable | RY (n=70) | uRY (n=75) | BB (n=76) | P-value |
|---|---|---|---|---|
| Body weight change after surgery (%) | ||||
| 3 mo | 7.1±6.8 | 5.9±6.8 | 7.2±5.5 | a),b) |
| 6 mo | 6.1±6.8 | 5.7±7.8 | 6.5±7.5 | 0.171 |
| 12 mo | 5.7±8.7 | 5.8±8.3 | 6.4±8.7 | 0.123 |
| 24 mo | 5.3±8.6 | 5.4±8.3 | 5.7±8.0 | 0.193 |
|
| ||||
| Hemoglobin change (%) | ||||
| 3 mo | 7.1±16.8 | 7.0±14.5 | 7.7±13.6 | 0.202 |
| 6 mo | 6.1±15.5 | 6.6±17.2 | 7.3±18.6 | 0.157 |
| 12 mo | 5.7±14.7 | 5.6±23.2 | 6.8±26.9 | 0.226 |
| 24 mo | 4.6±13.8 | 5.2±18.2 | 6.4±22.3 | 0.118 |
|
| ||||
| Platelet change (%) | ||||
| 3 mo | 7.9±33.9 | 5.5±20.6 | 7.4±29.8 | 0.273 |
| 6 mo | 8.2±29.3 | 2.8±20.9 | 7.2±24.0 | 0.170 |
| 12 mo | 9.1±27.0 | 1.6±22.0 | 5.6±27.1 | 0.205 |
| 24 mo | 8.8±25.0 | 3.2±23.0 | 8.0±26.1 | 0.128 |
|
| ||||
| Total protein change (%) | ||||
| 3 mo | 0.5±15.7 | 0.2±7.7 | 0.6±11.2 | 0.215 |
| 6 mo | 0.5±11.4 | 0.9±8.6 | 1.0±10.6 | 0.154 |
| 12 mo | 0.8±10.7 | 0.6±7.7 | 0.6±13.2 | 0.314 |
| 24 mo | 0.6±11.9 | 0.5±8.8 | 0.8±11.2 | 0.222 |
|
| ||||
| Albumin change (%) | ||||
| 3 mo | 0.5±13.5 | 0.7±9.9 | 1.2±11.9 | 0.138 |
| 6 mo | 0.4±13.0 | 1.7±10.0 | 1.8±13.2 | 0.058 |
| 12 mo | 0.5±12.0 | 1.3±11.1 | 1.4±12.5 | 0.123 |
| 24 mo | 0.4±11.0 | 1.2±11 | 1.1±12.5 | 0.153 |
|
| ||||
| Cholesterol change (%) | ||||
| 3 mo | 14.8±19.7 | 12.2±15.3 | 6.2±21.5 | 0.348 |
| 6 mo | 2.2±24.3 | 7.0±25.8 | 3.2±23.2 | 0.078 |
| 12 mo | 3.0±27.3 | 5.4±20.4 | 3.5±23.0 | 0.153 |
| 24 mo | 1.9±25.8 | 2.8±19.5 | 2.4±22.0 | 0.205 |
Values are presented as mean±SD.
RY, Roux-en-Y; uRY, uncut Roux-en-Y; BB, Billroth II with Braun anastomosis; SD, standard deviation.
P<0.05 between uRY and RY;
P<0.05 between uRY and BB.
RGB scores
From the endoscopic assessment, each residual, gastritis, bile scores and overall RGB scores were collected at 6, 12, and 24 months postoperatively (Table 4). The mean residual food scores in the RY group were 0.25±0.55 at 6 months, 0.20±0.58 at 12 months, and 0.17±0.56 at 24 months. In the uRY group, the values were 0.21±0.48, 0.17±0.55, and 0.16±0.55, respectively. In the BB group, the scores were 0.29±0.30, 0.24±0.40, and 0.12±0.41, respectively. No statistical differences were observed among the groups at any time point (P=0.734, P=0.620, and P=0.799, respectively). For gastritis scores, the RY group showed mean values of 0.23±0.54 at 6 months, 0.36±0.72 at 12 months, and 0.41±0.77 at 24 months. In the uRY group, the values were 0.59±0.79, 0.84±0.94, and 0.83±1.06. In the BB group, the scores were 1.37±0.77, 1.67±0.96, and 2.53±1.18, respectively. The overall difference among groups was statistically significant at each time point (P<0.001). For bile reflux scores, the RY group showed mean values of 0.03±0.17 at 6 months, 0.04±0.20 at 12 months, and 0.03±0.17 at 24 months. In the uRY group, the values were 0.02±0.07, 0.07±0.25, and 0.11±0.39, respectively. In the BB group, the scores were 0.73±0.46, 0.77±0.49, and 0.82±0.39, respectively. The overall difference among groups was statistically significant at each time point (P<0.001). The mean overall RGB scores in the RY group were 0.46±1.07 at 6 months, 0.60±0.94 at 12 months, and 0.61±0.94 at 24 months and 0.73±0.99, 1.08±1.26, and 1.09±1.43 in the uRY group, respectively. In the BB group, the scores were 2.29±0.63, 2.55±1.30, and 3.46±1.31, which were consistently higher than those of the RY and uRY groups. The overall difference among the three groups was statistically significant at each time point (P<0.001).
Table 4.
Longitudinal changes in RGB scores by reconstruction type
| Score type | Time point | RY (n=70) | uRY (n=75) | BB (n=76) | P-value |
|---|---|---|---|---|---|
| RGB score | 6 mo | 0.46±1.07 | 0.73±0.99 | 2.29±0.63 | <0.001 |
| 12 mo | 0.60±0.94 | 1.08±1.26 | 2.55±1.30 | <0.001 | |
| 24 mo | 0.61±0.94 | 1.09±1.43 | 3.46±1.31 | <0.001 | |
|
| |||||
| Residual | 6 mo | 0.25±0.55 | 0.21±0.48 | 0.29±0.30 | 0.734 |
| 12 mo | 0.20±0.58 | 0.17±0.55 | 0.24±0.40 | 0.620 | |
| 24 mo | 0.17±0.56 | 0.16±0.55 | 0.12±0.41 | 0.799 | |
|
| |||||
| Gastritis | 6 mo | 0.23±0.54 | 0.59±0.79 | 0.87±0.77 | <0.001 |
| 12 mo | 0.36±0.72 | 0.84±0.94 | 1.07±0.96 | <0.001 | |
| 24 mo | 0.41±0.77 | 0.83±1.06 | 1.53±1.18 | <0.001 | |
|
| |||||
| Bile | 6 mo | 0.03±0.17 | 0.02±0.12 | 1.30±0.46 | <0.001 |
| 12 mo | 0.04±0.20 | 0.07±0.25 | 1.37±0.49 | <0.001 | |
| 24 mo | 0.03±0.17 | 0.11±0.39 | 1.82±0.39 | <0.001 | |
Values are presented as mean±SD.
RGB, residual food, gastritis, and bile reflux; RY, Roux-en-Y; uRY, uncut Roux-en-Y; BB, Billroth II with Braun anastomosis; SD, standard deviation.
Regression analysis
Multivariable linear regression analysis was conducted using the 24-month RGB score as the dependent variable (Table 5). The reconstruction method was the only factor significantly associated with the 24-month RGB score. Both BB (P<0.001) and uRY (P=0.049) were independently associated with higher RGB scores at 24 months compared with RY, whereas no other covariates (age, BMI, operation time, tumor length) or factors (sex, lymph node dissection) showed significant associations (all P>0.05). The mixed linear model showed significant main effects for Group (P<0.001) and Time (P<0.001), and a significant Group×Time interaction (P<0.001) (Table 6, Fig. 2). The BB group had the greatest increase in RGB scores over time, while RY maintained the lowest scores. The uRY group showed a gradual rise, plateauing at 12 months.
Table 5.
Multivariate linear regression analysis for RGB score
| Variable | B | SE | 95% CI | P-value |
|---|---|---|---|---|
| Age | −0.01 | 0.01 | −0.02 to 0.01 | 0.404 |
| Operation time | 0.00 | 0.00 | −0.00 to 0.00 | 0.895 |
| Body mass index | 0.01 | 0.03 | −0.04 to 0.07 | 0.635 |
| Tumor length | 0.05 | 0.04 | −0.03 to 0.13 | 0.261 |
| Female (vs. male) | −0.15 | 0.21 | −0.56 to 0.25 | 0.459 |
| D1+ Dissection (vs. D2) | −0.09 | 0.20 | −0.48 to 0.31 | 0.660 |
| RY (vs. BB) | −2.54 | 0.23 | 2.08 to 3.00 | <0.001 |
| uRY (vs. BB) | −0.83 | 0.23 | −1.28 to −0.38 | <0.001 |
The table summarizes the results of a linear regression assessing the relationship between clinical variables and the outcome. The BB group was used as the reference category for reconstruction method. Variables with P<0.05 were considered statistically significant.
RGB, residual food, gastritis, and bile reflux; SE, standard error; CI, confidence interval; RY, Roux-en-Y; uRY, uncut Roux-en-Y; BB, Billroth II with Braun anastomosis.
Table 6.
Fixed-effect estimates from the mixed linear model for RGB score
| Fixed effect | Estimate (β) | SE | 95% CI | P-value |
|---|---|---|---|---|
| Intercept (BB, 6 mo) | 2.29 | 0.14 | 2.02 to 2.56 | <0.001 |
| RY | −1.41 | 0.47 | −2.33 to −0.49 | 0.004 |
| uRY | −0.85 | 0.47 | −1.77 to 0.07 | 0.07 |
| Month: 12 mo | 0.27 | 0.20 | −0.13 to 0.66 | 0.181 |
| Month: 24 mo | 1.17 | 0.20 | 0.77 to 1.56 | <0.001 |
| RY×Month: 24 mo | −1.91 | 0.66 | −3.22 to −0.61 | 0.005 |
| uRY×Month: 24 mo | −0.61 | 0.66 | −1.91 to 0.69 | 0.351 |
The model included reconstruction group (BB as reference), time (6 months as reference), and group×time interaction as fixed effects, with subject ID as a random effect. Compared to BB, the RY group showed significantly lower RGB scores across time (β=−1.41, P=0.004), while the uRY group did not reach statistical significance (β=−0.85, P=0.07). A significant group-by-time interaction was observed for RY at 24 months (β=−1.91, P=0.005), indicating a more stable course over time.
RGB, residual food, gastritis, and bile reflux; SE, standard error; CI, confidence interval; BB, Billroth II with Braun anastomosis; RY, Roux-en-Y; uRY, uncut Roux-en-Y.
Fig. 2.
Fixed-effect estimates from the mixed linear model for RGB score. Error bars indicate 95% confidence intervals. Post hoc analysis showed that BB group had significantly higher RGB scores than RY and uRY at all time points (all P<0.001). The difference between RY and uRY was not significant at 6 months but became significant at 12 and 24 months (P<0.05). RGB, residual food, gastritis, and bile reflux; RY, Roux-en-Y; uRY, uncut Roux-en-Y; BB, Billroth II with Braun anastomosis.
DISCUSSION
This study compared clinical and nutritional outcomes and is the first to report endoscopic findings of RY, uRY, and BB reconstructions using the RGB score in patients who underwent laparoscopic distal gastrectomy. The RGB score was proposed by Kubo et al. [11] to standardize the assessment of the remnant stomach after gastrectomy. Bile reflux in the remnant stomach and esophagus can cause gastritis, esophagitis and esophageal cancer [12]. Experimental models support gastric bile reflux may act as a potential gastric carcinogen [13]. Beyond cancer risk, bile reflux is associated with symptoms such as heartburn and regurgitation, which impair quality of life and may require long-term medical therapy [9,14]. Gastric stasis, reflected by food residue, can also negatively affect dietary habits and nutrition, while chronic gastritis of the remnant stomach contributes to mucosal injury and patient discomfort. Recent multicenter data from the KLASS-07 database further demonstrated that higher bile reflux grades were directly associated with worse patient-reported outcomes, including nausea (QLQ-C30) and reflux symptoms (QLQ-STO22), indicating a close correlation between endoscopic findings and postoperative quality of life [15].
In our study, while perioperative outcomes and nutritional recovery were comparable among groups, long-term endoscopic were differed. The RY group maintained the lowest RGB scores, the BB group demonstrated the highest, and the uRY group showed intermediate results with early benefits that diminished during follow-up. In multivariable analysis, RY demonstrated a protective effect from an increase in the 24-month RGB score.
The gradual deterioration observed in the uRY group may be explained by recanalization of the uncut staple line. In our series, two cases were identified, and previous studies have reported partial recanalization in up to 13% of patients, with animal experiments suggesting a time-dependent increase [16]. Although the use of a hexa-row no-knife stapler has reduced this risk concerns regarding long-term anatomical durability remain [8,16]. A Cochrane systematic review also noted that uRY may reduce bile reflux compared to B-II, but its long-term efficacy is less consistent when compared with RY [1]. These findings show that although uRY may have potential advantages, its long-term effect may be limited in some cases.
Our findings reaffirm the anti-reflux efficacy of RY reconstruction [17]. Previous clinical studies reported that RY results in lower incidences of gastritis, bile reflux, and reflux esophagitis compared with B-II or BB, both at 1 and 2 years postoperatively [1]. More recent multicenter studies have similarly shown superior endoscopic outcomes for RY compared to BB during long-term follow-up [18].
Although Roux stasis syndrome remains a potential concern, our results showed no difference in DGE among groups, consistent with prior reports indicating that gastric emptying is determined more by neuromuscular integrity and anatomical alignment than by Roux division itself [17,19,20]. These findings align with network meta-analyses concluding that RY provides the most reliable control of bile reflux without increasing stasis when properly constructed [21].
The BB technique was originally designed to preserve intestinal continuity while reducing bile reflux through Braun anastomosis. Some studies reported short-term benefits, including less reflux symptoms and lower use of acid-suppressive or prokinetic medications [22]. However, according to a meta-analysis of randomized trials, these advantages are not sustained beyond the first year [21]. In our study, the BB group showed the highest RGB scores and the steepest deterioration over time, consistent with previous findings that BB reconstruction provides only limited benefit [1].
Nutritional outcomes were preserved across all three groups over the 24-month follow-up, with no significant intergroup differences. This is consistent with previous multicenter trials and meta-analyses showing that nutritional recovery is largely unaffected by reconstruction type [1,18,23]. Similarly, quality of life measures was generally comparable among groups, except in the reflux domain, where RY demonstrated superiority. Together, these results indicate that reflux control, rather than nutritional recovery, represents the key differentiator among reconstruction methods.
The operation time showed no difference between the groups. Since in our experience, the reconstruction methods share main procedures such as side to side GJ and jejunojejunostomy. Additional division of jejunum in the RY or application of additional knife less stapler in uRY did not consume considerable operation time. There were no significant differences in the number of retrieved lymph nodes or in postoperative complication, indicating that these reconstruction variations did not compromise surgical radicality or safety. The surgeon’s familiarity with reconstruction and the number of anastomoses may have impacted surgical outcomes, however, were not shown in our study. Our findings align with current evidence and clinical guidelines regarding reconstruction after distal gastrectomy. Studies have shown that perioperative outcomes, including operative time, number of retrieved lymph nodes, and postoperative complications, do not differ significantly among reconstruction types, supporting the safety and feasibility of each approach [21,24]. The Korean Gastric Cancer Association guidelines also recognize Billroth I, Billroth II (with or without Braun), and RY as acceptable methods, with the choice depending on patient condition and surgeon expertise [3]. Taken together, these findings highlight that no single reconstruction is universally superior; rather, the selection should be individualized, balancing reflux prevention, nutritional recovery, and technical considerations.
One of the implications of our study is that future studies should investigate the correlation between the RGB score and quality of life instruments. There were many attempts to connect bile reflux and gastric residue to worse quality of life. However, to date, validation of RGB scores with patient-reported outcomes has not yet been established.
Several limitations should be noted. First, the retrospective nature of this study introduces inherent selection bias, as reconstruction methods were determined by surgeon preference. Second, follow-up was limited to 24 months, which may underestimate the long-term risk of recanalization in uRY. Third, quality of life outcome was not included, although they are increasingly recognized as essential in post-gastrectomy evaluation.
In conclusion, BB, RY, and uRY reconstructions show no difference in nutritional recovery. RY shows the most favorable endoscopic outcomes, whereas uRY may be considered in selected patients, since its stability remains uncertain due to the risk of recanalization in some cases. BB reconstruction provides limited effect in RGB scores compared to RY and uRY. Future randomized prospective studies with extended follow-up and incorporation of patient-reported outcomes are needed to determine the optimal strategy in reconstruction following gastrectomy in gastric cancer surgery.
Footnotes
Conflict of Interest
No potential conflict of interest relevant to this article was reported.
Funding
None.
Author Contributions
Conceptualization: THK. Methodology: THK. Validation: THK, JHP, YJL. Formal analysis: THK. Investigation: JHP. Resources: SHJ, RRC. Data curation: HJK, RRC, JHP. Writing–original draft: THK. Writing–review & editing: SHJ, RRC, HJK, JHP, YJL. Visualization: THK. Supervision: YJL. Project administration: THK. All authors contributed to the manuscript and approved the final version for publication.
Institutional Review Board Statement
The study protocol was approved by the Institutional Review Board of Gyeongsang National University Hospital (No. GNUH 2020-12-010). The requirement for written informed consent was waived by the Institutional Review Board.
Data Availability Statement
Data analyzed in this study are available from the corresponding author upon reasonable request.
Supplementary Materials
Supplementary materials are available at the Korean Journal of Clinical Oncology website (http://www.kjco.org/).
REFERENCES
- 1.Cai Z, Zhou Y, Wang C, Yin Y, Yin Y, Shen C, et al. Optimal reconstruction methods after distal gastrectomy for gastric cancer: a systematic review and network meta-analysis. Medicine (Baltimore) 2018;97:e10823. doi: 10.1097/MD.0000000000010823. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Kim TH, Kim IH, Kang SJ, Choi M, Kim BH, Eom BW, et al. Korean Practice Guidelines for gastric cancer 2022: an evidence-based, multidisciplinary approach. J Gastric Cancer. 2023;23:3–106. doi: 10.5230/jgc.2023.23.e11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kim IH, Kang SJ, Choi W, Seo AN, Eom BW, Kang B, et al. Korean Practice Guidelines for gastric cancer 2024: an evidence-based, multidisciplinary approach (Update of 2022 Guideline) J Gastric Cancer. 2025;25:5–114. doi: 10.5230/jgc.2025.25.e11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Vogel SB, Hocking MP, Woodward ER. Clinical and radionuclide evaluation of Roux-Y diversion for postgastrectomy dumping. Am J Surg. 1988;155:57–62. doi: 10.1016/s0002-9610(88)80258-7. [DOI] [PubMed] [Google Scholar]
- 5.Mathias JR, Fernandez A, Sninsky CA, Clench MH, Davis RH. Nausea, vomiting, and abdominal pain after Roux-en-Y anastomosis: motility of the jejunal limb. Gastroenterology. 1985;88(1 Pt 1):101–7. doi: 10.1016/s0016-5085(85)80140-2. [DOI] [PubMed] [Google Scholar]
- 6.Noh SM. Improvement of the Roux limb function using a new type of “uncut Roux” limb. Am J Surg. 2000;180:37–40. doi: 10.1016/s0002-9610(00)00421-9. [DOI] [PubMed] [Google Scholar]
- 7.Van Stiegmann G, Goff JS. An alternative to Roux-en-Y for treatment of bile reflux gastritis. Surg Gynecol Obstet. 1988;166:69–70. [PubMed] [Google Scholar]
- 8.Park YS, Shin DJ, Son SY, Kim KH, Park DJ, Ahn SH, et al. Roux stasis syndrome and gastric food stasis after laparoscopic distal gastrectomy with uncut Roux-en-Y reconstruction in gastric cancer patients: a propensity score matching analysis. World J Surg. 2018;42:4022–32. doi: 10.1007/s00268-018-4715-6. [DOI] [PubMed] [Google Scholar]
- 9.Park JY, Kim YJ. Uncut Roux-en-Y reconstruction after laparoscopic distal gastrectomy can be a favorable method in terms of gastritis, bile reflux, and gastric residue. J Gastric Cancer. 2014;14:229–37. doi: 10.5230/jgc.2014.14.4.229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Japanese Gastric Cancer Association. Japanese gastric cancer treatment guidelines 2021 (6th edition) Gastric Cancer. 2023;26:1–25. doi: 10.1007/s10120-022-01331-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kubo M, Sasako M, Gotoda T, Ono H, Fujishiro M, Saito D, et al. Endoscopic evaluation of the remnant stomach after gastrectomy: proposal for a new classification. Gastric Cancer. 2002;5:83–9. doi: 10.1007/s101200200014. [DOI] [PubMed] [Google Scholar]
- 12.Taylor PR, Mason RC, Filipe MI, Vaja S, Hanley DC, Murphy GM, et al. Gastric carcinogenesis in the rat induced by duodenogastric reflux without carcinogens: morphology, mucin histochemistry, polyamine metabolism, and labelling index. Gut. 1991;32:1447–54. doi: 10.1136/gut.32.12.1447. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Dixon MF, Neville PM, Mapstone NP, Moayyedi P, Axon AT. Bile reflux gastritis and Barrett’s oesophagus: further evidence of a role for duodenogastro-oesophageal reflux? Gut. 2001;49:359–63. doi: 10.1136/gut.49.3.359. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Nakamura M, Nakamori M, Ojima T, Iwahashi M, Horiuchi T, Kobayashi Y, et al. Randomized clinical trial comparing long-term quality of life for Billroth I versus Roux-en-Y reconstruction after distal gastrectomy for gastric cancer. Br J Surg. 2016;103:337–47. doi: 10.1002/bjs.10060. [DOI] [PubMed] [Google Scholar]
- 15.Park SH, Hur H, Park JH, Lee CM, Son YG, Jung MR, et al. Reappraisal of optimal reconstruction after distal gastrectomy: a study based on the KLASS-07 database. Int J Surg. 2024;110:32–44. doi: 10.1097/JS9.0000000000000796. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wu F, Ni Z, Diao H, Huang C, Wang S, Ge B, et al. Recanalization in uncut Roux-en-Y reconstruction: an animal experiment and a clinical study. Front Surg. 2021;8:644864. doi: 10.3389/fsurg.2021.644864. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Van der Mijle HC, Kleibeuker JH, Limburg AJ, Beekhuis H, Lamers CB, van Schilfgaarde R. Role of vagal dysfunction in motility and transit disorders of jejunal Roux limb after Roux-en-Y gastrojejunostomy. Dig Dis Sci. 1994;39:827–33. doi: 10.1007/BF02087430. [DOI] [PubMed] [Google Scholar]
- 18.Chen YX, Huang QZ, Wang PC, Zhu YJ, Chen LQ, Wu CY, et al. Short- and long-term outcomes of Roux-en-Y and Billroth II with Braun reconstruction in total laparoscopic distal gastrectomy: a retrospective analysis. World J Surg Oncol. 2023;21:361. doi: 10.1186/s12957-023-03249-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.van der Mijle HC, Kleibeuker JH, Limburg AJ, Bleichrodt RP, Beekhuis H, van Schilfgaarde R. Manometric and scintigraphic studies of the relation between motility disturbances in the Roux limb and the Roux-en-Y syndrome. Am J Surg. 1993;166:11–7. doi: 10.1016/s0002-9610(05)80574-4. [DOI] [PubMed] [Google Scholar]
- 20.van der Mijle HC, Beekhuis H, Bleichrodt RP, Kleibeuker JH. Transit disorders of the gastric remnant and Roux limb after Roux-en-Y gastrojejunostomy: relation to symptomatology and vagotomy. Br J Surg. 1993;80:60–4. doi: 10.1002/bjs.1800800122. [DOI] [PubMed] [Google Scholar]
- 21.Lombardo F, Aiolfi A, Cavalli M, Mini E, Lastraioli C, Panizzo V, et al. Techniques for reconstruction after distal gastrectomy for cancer: updated network meta-analysis of randomized controlled trials. Langenbecks Arch Surg. 2022;407:75–86. doi: 10.1007/s00423-021-02411-6. [DOI] [PubMed] [Google Scholar]
- 22.Kim JC, Lee MJ, Lee HJ, Park K, Kang MK, Kim SH, et al. Is Braun jejunojejunostomy necessary? Comparison between Billroth-II alone and Billroth-II with Braun anastomosis after distal gastrectomy. J Gastric Cancer. 2025;25:318–29. doi: 10.5230/jgc.2025.25.e13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.So JB, Rao J, Wong AS, Chan YH, Pang NQ, Tay AY, et al. Roux-en-Y or Billroth II reconstruction after radical distal gastrectomy for gastric cancer: a multicenter randomized controlled trial. Ann Surg. 2018;267:236–42. doi: 10.1097/SLA.0000000000002229. [DOI] [PubMed] [Google Scholar]
- 24.Sun MM, Fan YY, Dang SC. Comparison between uncut Roux-en-Y and Roux-en-Y reconstruction after distal gastrectomy for gastric cancer: a meta-analysis. World J Gastroenterol. 2018;24:2628–39. doi: 10.3748/wjg.v24.i24.2628. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.