M. Nakamura 1, M. Nakamori 1,T.Ojima 1, M. Iwahashi 1, T. Horiuchi 6, Y. Kobayashi 2, N. Yamade 3, K. Shimada 4,M.Oka 5 and H.

Transcription

1 Randomized clinical trial Randomized clinical trial comparing long-term quality of life for Billroth I versus Roux-en-Y reconstruction after distal gastrectomy for gastric cancer M. Nakamura 1, M. Nakamori 1,T.Ojima 1, M. Iwahashi 1, T. Horiuchi 6, Y. Kobayashi 2, N. Yamade 3, K. Shimada 4,M.Oka 5 and H. Yamaue 1 1 Second Department of Surgery, Wakayama Medical University, School of Medicine, and Departments of Surgery, 2 Japan Labour Health and Welfare Organization Wakayama Rosai Hospital, 3 Shingu Municipal Medical Centre, 4 Hashimoto Municipal Hospital and 5 National Hospital Organization Minami Wakayama Medical Centre, Wakayama and 6 Department of Surgery, National Hospital Organization Osaka Minami Medical Centre, Osaka, Japan Correspondence to: Professor H. Yamaue, Second Department of Surgery, Wakayama Medical University, School of Medicine, Kimiidera, Wakayama , Japan ( yamaue-h@wakayama-med.ac.jp) Background: Patients quality of life (QoL) deteriorates remarkably after gastrectomy. Billroth I reconstruction following distal gastrectomy has the physiological advantage of allowing food to pass through the duodenum. It was hypothesized that Billroth I reconstruction would be superior to Roux-en-Y reconstruction in terms of long-term QoL after distal gastrectomy. This study compared two reconstructions in a multicentre prospective randomized clinical trial to identify the optimal reconstruction procedure. Methods: Between January 2009 and September 2010, patients who underwent gastrectomy for gastric cancer were randomized during surgery to Billroth I or Roux-en-Y reconstruction. The primary endpoint was assessment of QoL using the Functional Assessment of Cancer Therapy Gastric (FACT-Ga) questionnaire 36 months after surgery. Results: A total of 122 patients were enrolled in the study, 60 to Billroth I and 62 to Roux-en-Y reconstruction. There were no differences between the two groups in terms of postoperative complications or mortality, and no significant differences in FACT-Ga total score (P = 0 496). Symptom scales such as epigastric fullness (heaviness), diarrhoea and fatigue were significantly better in the Billroth I group at 36 months after gastrectomy (heaviness, P = 0 040; diarrhoea, P = 0 046; fatigue, P = 0 029). The rate of weight loss in the third year was lower for patients in the Billroth I group (P = 0 046). Conclusion: The choice of anastomotic reconstruction after distal gastrectomy resulted in no difference in long-term QoL in patients with gastric cancer. Registration number: NCT ( Paper accepted 21 October 2015 Published online 3 February 2016 in Wiley Online Library ( DOI: /bjs Introduction Gastric cancer is the fourth most common cancer and the second most common cause of cancer-related death in the world 1. Survival rates of patients with gastric cancer have improved as a result of early detection and advances in treatment, including perioperative chemotherapy, and health-related quality of life (QoL) is now the focus for patients after gastrectomy 2 6. Patients who have undergone gastrectomy may have various symptoms, including dumping syndrome and nutritional problems such as loss of bodyweight and malnutrition 5 8, that cause significant deterioration in their QoL 5 7. Worldwide, Billroth I reconstruction has commonly been performed after distal gastrectomy 9,10. However, gastroduodenal reflux occurs frequently after this procedure, and may cause remnant gastritis and occasionally remnant gastric ulcer and reflux oesophagitis Moreover, studies 14,15 have reported that reflux of bile and pancreatic secretions causes oesophageal and gastric cancer in the rat foregut. Consequently, the frequency of Roux-en-Y reconstruction after distal gastrectomy has increased gradually 16. Roux-en-Y reconstruction can prevent reflux of bile and pancreatic secretions into the gastric and oesophageal mucosa after gastrectomy, which might decrease postoperative reflux gastritis and oesophagitis 16,17. However, patients who have undergone this procedure often suffer from delayed gastric emptying (Roux stasis syndrome), 2016 BJS Society Ltd BJS 2016; 103:

2 338 M. Nakamura, M. Nakamori, T. Ojima, M. Iwahashi, T. Horiuchi, Y. Kobayashi et al. Gastroduodenal anastamosis Duodenal stump Gastro-jejunal anastamosis Colon Colon Middle colic artery Ligament of Treitz Ligament of Treitz Roux limb Biliopancreatic limb a Billroth I b Roux-en-Y Fig. 1 Reconstruction procedures: a Billroth I and b Roux-en-Y with functional obstruction of the Roux limb 11,18. Moreover the procedure is technically more difficult and complex than the Billroth I procedure. Indeed, because an additional anastomosis is required in the Roux-en-Y procedure, the number of postoperative complications might be expected to be increased. Several prospective randomized clinical trials (RCTs) 17,19 21 have compared different reconstructions after distal gastrectomy. In these studies 17,19,20, minimal differences in the frequency of reflux oesophagitis and remnant gastritis were reported. In addition, postoperative QoL was not evaluated fully, as these RCTs assessed short-term QoL during the 12 months after surgery 19,20, or only once for each patient at varying time points after surgery 21. Therefore, the optimal type of surgical reconstruction after distal gastrectomy remains unclear. The aim of the present study was to investigate the optimal procedure for reconstruction after distal gastrectomy in patients who had a Billroth I or Roux-en-Y procedure, with respect to long-term QoL, which was evaluated at regular intervals for 36 months after surgery. Methods The study was designed as a multicentre RCT comparing Billroth I and Roux-en-Y reconstructions after distal gastrectomy (Fig. 1). Operations included in the trial were performed between January 2009 and September 2010 in patients with gastric cancer treated at Wakayama Medical University Hospital (WMUH) and its associated teaching hospitals, which had been approved as a training institute by the Japanese Society of Gastroenterological Surgery. Each hospital was the base hospital in the region with a population of about a million people, and conducted more than 50 gastrectomies per year. This trial was approved by the Ethical Committee on Clinical Investigation at WMUH. The study was registered at ClinicalTrials.gov (NCT ). Informed consent was obtained from all participating patients before surgery; patients also agreed to attend for follow-up for at least 36 months after the operation. Study population Eligible patients were recruited into the study before surgery, on the basis of whether Billroth I or Roux-en-Y

3 Quality of life after distal gastrectomy 339 reconstruction could be performed following distal gastrectomy. Inclusion criteria were: histologically confirmed adenocarcinoma of the stomach; performance status 0 or 1 according to Eastern Cooperative Oncology Group criteria 22 ; patient aged between 20 and 80 years; tumour located in the antrum, angle or lower body of the stomach; no evidence of distant metastasis; and appropriate informed consent obtained. Exclusion criteria were: severe co-morbidity such as myocardial infarction, respiratory disorder requiring oxygen inhalation, liver cirrhosis or chronic renal failure requiring haemodialysis, which may prolong the hospital stay; history of other organ malignancies; proven mental illness; patients who were diagnosed as inappropriate for the study by a physician; and no informed consent. Randomization process Patients were randomized to either a Billroth I or Roux-en-Y reconstruction once the surgeon had completed the gastric resection and confirmed that either type of reconstruction could be performed. Surgeons contacted the non-profit organization Japan Clinical Research Support Unit (J-CRSU), located in Tokyo, by telephone to receive a randomly generated assignment (allocation ratio 1 : 1) into one of the treatment groups. A research physician conducted the randomization using a computer-generated random number pattern (block size 4) in a central registry at the J-CRSU. Patients were stratified according to the approach status (open or laparoscopy-assisted gastrectomy) and institution. Patients were not blinded for the type of reconstruction. Surgical procedures Distal gastrectomy was performed with D1/D1+ or D2 lymphadenectomy 23, and all resected primary lesions and dissected lymph nodes were examined histologically with haematoxylin and eosin staining. All gastrectomies were performed by senior surgeons, who were members of the Japanese Gastric Cancer Association, according to the institutional protocol guidelines. The stomach was dissected with a linear stapling device at least 3 cm proximal to the tumour site for early cancer, or 5 cm for advanced cancer. During the reconstruction phase, the Billroth I procedure was performed by means of end-to-side anastomosis using a circular stapling device, size 29 mm (CDH, Ethicon Endo-Surgery, Cincinnati, Ohio, USA; or PCEEA TM, Covidien, Mansfield, Massachusetts, USA). In Roux-en-Y reconstruction, the jejunum was divided cm distal to the ligament of Treitz. About 5 cm of jejunum was sacrificed to preserve the nerves in the mesentery. The jejunal loop was brought up the retrocolic route, and gastrojejunostomy was performed by end-to-end anastomosis using a circular stapling device, size 29 mm (CDH or PCEEA TM ). An anastomosis of the proximal end of the jejunum to the distal jejunum was created cm distal from the jejunal division using a circular stapling device, size 21 mm (CDH ). The common entry hole of the remnant stomach was closed using the linear stapling device. In the laparoscopic approach, the reconstruction procedure was performed through a small midline upper abdominal incision 4 5 cm in length. Postoperative management Patients were allowed to drink water on postoperative day 3. Solid diet was usually resumed on day 4. Patients were discharged home on days Patients with pathological stage II or III disease received adjuvant chemotherapy by S Postoperative complications were divided into early (0 30 days) and late (30 days onwards), and were analysed in each group using the Clavien Dindo classification system 25. Nutritional parameters after operation were assessed by changes in bodyweight and laboratory findings (serum total protein, serum albumin, serum prealbumin, retinol-binding protein, total cholesterol, triglyceride, cholinesterase and fasting blood sugar levels). In addition, endoscopic findings were analysed to examine the relationship with long-term QoL or symptoms after distal gastrectomy. Endoscopic examinations were performed at 12 and 36 months after surgery to evaluate the condition of the lower oesophageal and remnant gastric mucosa, and to assess the amount of residue and presence of bile in the remnant stomach. The degree of reflux oesophagitis was classified according to the Los Angeles Classification System 26, and the degree of gastritis and residual food and the presence of bile reflux were evaluated according to the RGB (Residue, Gastritis and Bile) classification 27. Endoscopic evaluation was performed by endoscopic specialists who were not involved with the original operation. Patients were followed up routinely at the WMUH outpatient clinic 1 and 3 months after surgery, and then every 3 months. Histopathological assessment The pathological classification of the primary tumour, the degree of lymph node involvement and the presence of organ metastasis were determined according to the TNM classification system of the American Joint Committee on Cancer 28.

4 340 M. Nakamura, M. Nakamori, T. Ojima, M. Iwahashi, T. Horiuchi, Y. Kobayashi et al. Assessed for eligibility n = 136 Enrolment Excluded n = 14 Did not meet inclusion criteria n = 12 Refused to participate n = 2 Randomized n = 122 Allocation Allocated to Billroth I procedure n = 60 Received intervention n = 60 Did not receive intervention n = 0 Allocated to Roux-en-Y procedure n = 62 Received intervention n = 59 Did not receive intervention n = 3 Changed reconstruction n = 3 Follow-up Completed questionnaire survery n = 60 Lost to follow-up n = 0 Discontinued intervention n = 0 Completed questionnaire survey n = 59 Lost to follow-up n = 0 Discontinued intervention n = 0 Analysis QoL questionnaires completed Baseline 60 of 60 (100%) 12 months 56 of 58 (97%) 36 months 53 of 56 (95%) QoL questionnaires completed Baseline 59 of 59 (100%) 12 months 56 of 58 (97%) 36 months 52 of 55 (95%) Fig. 2 CONSORT diagram for the study. QoL, quality of life Quality-of-life instruments and data collection QoL was recorded using version 4 of the Functional Assessment of Cancer Therapy Gastric (FACT-Ga) questionnaire 29 31, which is part of the Functional Assessment of Chronic Illness Therapy measurement system. The general core questionnaire (Functional Assessment of Cancer Therapy General; FACT-G) is comprised of four general subscales: physical, social, emotional and functional well-being. The FACT-Ga questionnaire combines the 27 items of FACT-G and the 19 newly validated items of the Gastric Cancer Subscale, which assesses gastric cancer-specific domains of postoperative gastrointestinal symptoms including dumping syndrome, gastric fullness, appetite loss, diarrhoea and bile reflux gastritis. As the FACT-Ga questionnaire has been validated and translated into Japanese 29,31, the Japanese version was used. QoL was assessed before operation and at 12 and 36 months afterwards. A trained coordinator collected the postoperative QoL data by direct contact with patients. Patients completed the questionnaire by themselves to reduce bias, with their family filling it in when the patient was unable to write. If there were missing items in the FACT-Ga questionnaire, subscale scores were prorated according to version 4 of the FACT-Ga scoring guidelines. Subscale scores with missing items were calculated by multiplying the sum of subscale by the number of items in the subscale, then dividing by the number of items actually answered. Prorating by subscale in this way is acceptable as long as more than 50 per cent of the items have been answered. Missing data in QoL questionnaires were excluded from analyses. All data on postoperative outcomes were collected by a trained coordinator and discussed by at least three surgeons, who were not blinded to the allocation. Study endpoints The primary endpoint was to demonstrate the superiority of Billroth I reconstruction with regard to the FACT-Ga total score as QoL assessment using the FACT-Ga questionnaire at 36 months after surgery. Secondary endpoints were other QoL assessments apart from the FACT-Ga total score (scores of physical, social, emotional and functional well-being, Gastric Cancer Subscale, FACT Gastric Trial Outcome Index (FACT-Ga TOI), FACT-G and symptom scales), perioperative and postoperative complications, loss of bodyweight, nutritional status and endoscopic evaluation. Sample size A pilot study was performed to compare postoperative QoL following Billroth I and Roux-en-Y reconstructions,

5 Quality of life after distal gastrectomy 341 Table 1 Characteristics of enrolled patients Billroth I (n = 60) Roux-en-Y (n = 62) Age (years)* 66 (40 80) 67 (43 80) Sex ratio (M : F) 40 : : 17 ECOG performance status 0 56 (93) 58 (94) 1 4 (7) 4 (6) Preoperative weight (kg) 58. 4(9. 3) 57. 7(10. 8) Concomitant disease 22 (37) 22 (35) Macroscopic appearance 0 35 (58) 36 (58) 1 2 (3) 1 (2) 2 10 (17) 14 (23) 3 11 (18) 11 (18) 5 2 (3) 0 (0) Location Middle 36 (60) 32 (52) Lower 24 (40) 30 (48) Surgical approach Open 35 (58) 37 (60) Laparoscopic 25 (42) 25 (40) Duration of surgery (min) (64. 3) (77. 9) Blood loss (ml) (181. 1) (239. 0) Lymph node dissection D1/D1+ 30 (50) 30 (48) D2 30 (50) 32 (52) No. of retrieved lymph nodes 29. 8(12. 6) 27. 3(10. 8) Postoperative hospital stay (days)* 11 (7 63) 11 (7 88) Histological type Well differentiated 35 (58) 38 (61) Undifferentiated 25 (42) 24 (39) pt category 1 39 (65) 41 (66) 2 11 (18) 11 (18) 3 4 (7) 5 (8) 4a 6 (10) 5 (8) pn category 0 42 (70) 40 (65) 1 6 (10) 7 (11) 2 6 (10) 8 (13) 3 6 (10) 7 (11) ptnm stage IA 34 (57) 33 (53) IB 10 (17) 12 (19) IIA 1(2) 3(5) IIB 7 (12) 5 (8) IIIA 2 (3) 1 (2) IIIB 4 (7) 5 (8) IIIC 2 (3) 3 (5) Adjuvant treatment 14 (23) 15 (24) Recurrence 5 (8) 6 (10) Values in parentheses are percentages unless indicated otherwise; values are *median (range) and mean(s.d.). According to the Japanese classification of gastric carcinoma 32. ECOG, Eastern Cooperative Oncology Group. Table 2 Early postoperative complications according to the Clavien Dindo classification 25 Billroth I (n = 60) Roux-en-Y (n = 59) P* Grade II Surgical-site infection 1 (2) 0 (0) Delayed gastric emptying 0 (0) 4 (7) Pancreatic fistula 0 (0) 1 (2) Grade IIIa Anastomotic leakage 1 (2) 0 (0) Anastomotic stricture 1 (2) 1 (2) Pancreatic fistula 0 (0) 1 (2) Grade IIIb Anastomotic leakage 1 (2) 0 (0) Anastomotic stricture 0 (0) 1 (2) Postoperative bleeding 0 (0) 1 (2) Total 4 (7) 9 (15) Values in parentheses are percentages. *Fisher s exact test. assessed using the FACT-Ga questionnaire. It was found that mean changes from the preoperative FACT-Ga total score were 2 points in the Billroth I group and 12 points in the Roux-en-Y group at about 3 years after surgery. Billroth I reconstruction has the physiological advantage of allowing food to pass through the duodenum. It was considered that, as the duodenal passage would improve dumping syndrome and nutritional status, it might contribute to recovery in FACT-Ga total scores for FACT-G and the Gastric Cancer Subscale in the longer term. It was hypothesized, therefore, that Billroth I reconstruction would be superior to Roux-en-Y reconstruction in terms of QoL at 3 years after gastrectomy. Thus, the study design to predict the number of patients necessary for statistical validity (2-sided) was based on 10 points of improvement in mean changes from the preoperative score in the FACT-Ga total score, from 12 points in the Roux-en-Y group to 2 points in the Billroth I group, with α setat0 05 and β set at 0 09, yielding a power of 91 per cent. Consequently, a total sample size of 100 patients was targeted. When the statistical power was set to 90 per cent at initial planning of the study, the sample size of a single group was 49 patients. A statistical power of 91 per cent was therefore chosen to set the sample size of each group to just 50 patients. Anticipating loss to follow-up, it was calculated that 60 patients would be required in each arm, for a total study population of 120. Statistical analysis Statistical analysis was performed with the SPSS version 21.0 software program (IBM, Armonk, New York, USA). Data for QoL were expressed as mean changes

6 342 M. Nakamura, M. Nakamori, T. Ojima, M. Iwahashi, T. Horiuchi, Y. Kobayashi et al. Physical well-being FACT-Ga Social well-being FACT-G Emotional well-being Billroth I, 12 months Roux-en-Y, 12 months Billroth I, 36 months Roux-en-Y, 36 months FACT-Ga TOI Functional well-being Gastric Cancer Subscale Fig. 3 Mean changes from preoperative scores in function scales, based on version 4 of the Functional Assessment of Cancer Therapy Gastric (FACT-Ga) questionnaire over a 36-month period, in physical, social, emotional and functional well-being, Gastric Cancer Subscale, FACT-Ga Trial Outcome Index (TOI), Functional Assessment of Cancer Therapy General (FACT-G) and FACT-Ga scores in patients undergoing distal gastrectomy followed by Billroth I or Roux-en-Y reconstruction. Scores above zero reflect an improved quality of life (QoL) and scores below zero reflect a worse QoL relative to the preoperative status. Scores for all parameters were similar between the two groups at both 12 and 36 months in FACT-Ga scores from preoperative scores (mean difference between preoperative and postoperative scores). A change greater than zero reflects an improved QoL relative to the preoperative status, whereas a change of less than zero reflects a worse QoL. Data for bodyweight were expressed as percentage weight loss compared with preoperative values. Laboratory findings were expressed as percentage change from preoperative values. Kaplan Meier survival curves were determined, and compared with the log rank test. All values except postoperative hospital stay were expressed as mean(s.d.) values, and comparisons between the two groups were done using χ 2 test or Fisher s exact test, as appropriate, and Student s t test. Length of postoperative hospital stay was expressed as the median value. All statistical tests were two-sided. P < was considered to be statistically significant. Results Clinicopathological characteristics A CONSORT diagram for the study is shown in Fig. 2. During the study interval, 136 patients with gastric cancer were registered. Fourteen patients were excluded from the study because they did not meet the inclusion criteria (12) or refused to participate (2). The remaining 122 patients, who underwent distal gastrectomy and lymph nodedissection (D1/D1+ or D2), were randomized to either Billroth I (60 patients) or Roux-en-Y (62) reconstruction. Three patients in the Roux-en-Y group had to be changed to a Billroth I reconstruction or total gastrectomy because of intra-abdominal adhesions or microscopic extent of the lesion, and were excluded. Thus, 119 patients were analysed (60 had a Billroth I and 59 a Roux-en-Y reconstruction). Eight patients, four in the Billroth I group and four in the Roux-en-Y group, died within 3 years of surgery. Some 97 per cent of patients in both groups returned completed questionnaires at 12 months after surgery, and 95 per cent at 36 months (Fig. 2). Characteristics of the 122 randomized patients are shown in Table 1. No differences were observed between the two groups. With regard to surgical outcomes, duration of surgery was significantly shorter for Billroth I than for Roux-en-Y (P = 0 015). The median length of follow-up for all patients was 44 (range 3 56) months. No patient was lost to follow-up. Missing follow-up data At 12 and 36 months follow-up, QoL questionnaire data were missing for 3 per cent (2 of 58) and 5 per cent (3 of 56) of patients respectively in the Billroth I group, and for 3 per cent (2 of 58) and 5 per cent (3 of 55) of those in the Roux-en-Y group. QoL questionnaires were missing items for 5 per cent (3 of 56) and 8 per cent (4 of 53) of patients in the Billroth I group, and for 4 per cent (2 of 56) and 8

7 Quality of life after distal gastrectomy 343 Epigastric fullness (heaviness) 0 Flatus Diarrhoea Epigastric discomfort (pain) Fatigue Billroth I, 12 months Roux-en-Y, 12 months Billroth I, 36 months Roux-en-Y, 36 months Reflux symptom (heartburn) Nausea Appetite loss Fig. 4 Mean changes from preoperative scores in symptom scales, based on version 4 of the Functional Assessment of Cancer Therapy Gastric (FACT-Ga) questionnaire over a 36-month period, in epigastric fullness (heaviness), diarrhoea, fatigue, nausea, appetite loss, reflux symptom (heartburn), epigastric discomfort (pain) and flatus scores in patients undergoing distal gastrectomy followed by Billroth I or Roux-en-Y reconstruction. Scores above zero reflect an improved quality of life (QoL) and scores below zero reflect a worse QoL relative to the preoperative status. Scores for parameters at 12 months did not differ between the two groups, but scores for epigastric fullness (P = 0 040), diarrhoea (P = 0 046) and fatigue (P = 0 029) were significantly better in the Billroth I group at 36 months after surgery (Student s t test) Total protein 120 Fasting blood sugar Albumin 60 Cholinesterase Prealbumin Billroth I, 12 months Roux-en-Y, 12 months Billroth I, 36 months Roux-en-Y, 36 months Triglyceride Retinol-binding protein Total cholesterol Fig. 5 Comparison of the rates of change of nutritional parameters (%) serum total protein, serum albumin, serum prealbumin, retinol-binding protein, total cholesterol, triglyceride, cholinesterase and fasting blood sugar levels in Billroth I and Roux-en-Y groups at 12 and 36 months after operation. No difference was found in any parameter per cent (4 of 52) of those in the Roux-en-Y group at 12 and 36 months respectively. Data on endoscopic findings were missing for 0 per cent (0 of 58) and 5 per cent (3 of 56) of patients in the Billroth I group, and for 2 per cent (1 of 58) and 5 per cent (3 of 55) in the Roux-en-Y group at 12 and 36 months respectively. There were no missing data for postoperative complications, survival or nutritional parameters. Postoperative complications and survival There were no significant differences between the groups regarding total early complications (0 30 days) (Table 2). In the Roux-en-Y group, four patients (7 per cent) experienced delayed gastric emptying, one patient (2 per cent) had postoperative bleeding that required surgery, and two (3 per cent) developed a pancreatic fistula; one fistula was

8 344 M. Nakamura, M. Nakamori, T. Ojima, M. Iwahashi, T. Horiuchi, Y. Kobayashi et al. treated conservatively and the other needed percutaneous intra-abdominal drainage. In the Billroth I group, two patients (3 per cent) developed anastomotic leakage; one patient required percutaneous intra-abdominal drainage and the other needed surgery. There were no late complications (30 days onwards) in either group, and no unexpected effects. There were no differences in overall survival between the two groups: the 3-year survival rate was 93 per cent in both groups (P = 0 990). Assessment of quality of life The scores for physical, social, emotional and functional well-being, FACT-Ga TOI and FACT-G were similar in the two groups at both 12 and 36 months after surgery (Fig. 3). Regarding the Gastric Cancer Subscale and the FACT-Ga total score, scores at 36 months in the Billroth I group were slightly higher than those in the Roux-en-Y group, although the differences were not statistically significant (Gastric Cancer Subscale: Billroth I 2 0 (95 per cent c.i. 4 7 to 0 7)versus Roux-en-Y 4 8( 7 9 to 1 6), P = 0 337; FACT-Ga: Billroth I 2 8 ( 8 3 to 2 7) versus Roux-en-Y 5 7 ( 11 9 to0 5), P = 0 496) (Fig. 3). For the symptom scale, scores at 12 months were not significantly different between the two groups, but at 36 months epigastric fullness (heaviness) ( 0 1 (95 per cent c.i. 0 3to0 1) for Billroth I versus 0 5( 0 8 to 0 2) for Roux-en-Y; P = 0 040),diarrhoea ( 0 4( 0 6to 0 1)versus 0 8( 1 1to 0 4) respectively; P = 0 046) and fatigue ( 0 1 ( 0 4 to0 2) versus 0 9 ( 1 2 to0 5) respectively; P = 0 029) were significantly better following Billroth I than Roux-en-Y reconstruction (Fig. 4). Nutritional status The mean(s.d.) rate of bodyweight loss was significantly lower in the Billroth I group at 36 months after surgery (6 8(7 9) per cent versus 10 0(8 0) per cent in the Roux-en-Y group; P = 0 046). Rates of change in serum total protein, serum albumin, serum prealbumin, retinol-binding protein, total cholesterol, triglyceride, cholinesterase and fasting blood sugar levels were similar between the two groups at 12 and 36 months (Fig. 5). Endoscopic findings Endoscopic findings for the gastric remnant and lower oesophagus were evaluated 12 and 36 months; the results at 36 months are shown in Table 3. No significant differences were found in reflux oesophagitis, amount of residual food or anastomotic stricture between the two groups at either Table 3 Postoperative endoscopic examination results 36 months after surgery Billroth I (n = 53) Roux-en-Y (n = 52) P Reflux oesophagitis No 45 (85) 42 (81) Yes 8 (15) 10 (19) Grade of oesophagitis* Normal 45 (85) 42 (81) Minimal change 4 (8) 4 (8) A 3 (6) 6 (12) B 1 (2) 0 (0) C 0 (0) 0 (0) D 0 (0) 0 (0) Remnant gastritis < No 18 (34) 37 (71) Yes 35 (66) 15 (29) Degree of gastritis Grade 0 18 (34) 37 (71) Grade 1 27 (51) 15 (29) Grade 2 6 (11) 0 (0) Grade 3 1 (2) 0 (0) Grade 4 1 (2) 0 (0) Extent of gastritis Grade 0 18 (34) 37 (71) Grade 1 18 (34) 9 (17) Grade 2 8 (15) 4 (8) Grade 3 9 (17) 2 (4) Residual food in stomach No 42 (79) 38 (73) Yes 11 (21) 14 (27) Grade 0 42 (79) 38 (73) Grade 1 2 (4) 1 (2) Grade 2 5 (9) 9 (17) Grade 3 3 (6) 3 (6) Grade 4 1 (2) 1 (2) Bile reflux < No 39 (74) 52 (100) Yes 14 (26) 0 (0) Anastomotic stricture No 53 (100) 52 (100) Yes 0 (0) 0 (0) Values in parentheses are percentages. *Los Angeles Classification System 26 ; RGB (Residue, Gastritis and Bile) classification 27. χ 2 test. 12 or 36 months. The incidence of remnant gastritis was significantly lower following Roux-en-Y reconstruction at 12months (37 per cent versus 64 per cent following Billroth I; P = 0 004) and 36 months (29 versus 66 per cent respectively; P < 0 001) (Table 3). The incidence of bile reflux into the remnant stomach was also significantly lower in the Roux-en-Y group at 12 months (2 per cent versus 26 per cent for Billroth I; P < 0 001) and 36 months (0 versus 26 per cent; P < 0 001) (Table 3). Discussion In this study, which compared two reconstructive procedures for their effect on long-term QoL after distal

9 Quality of life after distal gastrectomy 345 gastrectomy for gastric cancer, the choice of reconstruction did not influence long-term QoL. Several studies have compared different reconstructions after distal gastrectomy with respect to postoperative QoL. A meta-analysis 33 of Billroth I versus Billroth II versus Roux-en-Y following distal gastrectomy showed that the incidence of reflux symptoms was significantly reduced with the Roux-en-Y procedure compared with Billroth I or II. However, as almost all studies in that meta-analysis were non-rcts or small-scale trials, the level of evidence was low. Moreover, one RCT 21 evaluated QoL following Billroth I or Roux-en-Y reconstruction at varying time points after surgery, and showed that scores on the dyspnoea symptom scale were superior for Roux-en-Y. Another RCT 19 found no significant difference in postoperative QoL between Billroth I, Billroth II with Braun, and Roux-en-Y during the 12-month interval after distal gastrectomy. One study 16 evaluated the efficacy of Roux-en-Y and Billroth I reconstruction after laparoscopy-assisted distal gastrectomy, and reported that the incidence of heartburn was significantly higher in the Billroth I group at 12 months after surgery. Although these studies have examined short-term QoL, the present RCT assessed long-term QoL for 36 months after the procedure. In this study there were no significant differences between the two groups in terms of the primary endpoint, the FACT-Ga total score. Moreover, scores for the Gastric Cancer Subscale, which assesses gastric cancer-specific domains, were similar in the two groups. Therefore, the authors consider that both Billroth I and Roux-en-Y reconstructions are standard procedures after distal gastrectomy with regard to long-term QoL. When individual symptom scores were analysed, the Billroth I reconstruction was found to have advantages over the Roux-en-Y procedure with regard to some long-term symptoms. Epigastric fullness (heaviness), diarrhoea and fatigue were significantly better in the Billroth I group at 36 months. These are symptoms of dumping syndrome, and the present findings suggest that dumping syndrome might be a long-term problem following the Roux-en-Y procedure. Moreover, some investigators 5,34 have reported that fatigue is one of the most severe problems experienced after gastrectomy. Fatigue might change the functional roles of patients or their families, as usual activities cannot be performed 35, and specific interventions (such as reduction of work capacity and nutritional management) are necessary to manage the fatigue in the longer term. Loss of bodyweight is a defining characteristic of postgastrectomy syndrome, and maintaining bodyweight after gastrectomy is difficult. Terashima and colleagues 8 reported that Billroth I procedures resulted in significantly less postoperative weight loss than Roux-en-Y procedures, but their study was not an RCT. In the present RCT, the Billroth I procedure were found to confer a significant advantage in terms of the change in bodyweight at 3 years. Billroth I reconstruction has the physiological advantage of allowing food to pass through the duodenum. Indeed, it has been reported 36 that patients with preservation of the duodenal passage have improved nutritional parameters (bodyweight, serum iron and haemoglobin levels) following total gastrectomy. Preservation of the duodenal passage has been reported to result in improved physiological enrichment of the chyme with bile and pancreatic juice, and better physiological regulation of gastrointestinal hormones such as cholecystokinin, known to cause early satiety, and somatostatin, which plays a role in dumping syndrome Thus, the present results suggest the possibility that a physiological reconstruction method, such as the Billroth I procedure, has an effect on the digestion and absorption of food, and the regulation of gastrointestinal hormones. Delayed gastric emptying occurred in 7 per cent of patients who had a Roux-en-Y reconstruction. Roux stasis syndrome, such as delayed gastric emptying, is characterized by symptoms of upper gut stasis after Roux-en-Y gastrojejunostomy 18,19. Some researchers 42,43 have proposed that Roux stasis syndrome is caused by separation of the Roux limb from the natural small-bowel pacemaker, which is located in the proximal duodenum. Roux stasis syndrome occurs in about 20 per cent of patients who have a Roux-en-Y procedure 20,44, but its frequency in the present study was lower than in previous reports. In the present authors approach, the vagal nerves were preserved in the mesentery of the jejunum as the mesentery was not divided, and about 5 cm of jejunum was sacrificed at separation of the Roux limb. This might decrease the frequency of Roux stasis syndrome. In endoscopic examinations, the degree of remnant gastritis and bile reflux in the remnant stomach was significantly more severe following the Billroth I than the Roux-en-Y reconstruction. These results showed that the Roux-en-Y procedure was advantageous with regard to decreasing the incidence of bile reflux and preventing reflux gastritis. However, there was no difference in reflux oesophagitis between the two groups. These results were similar to those of other reports 16,20. The similar incidence of reflux symptoms (heartburn) between the two groups in the present study might reflect the endoscopic finding of reflux oesophagitis. Remnant gastritis and bile reflux on endoscopy might have been expected to cause a decrease in QoL after gastrectomy. However, no difference was found in QoL between the two types of reconstruction in terms

10 346 M. Nakamura, M. Nakamori, T. Ojima, M. Iwahashi, T. Horiuchi, Y. Kobayashi et al. of reflux symptoms (heartburn) and epigastric discomfort (pain). The reason for this discrepancy may be that symptoms related to duodenogastric or gastro-oesophageal reflux are caused by acid reflux rather than bile reflux. Indeed, gastrectomy reduces acid production. Therefore, it was considered that the incidence of reflux symptoms such as heartburn and epigastric discomfort (pain) might not correlate well with the endoscopic findings of remnant gastritis and bile reflux after gastrectomy. Thus, remnant gastritis and bile reflux in Billroth I reconstruction might not impair QoL after gastrectomy. There are certain limitations to this RCT. It is a small-scale study and might be underpowered. If a larger study had been conducted, a significant difference may have been demonstrated between the two groups in terms of long-term postoperative QoL. However, the clinical relevance of this may be questioned. Moreover, Billroth I reconstruction had the benefit of improvement in some postoperative symptoms such as epigastric fullness (heaviness), diarrhoea and fatigue, and bodyweight loss. Disclosure The authors declare no conflict of interest. References 1 Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA Cancer J Clin 2011; 61: Kaptein AA, Morita S, Sakamoto J. Quality of life in gastric cancer. World J Gastroenterol 2005; 11: Fein M, Fuchs KH, Thalheimer A, Freys SM, Heimbucher J, Thiede A. Long-term benefits of Roux-en-Y pouch reconstruction after total gastrectomy: a randomized trial. Ann Surg 2008; 247: Kim YW, Baik YH, Yun YH, Nam BH, Kim DH, Choi IJ et al. Improved quality of life outcomes after laparoscopy-assisted distal gastrectomy for early gastric cancer: results of a prospective randomized clinical trial. Ann Surg 2008; 248: Kim AR, Cho J, Hsu YJ, Choi MG, Noh JH, Sohn TS et al. Changes of quality of life in gastric cancer patients after curative resection: a longitudinal cohort study in Korea. Ann Surg 2012; 256: Karanicolas PJ, Graham D, Gönen M, Strong VE, Brennan MF, Coit DG. Quality of life after gastrectomy for adenocarcinoma: a prospective cohort study. Ann Surg 2013; 257: Korenaga D, Orita H, Okuyama T, Moriguchi S, Maehara Y, Sugimachi K. Quality of life after gastrectomy in patients with carcinoma of the stomach. Br J Surg 1992; 79: Terashima M, Tanabe K, Yoshida M, Kawahira H, Inada T, Okabe H et al. Postgastrectomy Syndrome Assessment Scale (PGSAS)-45 and changes in body weight are useful tools for evaluation of reconstruction methods following distal gastrectomy. Ann Surg Oncol 2014; 21: S370 S Yoshino K. History of gastric cancer surgery. J Jpn Surg Soc 2000; 101: Weil PH, Buchberger R. From Billroth to PCV: a century of gastric surgery. World J Surg 1999; 23: Fukuhara K, Osugi H, Takada N, Takemura M, Higashino M, Kinoshita H. Reconstructive procedure after distal gastrectomy for gastric cancer that best prevents duodenogastroesophageal reflux. World J Surg 2002; 26: Osugi H, Fukuhara K, Takada N, Takemura M, Kinoshita H. Reconstructive procedure after distal gastrectomy to prevent remnant gastritis. Hepatogastroenterology 2004; 51: Kauer WK, Peters JH, DeMeester TR, Feussner H, Ireland AP, Stein HJ et al. Composition and concentration of bile acid reflux into the esophagus of patients with gastroesophageal reflux disease. Surgery 1997; 122: Miwa K, Hasegawa H, Fujimura T, Matsumoto H, Miyata R, Kosaka T et al. Duodenal reflux through the pylorus induces gastric adenocarcinoma in the rat. Carcinogenesis 1992; 13: Nishijima K, Miwa K, Miyashita T, Kinami S, Ninomiya I, Fushida S et al. Impact of the biliary diversion procedure on carcinogenesis in Barrett s esophagus surgically induced by duodenoesophageal reflux in rats. Ann Surg 2004; 240: Kojima K, Yamada H, Inokuchi M, Kawano T, Sugihara K. A comparison of Roux-en-Y and Billroth-I reconstruction after laparoscopy-assisted distal gastrectomy. Ann Surg 2008; 247: Hirao M, Takiguchi S, Imamura H, Yamamoto K, Kurokawa Y, Fujita J et al.; Osaka University Clinical Research Group for Gastroenterological Study. Comparison of Billroth I and Roux-en-Y reconstruction after distal gastrectomy for gastric cancer: one-year postoperative effects assessed by a multi-institutional RCT. Ann Surg Oncol 2013; 20: 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: Lee MS, Ahn SH, Lee JH, Park do J, Lee HJ, Kim HH et al. What is the best reconstruction method after distal gastrectomy for gastric cancer? Surg Endosc 2012; 26: Ishikawa M, Kitayama J, Kaizaki S, Nakayama H, Ishigami H, Fujii S et al. Prospective randomized trial comparing Billroth-I and Roux-en-Y procedures after distal gastrectomy for gastric carcinoma. World J Surg 2005; 29: Takiguchi S, Yamamoto K, Hirao M, Imamura H, Fujita J, Yano M et al.; Osaka University Clinical Research Group for

11 Quality of life after distal gastrectomy 347 Gastroenterological Study. A comparison of postoperative quality of life and dysfunction after Billroth I and Roux-en-Y reconstruction following distal gastrectomy for gastric cancer: results from a multi-institutional RCT. Gastric Cancer 2012; 15: Oken MM, Creech RH, Tormey DC, Horton J, Davis TE, McFadden ET et al. Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol 1982; 5: Japanese Gastric Cancer Association. Japanese gastric cancer treatment guidelines 2010 (ver. 3). Gastric Cancer 2011; 14: Sakuramoto S, Sasako M, Yamaguchi T, Kinoshita T, Fujii M, Nashimoto A et al.; ACTS-GC Group. Adjuvant chemotherapy for gastric cancer with S-1, an oral fluoropyrimidine. N Engl J Med 2007; 357: Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004; 240: Armstrong D, Bennett JR, Blum AL, Dent J, De Dombal FT, Galmiche JP et al. The endoscopic assessment of esophagitis: a progress report on observer agreement. Gastroenterology 1996; 111: 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: Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A. AJCC Cancer Staging Manual (7th edn). Springer: New York, Eremenco SL, Cashy J, Webster K, Ohashi Y, Locker GY, Pelletier G et al. FACT-Gastric: a new international measure of QOL in gastric cancer. J Clin Oncol 2004; 22(Suppl): GarlandSN,PelletierG,LaweA,BiagioniBJ,EasawJ, Eliasziw M et al. Prospective evaluation of the reliability, validity, and minimally important difference of the functional assessment of cancer therapy-gastric (FACT-Ga) quality-of-life instrument. Cancer 2011; 117: Kawai M, Tani M, Hirono S, Miyazawa M, Shimizu A, Uchiyama K et al. Pylorus ring resection reduces delayed gastric emptying in patients undergoing pancreatoduodenectomy: a prospective, randomized, controlled trial of pylorus-resecting versus pylorus-preserving pancreatoduodenectomy. Ann Surg 2011; 253: Japanese Gastric Cancer Association. Japanese classification of gastric carcinoma: 3rd English edition. Gastric Cancer 2011; 14: Zong L, Chen P. Billroth I vs. Billroth II vs. Roux-en-Y following distal gastrectomy: a meta-analysis based on 15 studies. Hepatogastroenterology 2011; 58: Zieren HU, Zippel K, Zieren J, Müller JM. Quality of life after surgical treatment of gastric carcinoma. Eur J Surg 1998; 164: Lee MK, Lee KM, Bae JM, Kim S, Kim YW, Ryu KW et al. Employment status and work-related difficulties in stomach cancer survivors compared with the general population. Br J Cancer 2008; 98: Yang YS, Chen LQ, Yan XX, Liu YL. Preservation versus non-preservation of the duodenal passage following total gastrectomy: a systematic review. J Gastrointest Surg 2013; 17: Kalmár K, Németh J, Kelemen D, Agoston E, Horváth OP. Postprandial gastrointestinal hormone production is different, depending on the type of reconstruction following total gastrectomy. Ann Surg 2006; 243: Schwarz A, Büchler M, Usinger K, Rieger H, Glasbrenner B, Friess H et al. Importance of the duodenal passage and pouch volume after total gastrectomy and reconstruction with the Ulm pouch: prospective randomized clinical study. World J Surg 1996; 20: Kim BJ, O Connell T. Gastroduodenostomy after gastric resection for cancer. Am Surg 1999; 65: Le Blanc-Louvry I, Ducrotte P, Chayvialle JA, Coquerel A, Michot F, Ténière P. Distal gastrectomy and Roux-Y limb in the rat: plasma motilin, pancreatic polypeptide concentrations, and duodenojejunal motility. Dig Dis Sci 2000; 45: Meyer BM, Werth BA, Beglinger C, Hildebrand P, Jansen JB, Zach D et al. Role of cholecystokinin in regulation of gastrointestinal motor functions. Lancet 1989; 2: Morrison P, Miedema BW, Kohler L, Kelly KA. Electrical dysrhythmias in the Roux jejunal limb: cause and treatment. Am J Surg 1990; 160: Miedema BW, Kelly KA. The Roux stasis syndrome: treatment by pacing and prevention by use of an uncut Roux limb. Arch Surg 1992; 127: Hirao M, Fujitani K, Tsujinaka T. Delayed gastric emptying after distal gastrectomy for gastric cancer. Hepatogastroenterology 2005; 52: