Survival outcomes following laparoscopic versus open D3 dissection for stage II or III colon cancer (JCOG0404): a phase 3, randomised controlled trial

Articles

Survival outcomes following laparoscopic versus open D3 dissection for stage II or III colon cancer (JCOG0404): a phase 3, randomised controlled trial Seigo Kitano, Masafumi Inomata, Junki Mizusawa, Hiroshi Katayama, Masahiko Watanabe, Seiichiro Yamamoto, Masaaki Ito, Shuji Saito, Shoichi Fujii, Fumio Konishi, Yoshihisa Saida, Hirotoshi Hasegawa, Tomonori Akagi, Kenichi Sugihara, Takashi Yamaguchi, Tadahiko Masaki, Yosuke Fukunaga, Kohei Murata, Masazumi Okajima, Yoshihiro Moriya, Yasuhiro Shimada

Summary

Background Although benefits of laparoscopic surgery compared with open surgery have been suggested, the long-term survival of patients undergoing laparoscopic surgery for colon cancer requiring Japanese D3 dissection remains unclear. We did a randomised controlled trial to establish non-inferiority of laparoscopic surgery to open surgery. Methods We did an open-label, multi-institutional, randomised, two-arm phase 3 trial in 30 hospitals in Japan. Patients aged 20–75 years who had histologically proven colon cancer; tumours located in the caecum or ascending, sigmoid, or rectosigmoid colon; T3 or deeper lesions without involvement of other organs, node stages N0–2, and metastasis stage M0; and tumour size of 8 cm or smaller were included. Only accredited surgeons did surgery as an operator or instructor. Patients were randomly assigned (1:1) preoperatively to undergo D3 resection either by an open route or a laparoscopic route, via phone call or fax to the Japan Clinical Oncology Group (JCOG) Data Center. Randomisation used a minimisation method with a biased-coin assignment according to tumour location (caecum, ascending vs sigmoid, rectosigmoid) and institution. The primary endpoint was overall survival and was analysed by intention to treat. The non-inferiority margin for the hazard ratio (HR) was set at 1·366. This study is registered with UMIN Clinical Trials Registry, number C000000105, and ClinicalTrials.gov, number NCT00147134. Findings Between Oct 1, 2004, and March 27, 2009, 1057 patients were randomly assigned to either open surgery (n=528) or laparoscopic surgery (n=529). 5-year overall survival was 90·4% (95% CI 87·5–92·6) for open surgery and 91·8% (89·1–93·8) for laparoscopic surgery. Laparoscopic D3 surgery was not non-inferior to open surgery for overall survival (HR 1·06, 90% CI 0·79–1·41; pnon-inferiority=0·073). 65 (13%) patients in the open surgery group and 53 (10%) patients in the laparoscopic surgery group had grade 2–4 adverse events. Grade 2–4 adverse events included diarrhoea (15 [3%] in the open surgery group vs 14 [3%] in the laparoscopic surgery group), paralytic ileus (six [1%] vs nine [2%]), and small intestine bowel obstruction (16 [3%] vs 11 [2%]). Two treatment-related deaths occurred in the open surgery group: one patient died 7 days after surgery (probably due to myocardial infarction), and one patient died from febrile neutropenia, pneumonia, diarrhoea, and gastrointestinal haemorrhage during postoperative chemotherapy. Interpretation Laparoscopic D3 surgery was not non-inferior to open D3 surgery in terms of overall survival for patients with stage II or III colon cancer. However, because overall survival in both groups was similar and better than expected, laparoscopic D3 surgery could be an acceptable treatment option for patients with stage II or III colon cancer. Funding National Cancer Center Research and Development Fund, Grant-in-Aid for Cancer Research, and Health and Labour Sciences Research Grant for Clinical Cancer Research from the Ministry of Health, Labour and Welfare of Japan.

Introduction In the USA and Europe, laparoscopic surgery for colon cancer is accepted as a standard treatment, which has some benefits over open surgery including decreased pain, improved postoperative pulmonary function, reduced postoperative ileus, improved incidence of wound infection, faster recovery, and shorter hospital stay. In addition to these short-term outcomes, the long-term outcome after laparoscopic surgery for colon cancer is comparable to that after open surgery, as shown by the results of several randomised controlled trials.1–10 However, these trials were done before the standard procedure in the USA and Europe—ie, complete mesocolic excision with central vascular ligation—had www.thelancet.com/gastrohep Vol 2 April 2017

emerged as a technique. These trials also had several weaknesses: the proportion of patients with tumours of pathological stage 0–I was high (21–37%); rectal cancer was included in some trials; and the extent of lymph node dissection was not specified. The D3 dissection technique used in Japan emphasises anatomical lymph node dissection, defined by the dissection of lymph nodes at the root of the tumour feeding artery and the longitudinal length of large intestine to be resected; by contrast, the complete mesocolic excision technique emphasises preservation of the anatomical planes of surgical resection and central vascular ligation. Although these two concepts are different, the purpose and the extent of lymph node

Lancet Gastroenterol Hepatol 2017; 2: 261–68 Published Online February 1, 2017 http://dx.doi.org/10.1016/ S2468-1253(16)30207-2 See Comment page 237 Department of Gastroenterological and Pediatric Surgery, Oita University Faculty of Medicine, Oita, Japan (Prof S Kitano MD, Prof M Inomata MD, T Akagi MD); JCOG Data Center, Operations Office, National Cancer Center, Tokyo, Japan (J Mizusawa ME, H Katayama MD); Department of Surgery, Kitasato University Hospital, Kanagawa, Japan (Prof M Watanabe MD); Division of Colorectal Surgery (S Yamamoto MD, Y Moriya MD), and Gastrointestinal Oncology Division (Prof Y Shimada MD), National Cancer Center Hospital, Tokyo, Japan; Division of Colorectal Surgery, National Cancer Center Hospital East, Chiba, Japan (M Ito MD); Division of Colon and Rectal Surgery, Shizuoka Cancer Center Hospital, Shizuoka, Japan (S Saito MD); Department of Surgery, Gastroenterological Centre, Yokohama City University, Yokohama, Japan (Prof S Fujii MD); Department of Surgery, Jichi Medical University Saitama Medical Center, Saitama, Japan (Prof F Konishi MD); Departments of Surgery, Toho University Ohashi Medical Center, Tokyo, Japan (Prof Y Saida MD); Department of Surgery, Keio University School of Medicine, Tokyo, Japan (H Hasegawa MD); Department of Surgical Oncology, Tokyo Medical and Dental University, Tokyo, Japan (Prof K Sugihara MD); Department of Surgery, National Hospital Organization, Kyoto Medical

261

Articles

Centre, Kyoto, Japan (T Yamaguchi MD); Department of Surgery, Kyorin University School of Medicine, Tokyo, Japan (Prof T Masaki MD); Department of Gastroenterological Surgery, Osaka City General Hospital, Osaka, Japan (Y Fukunaga MD); Department of Surgery, Suita Municipal Hospital, Osaka, Japan (K Murata MD); and Department of Gastroenterological and Transplant Surgery, Hiroshima University, Hiroshima, Japan (M Okajima MD) Correspondence to: Prof Masafumi Inomata, Department of Gastroenterological and Pediatric Surgery, Oita University Faculty of Medicine, Hasama-machi, Yufu, Oita 879–5593, Japan [email protected]

Research in context Evidence before this study We searched PubMed for randomised clinical trials published in English up to Nov 30, 2016, using the terms “colon cancer”, “stage II or III”, “laparoscopic surgery”, and “long-term outcome”. A quarter of a century has passed since laparoscopic techniques were first used in colorectal surgery, and the number of such surgeries continues to increase worldwide. Most randomised controlled trials in the USA and Europe have revealed the short-term benefits of laparoscopic resection, with no significant difference in long-term outcomes. However, these trials were done before the concept of complete mesocolic excision with central vascular ligation had emerged and had several weaknesses: the proportion of patients with pathological stage 0–I cancer was high (21–37%), rectal cancer was included in some trials, and the extent of lymph node dissection was not specified. Added value of this study To overcome the above weaknesses in the previous trials, the study population in this study was limited to clinical stage II or III

dissection are similar. Theoretically, the techniques should be equivalent because the principles of both are the same.11 In the Japan Clinical Oncology Group (JCOG) 0404 trial, we aimed to determine whether D3 dissection via a laparoscopic route was non-inferior to the same dissection via an open route, in terms of overall survival.12 To overcome the weaknesses in previous trials, the study population in this study was limited to patients with clinical stage II or III colon cancer. Moreover, a detailed surgical procedure was specified in the protocol. D3 dissection was mandatory in all patients in our study, and additional extended lymph node dissection was allowed if necessary. The previously reported short-term outcomes of this study showed that laparoscopic surgery for colon cancer was more beneficial than open surgery.13 Here, we report to the long-term survival data from the trial.

Methods

Study design and participants This open-label, multi-institutional, randomised, two-arm phase 3 trial was done within the framework of the JCOG Colorectal Cancer Study Group. Patients were recruited from 30 hospitals in Japan. Eligible patients were those with the main tumour lesion located in the caecum, ascending colon, sigmoid colon, and rectosigmoid colon; with histologically confirmed adeno­ carcinoma, mucinous adenocarcinoma, signet ring cell carcinoma, or adenosquamous carcinoma; clinically diagnosed as T3–4, N0–2, or M0 lesions without involvement of other organs based on colonoscopy and enhanced abdominal CT; no multiple lesions are observed on general diagnosis based on endoscopy and preoperative image assessment (barium enema or upper abdominal or pelvic CT; carcinoma in situ or lesions equivalent to mucosal 262

colon cancer. Moreover, a detailed surgical procedure was specified in the protocol: D3 dissection, which is similar to complete mesocolic excision with central vascular ligation, was mandatory in all patients. This study is, to our knowledge, the first to compare and report long-term outcomes of laparoscopic surgery versus open surgery for clinical stage II or III colon cancer. Implications of all the available evidence Laparoscopic surgery with Japanese D3 dissection could be an acceptable treatment option for patients with stage II or III colon cancer. Since Japanese D3 dissection and complete mesocolic excision with central vascular ligation, use of which is widespread in the USA and Europe, are almost identical (ie, lymph node dissection along the feeding arteries is same, whereas the length of resected colon is shorter in Japanese D3 than in complete mesocolic excision with central vascular ligation), these results may be of value in the USA and Europe.

carcinoma were not regarded as multiple lesions); no intestinal obstruction that would result in insufficient bowel cleansing; who had no previous intestinal resection surgery (including stomach, but excluding appendectomy); no previous chemotherapy or radio­ therapy including treatment for other types of carcinoma; with tumour size of 8 cm or smaller diagnosed by imaging; who were aged 20–75 years old; and who had adequate organ function such as a leucocyte count of 3 × 10⁹ cells per L or higher, platelet count of at least 100 × 10⁹ cells per L, aspartate or alanine aminotransferase concentration 100 IU/L or lower, total bilirubin concen­tration 34·2 μmol/L or lower, and serum creatinine 132·6 μmol/L or lower. Exclusion criteria were synchronous or metachronous (within 5 years) malignancy other than carcinoma in situ, severe pulmonary emphysema, interstitial pneumonitis, or ischaemic heart disease, pregnant or lactating women, severe mental disease, and continuous systemic steroid therapy. All patients gave written informed consent. The study protocol was approved by the clinical trial review committee of JCOG and was approved and overseen by the institutional review board of each participating hospital. Ethics approval was obtained at each participating hospital.

Randomisation and masking The JCOG Data Center randomly assigned patients to the surgery groups and managed the study data. Eligible patients were randomly assigned (1:1) to have open surgery or laparoscopic surgery after a phone call or fax to the JCOG Data Center. Randomisation was done by a minimisation method with a biased-coin assignment according to tumour location (caecum, ascending versus sigmoid, rectosigmoid) and institution. The allocation procedure was not masked from the investigators or patients. www.thelancet.com/gastrohep Vol 2 April 2017

Articles

Procedures Surgeons with experience of more than 30 laparoscopic operations and 30 open operations for colorectal cancer were credited by the study chair (SK) and did laparoscopic surgery, and surgeons with experience of more than 30 open operations for colorectal cancer did open surgery. We peer reviewed the surgical procedure centrally using intraoperative photographs in all patients and video in arbitrarily selected patients. The surgical quality control committee assessed the photographs, and problems regarding the surgical procedure were discussed at meetings held twice per year. In both groups, resection of the colon with D3 dissection was done according to the Japanese Classification of Colorectal Carcinoma.14 Approaches for colonic mobilisation (median-to-lateral, lateral-tomedian, and retroperitoneal) were decided by the surgeon in charge. For right-sided lesions, the vascular pedicles were divided at their origin together with removal of draining lymph nodes (stations 203, 213, 223) along the superior mesenteric vein (appendix). For left-sided lesions, removal of lymph nodes at the root of the inferior mesenteric artery (253) was done with high ligation or with preservation of the left colic artery and ligation of the root of the superior rectal artery. Details of the surgical procedure have previously been reported by our group.13 When pathological stage III cancer was confirmed by histological examination of the resected specimen, adjuvant chemotherapy was administered as three courses of the Roswell Park Memorial Institute 8-week regimen15 of fluorouracil (500 mg/m² by bolus intravenous infusion on days 1, 8, 15, 22, 29, and 36) and leucovorin (250 mg/m² by 2 h drip intravenous infusion on days 1, 8, 15, 22, 29, and 36). All patients were followed up for at least 5 years after patient recruitment was completed. For surveillance after curative surgery, patients were observed periodically by their surgeons every 4 months for the first 2 years and then every 6 months for 3 years. Blood tests, abdominal and pelvic CT, and plain chest radiograph were done at each visit. Operative methods and pathology results were recorded according to the Japanese Classification of Colorectal Carcinoma14 and TNM classification.16

Outcomes The primary endpoint of this study was overall survival, defined as the time from randomisation to death from any cause or to the last date of contact with the surviving patient. Secondary endpoints were relapse-free survival, short-term clinical outcomes (time to pass first flatus or stool, postoperative hospital stay, maximum body temperature during hospital stay), adverse events, and the proportion of patients converting from laparoscopic surgery to open surgery. All laparoscopic surgery cases that required a skin incision of larger than 8 cm were www.thelancet.com/gastrohep Vol 2 April 2017

counted as a conversion to open surgery, except for cases in which retrieval of the resected specimen alone required this length of incision. Relapse-free survival was defined as the time from randomisation to relapse or death from any cause or to the last date on which relapse-free survival was verified. Operation time, blood test results, and all postoperative morbidities during the hospital stay were included on the case report forms for prospective data collection. Hospital mortality was defined as postoperative death due to any cause within 30 days of the initial surgery, or death within the same hospital stay. Adverse events, including postoperative morbidity, were described according to the Common Terminology Criteria for Adverse Events 3.0. Short-term outcomes in this study, which showed favourable results of laparoscopic surgery compared with open surgery, have previously been reported.13

For the Common Terminology Criteria for Adverse Events see http://ctep.cancer.gov

Statistical analysis The initial sample size needed to observe 254 deaths at the primary analysis was 818 patients, which was determined with a one-sided alpha of 0·05, power of 0·80, and non-inferiority margin for a hazard ratio (HR) of 1·366. This power calculation was based on an expected 5-year overall survival of 75%. However, during the patient accrual period, it became apparent that this was an underestimate of 5-year overall survival for this patient population; therefore, the sample size was increased to 1050 patients based on a projected 5-year overall survival of 82% to maintain the required statistical power. We planned to recruit patients for 4·5 years and follow-up patients for 5 years. We did two planned interim analyses. We used the Lan-DeMets method with an O’Brien-Fleming type alpha spending function17 to maintain a study-wise one-sided alpha error of 5%. We planned for the first interim analysis to take place 1 year after half of the planned sample size was accrued, and for the second interim analysis to take place 1 year after patient recruitment was completed. We used a significance level of 0·04999 in the final analysis. We used a stratified Cox proportional hazard model with tumour location as a stratum to test the non-inferiority of laparoscopic surgery to open surgery in terms of overall survival and to estimate HR. We used an unstratified Cox proportional hazard model to estimate HR of relapse-free survival, and the Kaplan-Meier method to estimate overall survival and relapse-free survival. We did prespecified subgroup analyses for overall survival with the Cox proportional hazards model. The p value was one-sided at 5% significance. The efficacy analyses were done in all randomly assigned patients in this study by intention-to-treat. As a sensitivity analysis, we analysed overall survival in patients who underwent surgery as assigned, excluding those patients who did not have the treatment as assigned. We did safety analyses in all patients who had surgery as assigned.

See Online for appendix

263

Articles

Results

1057 randomly assigned

528 assigned to open surgery 5 found ineligible after randomisation*

529 assigned to laparoscopic surgery 6 found ineligible after randomisation*

8 had laparoscopic surgery

520 had open surgery

4 had open surgery

525 had laparoscopic surgery

346 did not have adjuvant chemotherapy

174 had adjuvant chemotherapy

326 did not have adjuvant chemotherapy

199 had adjuvant chemotherapy

Figure 1: CONSORT diagram *Included in intention-to-treat analysis as prespecified.

Open surgery (n=528)

Laparoscopic surgery (n=529)

Sex Male

312 (59%)

282 (53%)

Female

216 (41%)

247 (47%)

Age (years)

64 (57–69)

64 (28–69)

Body-mass index (kg/m²) ≤20

85 (16%)

91 (17%)

21–25

319 (60%)

304 (57%)

>25

124 (23%)

134 (25%)

Clinical stage II

366 (69%)

331 (63%)

III

160 (30%)

197 (37%)

IV

2 (<1%)

1 (<1%)

Tumour location Caecum

55 (10%)

46 (9%)

Ascending

100 (19%)

109 (21%)

Sigmoid

236 (45%)

252 (48%)

Rectosigmoid

137 (26%)

122 (23%)

Data are n (%) or median (IQR).

Table 1: Patient characteristics

All statistical analyses were done with SAS (version 9.2). This study is registered with ClinicalTrials.gov, number NCT00147134, and UMIN Clinical Trials Registry, number C000000105.

Role of the funding source The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication. 264

Between Oct 1, 2004, and March 27, 2009, 1057 eligible patients were enrolled from 30 hospitals in Japan. Of 1057 patients, 528 (50%) were assigned to the open surgery group and 529 (50%) were assigned to the laparoscopic surgery group. After randomisation, 11 patients (1%; five [1%] patients in the open surgery group and six [1%] in the laparoscopic group) were found to be ineligible. In the open surgery group, liver metastasis was found in three patients by additional imaging before surgery, lung metastasis was confirmed in one patient by lung resection one month after colectomy, and a different colon cancer was found in one patient by additional endoscopy before surgery. In the laparoscopic group, liver metastasis was found in three patients by additional imaging before surgery, double cancer (gastric cancer) was found in one patient before surgery, a different colon cancer was found in one patient during surgery, and bowel obstruction was present in one patient before surgery. According to the intention-to-treat policy, 528 patients in the open surgery group and 529 patients in the laparoscopic surgery group were included in the efficacy analyses. Eight (15%) patients in the open surgery group and four (1%) patients in the laparoscopic surgery group did not undergo their assigned surgery, and these patients were excluded from the population for safety analyses. Therefore, 520 (98%) of 528 patients in the open surgery group and 525 (99%) of 529 patients in the laparoscopic surgery group were included in the operative safety analyses (figure 1). There were no substantial differences in baseline characteristics between the two groups (table 1). Operative and pathological results are summarised in table 2. Conversion to open surgery was necessary in 29 (5%) patients in the laparoscopic surgery group. The reasons for conversion have been reported elsewhere.13 The median number of harvested lymph nodes was similar in both groups (22 [IQR 15–31] in the open group vs 21 [16–30] in the laparoscopic group). Japanese D3 dissection was done in 524 (99%) of 528 patients in the open surgery group and in 527 (99%) of 529 patients in the laparoscopic surgery group. There was no substantial difference in terms of pathological stage in either group. In patients with pathological stage III cancer, 174 (86%) of 203 patients in the open surgery group and 199 (86%) of 232 patients in the laparoscopic surgery group were given adjuvant chemotherapy. Two treatment-related deaths occurred in the open surgery group; one patient died 7 days after surgery (probably due to myocardial infarction), and the other died from febrile neutropenia, pneumonia, diarrhoea, and gastrointestinal haemorrhage during postoperative chemotherapy. One patient died during the hospital stay and was therefore excluded from the adverse events analysis. 65 (13%) of 519 patients in the open surgery group and 53 (10%) of 525 patients in the laparoscopic surgery group had one or more grade 2–4 adverse events www.thelancet.com/gastrohep Vol 2 April 2017

Articles

Open surgery (n=528)

Laparoscopic surgery (n=529)

Operative results lleo-caecal resection

54 (10%)

37 (7%)

Right hemicolectomy

99 (19%)

111 (21%)

Sigmoldectomy

214 (41%)

234 (44%)

Anterior resection

154 (29%)

138 (26%)

Hartmann’s procedure

1 (<1%)

Partial resection

4 (1%)

7 (1%)

Others

2 (<1%)

2 (<1%)

D3 dissection Conversion

524 (99%) ··

0

527 (100%)

Number of lymph nodes harvested

4·5 (3·5–5·5) 22 (15–31)

4·3 (3·5–5·5) 21 (16–30)

Pathological stage 0

4 (1%)

2 (<1%)

I

59 (11%)

58 (11%)

II

246 (47%)

222 (42%)

III

203 (38%)

232 (44%)

IV

15 (3%)

15 (3%)

Missing

1 (<1%)

0

Data are n (%) or median (IQR).

Table 2: Operative and pathological results

(table 3). Late grade 2–4 adverse events included constipation, diarrhoea, paralytic ileus, and small bowel obstruction (table 3). Median follow-up for all patients following random­ isation was 72·8 months (IQR 61·1–72·8). At the time of last follow-up on March 27, 2014, 128 (12%) of 1057 patients had died; 62 (12%) of 528 patients in the open surgery group and 66 (12%) of 529 patients in the laparoscopic surgery group. 64 patients (36 patients in the open surgery group and 28 patients in the laparoscopic surgery group) were lost to follow-up within 5 years of enrolment. Estimated 5-year overall survival was 90·4% (95% CI 87·5–92·6) in the open surgery group and 91·8% (89·1–93·8) in the laparoscopic surgery group. The HR for overall survival for laparoscopic surgery versus open surgery was 1·06 (90% CI 0·79–1·41, one-sided pnon-inferiority=0·073; figure 2A); laparoscopic surgery was not non-inferior to open surgery. In sensitivity analysis of 1045 patients who had surgery as assigned, the HR for overall survival for laparoscopic surgery versus open surgery was 1·03 (0·77–1·38, one-sided pnon-inferiority=0·057). Subgroup analyses for overall survival were done for sex (male vs female), age (<65 vs ≥65 years), tumour location (caecum, ascending, sigmoid colon vs rectosigmoid colon vs upper rectum), clinical stage (II vs III), clinical T stage (cT3 vs cT4), clinical N stage (cN0 vs cN1 vs cN2), and body-mass index (BMI; ≤20 vs >20 to 25 vs >25 kg/m²; figure 2B). Patients with tumours located in the www.thelancet.com/gastrohep Vol 2 April 2017

Number of patients who had one or more late grade 2–4 adverse events

65 (13%)

Laparoscopic surgery (n=525) 53 (10%)

Constipation

31 (6%)

23 (4%)

Diarrhoea

15 (3%)

14 (3%)

Paralytic ileus Bowel obstruction of small intestine

6 (1%)

9 (2%)

16 (3%)

11 (2%)

Data are n (%). *One patient died during their hospital stay and was excluded.

Table 3: Late grade 2–4 adverse events

29 (5%)

Pathological results Tumour size (cm)

Open surgery (n=519*)

rectosigmoid, who were clinical T4, clinical N2, or had high BMI (>25) who underwent laparoscopic surgery tended to show worse survival compared with patients in open surgery group (figure 2B). Relapse-free survival is shown in figure 2C. 228 (22%) of 1057 patients had recurrence or death; 111 (21%) of 528 patients in the open surgery group and 117 (22%) of 529 patients in the laparoscopic surgery group. 5-year relapse-free survival was 80% (95% CI 76·0–82·9) in the open surgery group and 79% (75·6–82·6) in the laparoscopic surgery group (figure 2C). The HR for relapse-free survival for laparoscopic surgery versus open surgery was 1·07 (95% CI 0·82–1·38). 89 (17%) of 520 patients in the open surgery group and 101 (19%) patients in the laparoscopic surgery group had recurrence after R0 resection. Of these patients, 39 (44%) in the open surgery group and 40 (40%) in the laparoscopic surgery group had liver metastasis, ten (11%) in the open surgery group and 16 (16%) in the laparoscopic surgery group had peritoneal metastasis, 31 (35%) in the open surgery group and 33 (33%) in the laparoscopic surgery group had lung metastasis, and 12 (13%) in the open surgery group and 15 (15%) in the laparoscopic surgery group had lymph node metastasis.

Discussion Our data show that overall survival after laparoscopic D3 surgery was not non-inferior to open D3 surgery. However, overall survival was more favourable than expected during protocol design in both groups. Despite not being non-inferior, because the overall survival in both groups was similar and better than expected, laparoscopic D3 seems to be acceptable as a treatment option for stage II or III colon cancer. To the best of our knowledge, this is the first randomised controlled trial focused on Japanese D3 or complete mesocolic excision with central vascular ligation to assess long-term outcomes. These techniques are reflective of the current standard of care in Japan and the USA and Europe, respectively, which has not been the case for previous randomised controlled trials. One possible explanation why non-inferiority was not shown was that survival was better than expected in both groups. The observed 5-year overall survival (90·4% in the 265

Articles

Overall survival (%)

A 100 90 80 70 60 50 40 30 20 10 0

Open surgery Laparoscopic surgery

HR 1·06 (90% CI 0·79–1·41; p=0·073)* 0

1

Number at risk (censored) Open surgery 528 (0) Laparoscopic surgery 529 (0)

2

3

4

5

6

7

8

9

10

164 (125) 163 (127)

71 (91) 75 (87)

12 (59) 11 (63)

0 (12) 0 (11)

Time after randomisation (years) 523 (1) 526 (1)

516 (2) 517 (1)

499 (4) 504 (2)

486 (2) 487 (5)

442 (27) 458 (19)

291 (143) 299 (147)

B Sex Male Female Age (years) <65 ≥65 Tumour location Caecum Ascending Sigmoid Rectosigmoid Clinical stage† II III T factor cT3 cT4 N factor cN0 cN1 cN2 Body-mass index (kg/m²) ≤20 >20 to 25 >25 Overall

Open surgery (event/n)

Laparoscopic surgery (event/n)

HR (95% CI)

37/312 25/216

37/282 29/247

1·09 (0·69–1·71) 1·02 (0·60–1·74)

32/275 30/253

32/286 34/243

0·93 (0·57–1·53) 1·22 (0·75–2·00)

6/56 11/99 30/235 15/138

4/46 15/109 27/250 20/124

0·80 (0·23–2·83) 1·23 (0·56–2·67) 0·82 (0·49–1·38) 1·55 (0·80–3·04)

39/366 22/160

37/331 28/197

1·04 (0·67–1·64) 1·01 (0·58–1·76)

44/407 18/121

43/410 23/119

0·97 (0·64–1·48) 1·27 (0·68–2·34)

39/367 21/126 2/35

37/331 22/161 7/37

1·05 (0·67–1·64) 0·78 (0·43–1·42) 3·61 (0·75–17·37)

11/85 43/319 8/124 62/528

13/91 35/304 18/134 66/529

1·12 (0·50–2·49) 0·82 (0·53–1·28) 2·22 (0·97–5·12) 1·06 (0·75–1·49) 0·25

266

2·0

Favours laparoscopic surgery

4·0

8·0

32·0

16·0

Favours open surgery

C

Relapse-free survival (%)

Figure 2: Overall and relapse-free survival (A) Kaplan-Meier curve for comparison of overall survival, (B) subgroup analysis of overall survival, and (C) Kaplan-Meier curve for comparison of relapse-free survival. HR=hazard ratio. *Stratified Cox proportional hazard model according to tumour location (for non-inferiority; non-inferiority margin of HR was set at 1·366). †Two patients in open surgery and one patient in laparoscopic surgery were excluded because they had stage IV cancer. ‡Unstratified Cox proportional hazard model.

1·0

0·5

100 90 80 70 60 50 40 30 20 10 0

Open surgery Laparoscopic surgery

HR 1·07 (95% CI 0·82–1·38)‡ 0

Number at risk (censored) Open surgery 528 (0) Laparoscopic surgery 529 (0)

1

2

3

4

5

6

7

8

9

10

135 (105) 126 (114)

60 (74) 56 (69)

10 (49) 6 (50)

0 (10) 0 (6)

Time after randomisation (years) 484 (3) 477 (1)

446 (2) 441 (1)

426 (3) 422 (1)

417 (1) 416 (3)

383 (30) 386 (28)

240 (140) 244 (139)

www.thelancet.com/gastrohep Vol 2 April 2017

Articles

open surgery group and 91·8% in the laparoscopic surgery group) was better than our initial assumption by about 15%. Despite increasing the sample size based on a better expected 5-year overall survival, the number of events observed—128 (50%) of 254 of expected deaths had occurred by the time of the primary analysis—was still insufficient. Had we maintained sufficient power with an expected 5-year overall survival of 90%, we would have needed more than 2500 patients, but this was not practically possible. The reasons for good long-term survival could include advancements in surgical techniques (eg, a high proportion of precise D3 lymph node dissection and reduced postoperative compli­cations using energy and anatomical devices); use of chemotherapy (eg, the antimolecular drugs during the present study); and a higher proportion of patients with tumours at stage II than stage III. Additionally, the enrolled patients were a carefully chosen population because of the use of strict inclusion and exclusion criteria, which might have led to good long-term outcomes. However, we did not collect data on post-recurrence treatments, which might have differed between groups and thus affected outcomes. This study differed from previous randomised controlled trials in several ways: the study population was limited to patients with clinical stage II and III disease; a uniform surgical procedure including the extent of lymph node dissection (Japanese D3) was done; surgical quality control and assurance were done; and uniform adjuvant chemotherapy was given. Definition of the degree of lymph node dissection and adjuvant chemotherapy was especially important to evaluate long-term outcomes. Additionally, surgical quality control led to lower conversion rates than noted in other trials. This study is, to our knowledge, the first to compare and report long-term outcomes of laparoscopic surgery versus open surgery for clinical stage II or III colon cancer and to be done to overcome issues noted for earlier randomised controlled trials. In terms of quality control of surgery, Japanese D3 dissection was done in almost all patients in both laparoscopic surgery and open surgery groups. Japanese D3 dissection and the current standard procedure in the USA and Europe of complete mesocolic excision with central vascular ligation are almost identical although the resected colon is slightly shorter in the Japanese D3 procedure. Theoretically, the procedures should be equivalent because the principles of both are the same. Findings from a 2012 study18 showed that complete mesocolic excision with central vascular ligation and Japanese D3 dissection proved to be superior to previously reported cases. However, long-term outcomes of laparoscopic complete mesocolic excision with central vascular ligation or D3 dissection have not been reported. In 2015, Bertelsen and colleagues19 reported in a retrospective study that complete mesocolic excision was superior to conventional colon surgery in terms of disease-free survival. However, there are no www.thelancet.com/gastrohep Vol 2 April 2017

reports of long-term outcome of complete mesocolic excision or Japanese D3 dissection in prospective studies. To our best knowledge, this study is the first to prospectively assess the long-term outcomes of patients with colon cancer who underwent Japanese D3 lymph node dissection. Since this technique is broadly equivalent to complete mesocolic excision with central vascular ligation we believe our results could be useful in western countries. Findings from subgroup analyses for overall survival highlight the continuing need to identify the patient subgroups associated with unfavourable outcomes who have rectosigmoid tumours, high BMI, advanced clinical node metastasis (N2), and T4 staging in the laparoscopic surgery group versus the open surgery group (figure 2B).20–27 We speculate that manipulations with forceps and pneumoperitoneum during laparoscopic surgery might affect long-term outcomes, especially in the patients with high BMI, advanced clinical node metastasis, and T4 staging. Although only patients with T4 disease without involvement of other organs were eligible in the present study, patients with T4 disease in the laparoscopic surgery group tended to show worse survival compared with those in the open surgery group. Careful consideration must be given as to whether to operate laparoscopically on patients with T4 disease. By contrast with the results of our subgroup analyses, previous reports showed the feasibility and safety of laparoscopic surgery compared with open surgery in terms of BMI, clinical node metastasis, T4 stage, and rectosigmoid tumours.20–28 Further examination will clarify whether there are patient subgroups for whom laparoscopic surgery is indicated who require additional attention. We are planning subgroup analyses to assess the efficacy of laparoscopic surgery for patients who are obese, and who have clinical node metastasis, T4, or rectosigmoid cancers. The results of our trial suggest that open surgery should be the first recommendation for surgery, but that laparoscopic surgery is also an option. Our previous study revealed the benefit of laparoscopic surgery on short-term outcomes including wound complications such as incisional hernia and wound dehiscence.13 We did not find any differences in late complications, perhaps because of the lower number of events in our study. In conclusion, laparoscopic surgery with Japanese D3 dissection was not non-inferior to open surgery in terms of overall survival for patients with stage II or III colon cancer. However, because overall survival in both groups was similar and better than expected, laparoscopic surgery with Japanese D3 dissection appears to be acceptable as a treatment option for patients with stage II or III colon cancer. Contributors All authors were involved in the conception and design of the study, and the provision of study material, patients, and data acquisition. JM and HK were responsible for data management, statistical analysis, and data interpretation. All authors were involved in the writing of the report and approved the final decision to submit for publication. 267

Articles

Declaration of interests SK reports grants from Ministry of Health, Labour and Welfare, Japan, during the conduct of the study; and non-financial support from Olympus Medical Science Company, Ethicon (a Johnson & Johnson company), and Covidien Japan, outside the submitted work. MIn reports grants from Ministry of Health, Labour and Welfare, Japan, during the conduct of the study; and non-financial support from Olympus Medical Science Company, Ethicon, and Covidien Japan, outside the submitted work. JM reports grants from Ministry of Health, Labour and Welfare, Japan, during the conduct of the study. HK report grants from Ministry of Health, Labour and Welfare, Japan, during the conduct of the study. MW reports grants from Ministry of Health, Labour and Welfare, Japan, during the conduct of the study; non-financial support from Olympus Medical Science Company, Johnson & Johnson, and Covidien Japan, outside the submitted work. SY reports grants from Ministry of Health, Labour and Welfare, Japan, during the conduct of the study; and personal fees from Johnson and Johnson, Covidien Japan, Takeda Pharmaceutical, Merk Serono, Chugai Pharmaceutical, Taiho Phamaceutical, Eli Lilly Japan, Daiichi Sankyo, Asahi Kasei Pharma Corporation, Eisai, Sumitomo Dainippon Pharma, outside the submitted work. SS reports grants from Ministry of Health, Labour and Welfare, Japan, during the conduct of the study; and personal fees from Ethicon Endo-surgery and Covidien Japan, outside the submitted work. KS reports grants from Ministry of Health, Labour and Welfare, Japan, during the conduct of the study; and non-financial support from Takeda Pharmaceutical, Merk Serono, Chugai Pharmaceutical, Taiho Phamaceutical, Eli Lilly Japan, Sumitomo Dainippon Pharma, Bayer Yakuhin, and Novartis Pharma, outside the submitted work. TM reports grants from Ministry of Health, Labour and Welfare, Japan, during the conduct of the study; and non-financial support from Olympus Medical Science Company, Ethicon, and Covidien Japan, outside the submitted work. YSh reports grants from Ministry of Health, Labour and Welfare, Japan, during the conduct of the study; grants and personal fees from Taiho, Chugai, Lilly, Merck Serono; and personal fees from Takeda, Yakult, Ono, Daiichi-Sankyo, Bayer, Novartis, and AstraZeneca, outside the submitted work. MIt, SF, FK, YSa, HH, TA, TY, YF, KM, MO, and YM declare no competing interests. Acknowledgments We thank all of the participating patients and their families, as well as investigators, research nurses, study coordinators, and operations staff; and the members of the JCOG Data Center and JCOG Operations Office for their support in preparing the report, data management, and overseeing the study management. This study was supported in part by the National Cancer Center Research and Development Fund (23-A-16, 23-A-19, 26-A-4), Grant-in-Aid for Cancer Research (14S-3, 14S-4, 17S-3, I7S-5, 20S-3 and 20S-6), and Health and Labour Sciences Research Grant for Clinical Cancer Research (Hl5-018, H18-013, H21-017, H24-005) from the Ministry of Health, Labour and Welfare of Japan. References 1 Lacy AM, García-Valdecasas JC, Delgado S, et al. Laparoscopy-assisted colectomy versus open colectomy for treatment of non-metastatic colon cancer: a randomised trial. Lancet 2002; 359: 2224–29. 2 Lacy AM, Delgado S, Castells A, et al. The long-term results of a randomized clinical trial of laparoscopy-assisted versus open surgery for colon cancer. Ann Surg 2008; 248: 1–7. 3 Clinical Outcomes of Surgical Therapy Study Group. A comparison of laparoscopically assisted and open colectomy for colon cancer. N Engl J Med 2004; 350: 2050–59. 4 Fleshman J, Sargent DJ, Green E, et al. Laparoscopic colectomy for cancer is not inferior to open surgery based on 5-year data from the COST Study Group trial. Ann Surg 2007; 246: 655–62. 5 Veldkamp R, Kuhry E, Hop WC, et al. Laparoscopic surgery versus open surgery for colon cancer: short-term outcomes of a randomised trial. Lancet Oncol 2005; 6: 477–84. 6 The Colon Cancer Laparoscopic or Open Resection Study Group. Survival after laparoscopic surgery versus open surgery for colon cancer: long-term outcome of a randomised clinical trial. Lancet Oncol 2009; 10: 44–52. 7 Guillou PJ, Quirke P, Thorpe H, et al. Short-term endpoints of conventional versus laparoscopic-assisted surgery in patients with colorectal cancer (MRC CLASICC trial): multicentre, randomised controlled trial. Lancet 2005; 365: 1718–26.

268

8

9

10 11

12

13

14 15

16 17 18

19

20 21 22 23 24 25

26

27 28

Jayne DG, Thorpe HC, Copeland J, Quirke P, Brown JM, Guillou PJ. Five-year follow-up of the Medical Research Council CLASICC trial of laparoscopically assisted versus open surgery for colorectal cancer. Br J Surg 2010; 97: 1638–45. Hewett PJ, Allardyce RA, Bagshaw PF, et al. Short-term outcomes of the Australasian randomized clinical study comparing laparoscopic and conventional open surgical treatments for colon cancer: the ALCCaS trial. Ann Surg 2008; 248: 728–38. Neudecker J, Klein F, Bittner R, et al. Short-term outcomes from a prospective randomized trial comparing laparoscopic and open surgery for colorectal cancer. Br J Surg 2009; 96: 1458–67. Higuchi T, Sugihara K. Complete mesocolic excision (CME) with central vascular ligation (CVL) as standardised surgical technique for colonic cancer: a Japanese multicentre study. Dis Colon Rectum 2010; 53: 646. Kitano S, Inomata M, Sato A, Yoshimura K, Moriya Y, for the Japan Clinical Oncology Group Study. Randomized controlled trial to evaluate laparoscopic surgery for colorectal cancer: Japan Clinical Oncology Group Study JCOG 0404. Jpn J Clin Oncol 2005; 35: 475–77. Yamamoto S, Inomata M, Katayama H, et al. Short-term surgical outcomes from a randomized controlled trial to evaluate laparoscopic and open D3 dissection for stage II/III colon cancer: Japan Clinical Oncology Group Study JCOG 0404. Ann Surg 2014; 260: 23–30. Watanabe T, Itabashi M, Shimada Y, et al. Japanese Society for Cancer of the Colon and Rectum (JSCCR) Guideline 2014 for treatment of colorectal cancer. Int J Clin Oncol 2015; 20: 207–39. Petrelli N, Douglass HO Jr, Herrera L, et al. The modulation of fluorouracil with leucovorin in metastatic colorectal carcinoma: a prospective randomized phase III trial. Gastrointestinal Tumor Study Group. J Clin Oncol 1989; 7: 1419–26. Sobin LH, Wittekind C. TNM Classification of Malignant Tumours, 6th edition. Hoboken, NJ: John Wiley & Sons, 2002. Lan KKG, DeMets DL. Discrete sequential boundaries for clinical trials. Biometrika 1983; 70: 659–63. West NP, Kobayashi H, Takahashi K, et al. Understanding optimal colonic cancer surgery: comparison of Japanese D3 resection and European complete mesocolic excision with central vascular ligation. J Clin Oncol 2012; 30: 1763–69. Bertelsen CA, Neuenschwander AU, Jansen JE, et al. Disease-free survival after complete mesocolic excision compared with conventional colon cancer surgery: a retrospective, population-based study. Lancet Oncol 2015; 16: 161–68. Elnahas A, Sunil S, Jackson TD, Okrainec A, Quereshy FA. Laparoscopic versus open surgery for T4 colon cancer: evaluation of margin status. Surg Endosc 2016; 30: 1491–96. Feliciotti F, Guerrieri M, Paganini AM, et al. Long-term results of laparoscopic versus open resections for rectal cancer for 124 unselected patients. Surg Endosc 2003; 17: 1530–35. Hotta T, Takifuji K, Yokoyama S, et al. The impact of obesity on learning laparoscopic surgery for colon cancer. J Laparoendosc Adv Surg Tech A 2012; 22: 635–40. Kim IY, Kim BR, Kim YW. The short-term and oncologic outcomes of laparoscopic versus open surgery for T4 colon cancer. Surg Endosc 2016; 30: 1508–18. Kim JH, Ahn BK, Park SJ, et al. Long-term outcomes of laparoscopic versus open surgery for rectal cancer: a single-center retrospective analysis. Korean J Gastroenterol 2015; 65: 273–82. Liang JT, Lai HS, Huang J, Sun CT. Long-term oncologic results of laparoscopic D3 lymphadenectomy with complete mesocolic excision for right-sided colon cancer with clinically positive lymph nodes. Surg Endosc 2015; 29: 2394–401. Makino T, Trencheva K, Shukla PJ, et al. The influence of obesity on short- and long-term outcomes after laparoscopic surgery for colon cancer: a case-matched study of 152 patients. Surgery 2014; 156: 661–68. Nakamura T, Miura H, Ikeda A, et al. Laparoscopic surgery for obese patients with colon cancer: a case-matched control study. Surg Today 2013; 43: 763–68. Watanabe J, Tatsumi K, Ota M, et al. The impact of visceral obesity on surgical outcomes of laparoscopic surgery for colon cancer. Int J Colorectal Dis 2014; 29: 343–51.

www.thelancet.com/gastrohep Vol 2 April 2017