Introduction

As described by Hohenberger et al., complete mesocolic excision (CME) with central vascular ligation (CVL) is an anatomically based resection of colon cancer and locoregional lymph nodes that do not breach the visceral fascia, avoiding tumor spread outside of the mesocolon and ensuring complete lymphadenectomy [1]. Multiple studies have indicated that CME with CVL provides better oncological outcomes than standard colectomy [2]. Japanese D3 dissection is a similar surgical technique that ensures oncological sharp dissection of the embryonic plane and complete lymphadenectomy with central vascular ligation and further “D3 lymph nodes”, which are the central lymph nodes along the superior mesenteric vein or artery. This procedure is strongly recommended for T3 and deeper colon cancers by the guideline of the Japanese Society for Cancer of the Colon and Rectum (JSCCR), based on the excellent oncological outcomes from this procedure [3, 4]. With recent technical improvements in laparoscopic colectomy, D3 dissection has been performed increasingly frequently through a laparoscopic approach in Japan and other countries [5, 6]. However, the oncological safety of a laparoscopic approach for CME with D3 dissection has been controversial. Although some studies reported safe short-term and oncological outcomes [7], a recent phase III randomized controlled trial failed to show the non-inferiority of laparoscopic CME with D3 dissection compared to open CME for stage II/III colon cancer [8]. Further, subgroup analysis identified that patients with cT4a, cN2 and obesity tended to display poorer survival in the laparoscopic arm. The JSCCR guideline recommends limiting the use of laparoscopic colectomy with D3 dissection only to expert surgeons due to a lack of evidence regarding the certainty of oncological safety [3]. Although previous randomized trials have depicted laparoscopic surgery as safe and feasible compared to open surgery, no studies have successfully shown statistical noninferiority in terms of OS or RFS [9,10,11]. On the other hand, a recent study reported better oncological outcomes from a laparoscopic approach compared to an open approach [12, 13].

In this study, we aimed to examine the oncological safety of laparoscopic CME with D3 dissection for pT3/4a M0 colon cancer, and to examine whether increased use of a laparoscopic approach changed the outcomes in a cohort at a comprehensive cancer center.

Patients and methods

This retrospective study collected data from 1510 consecutive patients with pathologically confirmed T3 and T4a colon cancer from the cecum to the rectosigmoid colon, who underwent surgery between July 2004 and December 2013 at the Cancer Institute Hospital of the Japanese Foundation for Cancer Research (CIH). Exclusion criteria were synchronous malignancies (n = 29), emergency surgery (n = 12), appendiceal tumor (n = 15), familial tumor (familial adenomatous polyposis or lynch syndrome) (n = 12), inflammatory bowel disease (n = 10), adenosquamous carcinoma, neuroendocrine tumor or signet-ring cell carcinoma (n = 6), and D2 lymph node dissection (n = 235). Finally, a total of 1191 patients were eligible. No patients received neoadjuvant chemotherapy or chemoradiotherapy.

Preoperative staging included total colonoscopy and contrast-enhanced computed tomography (CT) of the chest, abdomen and pelvis. Magnetic resonance imaging (MRI) and positron emission tomography (PET) were added at the discretion of the treating physician.

Postoperative complications were monitored for 30 days after surgery and were graded according to the Clavien-Dindo classification [14, 15]. Adjuvant chemotherapy was administered to patients with stage III or high-risk stage II disease after histological evaluation of surgical specimens, as recommended in national guidelines [3]. General practice for postoperative surveillance of stage I–III colon cancer was also in accordance with national guidelines, including physical examination, interval history, serum carcinoembryonic antigen (CEA) testing, and imaging (most frequently CT) of the chest, abdomen, and pelvis with intravenous contrast, at 3–6-month intervals for the first 3 years and at 6-month intervals thereafter for at least 5 years. Colonoscopy was typically performed at 1 year after surgery, then repeated every 2–3 years unless advanced adenoma was identified. Radiographic reports were reviewed, and a definitive diagnosis of recurrence was based on the appearance of new lesions on CT, MRI, and/or PET and/or histological confirmation through biopsy.

Data on patient demographics, perioperative clinical outcomes, pathological outcomes, and disease status at last follow-up were collected from the prospectively maintained database at CIH, and electronic medical records were reviewed. Informed consent was obtained in the form of an opt-out option on the hospital website. The protocols for this study were reviewed and approved by the Clinical Research Review Board of CIH (Research Registry No. 1025).

Surgical procedures

Surgical procedures for T3/4a colon cancer without distant metastasis were performed in accordance with JSCCR guidelines [3]. We performed a modified CME (mCME) using a similar surgical technique to the original approach, with some technical differences that have been described previously [1, 4, 6, 7]. After performing mobilization of the mesocolon, the supplying vessels were ligated to perform D3 lymph node dissection. The horizontal margin from the tumor was at least 10 cm from the tumor as defined by the Japanese guideline [3], which was shorter than the original CME reported by Hohenberger [1, 16]. For right colectomies, central vessel ligation and D3 dissection required complete removal of lymphatic tissue on the surface of the superior mesenteric vein (SMV). For left colectomies, the high tie of the inferior mesenteric artery (IMA) or low tie with preservation of the left colic artery and complete removal of lymphatic tissue from the root to the point of the division were performed. Both open and laparoscopic procedures were performed or supervised by attending colorectal surgeons who are board-certified and well experienced in the procedures.

Statistical analysis

Data from different groups were compared using Student’s t test. Continuous data were expressed as mean ± standard deviation (SD). In univariate analysis, comparison of categorical variables was performed using the chi-square test or Fisher’s exact test. Overall survival (OS) and relapse-free survival (RFS) were calculated using Kaplan–Meier methods, and differences were tested using the log-rank test. Cox hazard models were used to determine independent factors affecting survival. Statistical analysis was performed using JMP version 12.1.0 software (SAS Institute, Cary, NC). Values of p < 0.05 were considered to indicate statistical significance.

Results

Proportion of surgical approaches and patient characteristics

A total of 1191 patients were eligible for this study. The proportion of laparoscopic approaches increased markedly during the study period, from 3.2% in 2004 to 96.5% in 2013 (Fig. 1a). The cohort was divided into an early period (2004–2008, n = 450) and late period (2009–2014, n = 741). The number of patients with comorbidity was higher in the late period (Table 1). No significant differences in sex, age, tumor location, history of abdominal surgery, or preoperative CEA level were identified.

Fig. 1
figure 1

a Proportion of laparoscopic approaches during the study period. Comparison between study periods for relapse-free survival (b) and overall survival (c) in patients with colon cancer

Full size image
Table 1 Clinical characteristics of patients in the early and late periods
Full size table

Perioperative and pathological outcomes

Perioperative and pathological outcomes are shown in Table 2. Operative time was longer and blood loss were both lower in the late period (early vs late period: 200 min vs 185 min, p < 0.01; 20 mL vs 10 mL, p < 0.01). The late period included more pT4a tumors (16.9% vs. 23.2%, p < 0.01). Lymph node yield (18.1 vs. 21.6, p < 0.01) and the proportion of patients with < 12 lymph nodes (13.6% vs. 5.8%, p < 0.01) were lower in the late group. No significant differences were seen in N stage, TNM stage, or tumor size. Postoperative complication rate (Grade III or IV) (2.7% vs. 2.7%, p = 0.97) did not differ between groups. No mortality was encountered. Patients who received adjuvant chemotherapy with oxaliplatin were more common among Stage III patients (14.3% vs. 48.1%, p < 0.01). When analyzed by tumor laterality, lymph node yield was higher in the late period for both right and left colon cancer (Fig. 2a, b).

Table 2 Pathological and perioperative characteristics of patients between early and late periods
Full size table
Fig. 2
figure 2

Comparison of number of dissected lymph nodes between laparoscopic and open surgeries

Full size image

Oncological outcomes

Median durations of follow-up in the early and late periods were 64 and 54 months, respectively. Overall, the late period showed better RFS (75.3% vs. 82.7% at 5 years, p < 0.01) and OS (86.9% vs. 91.7% at 5 years, p = 0.01) compared to the early period (Fig. 3a, b). When cohorts were subdivided by stage, patients in the late period displayed better 5-year RFS (71.3% vs. 86.2% in stage II, p < 0.05; 63.3% vs. 79.5% in stage III, p < 0.05) and better 5-year OS (83.3% vs. 93.7% in stage II, p < 0.05; 73.2% vs. 91.9% in stage III, p < 0.05) (Fig. 1b, c).

Fig. 3
figure 3

Comparison between study periods for relapse-free survival (a) and overall survival (b) by TNM stage

Full size image

Prognostic factors for 5-year RFS and OS are shown in Table 3. Univariate analysis showed that elevated CEA, open surgery, combined resection, lymphovascular invasion, LN yield < 12, lymph node metastasis, early study period and pT4 were significantly associated with lower 5-year-RFS. Multivariate analysis revealed laparoscopic surgery (hazard ratio (HR) 0.67; 95% confidence interval (CI) 0.50–0.91, p = 0.01) as an independent prognostic factor along with lymphovascular invasion (HR 1.56; 95% CI 1.08–2.26, p = 0.01), LN yield < 12 (HR 1.56; 95% CI 1.08–2.26, p = 0.01) and lymph node metastasis (HR 1.41; 95% CI 1.06–1.87, p = 0.01), and pT4 (HR 2.25; 95% CI 1.70–2.99, p < 0.01). In univariate analysis for OS, age > 75 years, male sex, open surgery, longer operation time, combined resection, poor histological grade, lymphovascular invasion, lymph node metastasis, postoperative complications, early study period and T4 were associated with lower 5-year OS. Multivariate analysis revealed laparoscopic surgery (HR 0.56; 95% CI 0.37–0.83, p < 0.01) as an independent prognostic factor along with age > 75 years (HR 2.57; 95% CI 0.80–3.67, p < 0.001), male (HR 1.61; 95% CI 1.13–2.29, p = 0.007), longer operation time (HR 1.57; 95% CI 1.03–2.39, p = 0.03), poor histological grade (HR 2.57; 95% CI 1.56–4.24, p < 0.001), and pT4 (HR 1.78; 95% CI 1.21–2.63, p = 0.003). Repeat analyses using cancer-specific survival showed similar results for comparison of the early and late periods and multivariate analysis (Supple Fig. 1a, b, Supple Table 1). Overall, recurrences occurred at a median of 18.4 months (range 2–67 months). The most common site of recurrence was the liver, followed by lung and peritoneal dissemination. No differences in site of recurrence were seen between laparoscopic and open procedures or between the early and late periods (Table 4).

Table 3 a Uni- and multivariate analyses for prediction of relapse-free survival, b Uni- and multivariate analyses for prediction of overall survival
Full size table
Table. 4 a Comparison of recurrence pattern between laparoscopic and open procedure, b comparison of recurrence pattern between the early and late period
Full size table

Discussion

In the present study, a cohort of consecutive patients with pT3/4a M0 colon cancer who underwent mCME with D3 dissection was analyzed. With the increased use of laparoscopic approaches, patients in the late period (2009–2013) exhibited better survival outcomes compared to the early period (2004–2008). After multivariate analyses, laparoscopic surgery remained as a prognostic factor associated with better RFS and OS The improvement in outcomes we observed in the 2009–2013 period relative to the 2004–2008 period likely results from multiple factors, including better staging by enhanced imaging using multidetector-row helical CT and liver MRI, increased use of oxaliplatin-based regimens in adjuvant chemotherapy and a better understanding of the CME concept with greater lymph node yield for curative resection. With these bundled together, improved oncologic outcomes with an evident increase in laparoscopic approach would justify this approach for mCME with D3 dissection in pT3/4a M0 colon cancer.

Adequate lymph node evaluation is central to the prognosis of colon cancer patients, possibly serving as a surrogate marker for surgical quality [17,18,19]. Several guidelines have shown < 12 harvested lymph nodes as a predictor of poor prognosis, and have recommended postoperative adjuvant chemotherapy for Stage II patients [3, 20]. Modified CME plus CVL and Japanese D3 apply the same concept of surgical resection of the embryological plane and true central ligation of the supplying artery [1, 2]. Hohenberger et al. reported an average of 32 harvested lymph nodes in CME and discussed the number of harvested lymph nodes as an indicator of CME quality [1]. Multiple studies have investigated the benefits of laparoscopic CME or Japanese D3 [7, 12, 21]. Although some studies have reported fewer harvested lymph nodes in laparoscopic surgery compared to open surgery [22], recent studies have reported similar lymph node yields between these two approaches [7, 8]. In the present study, more lymph nodes were harvested from patients in the late period than in the early period, regardless of the use of a laparoscopic or open approach (Fig. 2), resulting in fewer patients with < 12 harvested lymph nodes. Multivariate analysis revealed laparoscopic surgery and number of harvested lymph nodes as independently associated with better RFS. Such data suggest that the increased use of a laparoscopic approach in the latter period was accompanied by a better quality of CME consequently improving oncologic outcomes.

The JSCCR guideline recommended that the use of laparoscopic surgery for D3 resection should be limited to experts, due to technical difficulties and a lack of sufficient evidence for this procedure [3]. A subgroup analysis of the JCOG 0404 randomized trial, which compared laparoscopic and open D3 dissection for cStage II–III colon cancer, revealed some differences in short-term outcomes between high- and low-volume centers [8]. In Japan, a board-certification system called the Japanese Endoscopic Surgical Skill Qualification System (JESSQS) has been established to assess the skill of laparoscopic surgeons [23]. For JESSQS accreditation, two expert referee surgeons evaluate unedited videos of laparoscopic colectomy from the applicant in a double-blinded fashion, and the pass rate in the field of colorectal surgery is below 30% each year. In the present study, all open and laparoscopic procedures were performed or supervised by board-certificate surgeons [24]. Interestingly, improved oncologic outcomes in laparoscopic D3 dissection compared to open surgery have also been reported from a large center in Korea [12, 13]. The laparoscopic approach to colon cancer by an experienced surgeon might contribute to improved oncological outcomes.

A subgroup analysis of the JCOG0404 randomized trial for OS suggested poorer survival from the laparoscopic approach among patients with cT4 and/or cN2 disease [8]. The authors speculated that pneumoperitoneum and manipulations with forceps during the operation might have affected long-term outcomes. Some studies have also indicated unfavorable effects of pneumoperitoneum and instrumental manipulation during laparoscopic surgery causing peritoneal dissemination [25, 26]. Several randomized trials have also demonstrated a higher incidence of perineal dissemination in laparoscopic surgery compared to open surgery [27,28,29]. The present study did not identify any significant differences in sites of recurrence (including peritoneal dissemination) between laparoscopic and open procedures. We pay careful attention to avoiding manipulation of tissues around the tumor during laparoscopic surgery, to prevent microscopic dissemination. A recent study from a large center also demonstrated the oncological safety of laparoscopic surgery in cT4 colon cancer [30, 31].

Strengths of the present study include the relatively large cohort of patients who underwent a standardized resection procedure. Other strengths have included the availability of granular clinical and demographic information. However, the study was subject to the selection bias inherent in observational retrospective studies. In the present study, difficult procedures or advanced disease were more likely treated with open surgery particularly in the early period. Although we tried to minimize such selection bias by multivariate analyses, the results need careful interpretation as we cannot eliminate the effects of confounding factors. Data from a specialty institution in Japan, where patients are generally fit and non-obese, have the potential for limited applicability. Survival outcomes in this study resembled those of a previous Japanese RCT (5-year OS, 91.8%) [8]. Further external validation with contemporary data is thus needed. Improved outcomes in the late period could have resulted from a shorter follow-up compared to an early period. However, the median follow-up of 54 months in the late period would still cover the majority of recurrences. Despite such limitations, our findings support the oncological safety of a laparoscopic approach for D3 dissection in pT3/4a colon cancer. Further studies are needed to reveal the benefits of a minimally invasive approach in this procedure compared to open surgery.

In summary, a cohort of patients with pT3/4a M0 colon cancer who underwent CME with D3 dissection at a single cancer center demonstrated improved oncological outcomes with an evident increase in the laparoscopic approach during the 10-year period from 2004 to 2013. Improvements in outcome are encouraging and likely reflect advances in surgical techniques along with the spectrum of care, including staging and adjuvant chemotherapy. Laparoscopic modified CME appears oncologically safe and feasible under a bundle of modern improvements in cancer care, and future prospective studies investigating the true benefits of this specific procedure are warranted.