Laparoscopy-assisted colectomy as an Oncologically safe alternative for patients with stage T4 Colon Cancer: a propensity-matched cohort study

A retrospective study of a maintained colon cancer databases was performed. This database collected information of patients who were diagnosed as colon cancer and treated at Department of General Surgery of Nanfang Hospital of Southern Medical University [20]. All analyzed patients were staged according to the seventh edition staging system proposed by the American Joint Committee on Cancer (AJCC).

Between March 2004 and September 2014, a total of 447 patients with clinical T4 colon cancer who underwent colectomy in our center were collected from the above database [20]. Among 447 patients, 18 patients were excluded, including 7 patients with pathology-proven T3 tumor and 11 patients with distant metastasis. Finally, 429 eligible patients with pT4 colon cancer who underwent either LAC (n = 303) or OC (n = 126) with curative intent was analyzed (Fig. 1).

The decision to proceed laparoscopic or open surgery was made depending on an understanding of the risks and benefits adherent to the laparoscopic and open procedure, without any pressure from the surgeons. All patients underwent a standardized colectomy regimen which included segmental resection of the primary tumor, radical lymphadenectomy and combined resection of invaded adjacent organs if necessary. The bowel resection contained left colectomy, extended left colectomy, transverse colectomy, right colectomy, extended right colectomy, sigmoid colectomy and total colectomy, which was selected adequately depending on primary tumor situation.

All patients analyzed were suggested to receive adjuvant chemotherapy. Personalized adjuvant chemotherapy regimens were prescribed for patients, mainly using capecitabine, oxaliplatin, leucovorin, and irinotecan, according to the updating editions of colon cancer treatment guideline from the National Comprehensive Cancer Network (NCCN) guideline [21].

All patients were followed up every 3 months for the first 2 years, 6 months for the next 3 years and annually till the patient’s death. The physical examination, serum tumor marks, chest x-ray, abdominal ultrasonography, gastroendoscopy and (positron emission) computed tomography were chosen adequately for each time of assessments. The last follow-up was in September 2016.

Patients were divided into two groups based on the surgical approach, namely the LAC and OC groups. Conversion to open surgery was defined as an abdominal incision longer than 7 cm or an abdominal incision made different from the previously planned procedure. To minimize the confounding bias, propensity score matching (PSM) was performed [18, 19]. Four covariates (i.e., age, previous abdominal surgery history, comorbidity and primary tumor size) with unbalanced distribution between two groups were used for PSM. The score was estimated using a logistic regression model and greedy matching (ratio = 1:1 without replacement) with a caliper of width 0.2 standard deviations of the logit of the estimated propensity score. Demographic, clinicopathologic, surgical finding and recovery course, and long-term oncologic outcomes were compared between the LAC and OC groups. Long-term outcomes included 5-year overall survival (OS) and 5-year disease free survival (DFS).

Data are presented as mean (±SD) for continuous variables and as frequency (%) for categorical variables. χ² tests (or Fisher’s exact tests, if applicable) were used for comparing categorical data and Student’s t tests were applied for continuous variables. Survival curves were estimated by using Kaplan-Meier method and were then compared by log-rank test. A p value less than 0.05 (two-tailed) was considered as statistically significant. All statistical analyses were performed with SPSS for Windows, version 20.0 (SPSS, Chicago, IL, USA).

Four hundred and twenty nine patients were finally included in this analysis, including 303 in LAC group and 126 in OC group. The details of patients’ characteristics are presented in Table 1. There were significantly unequal distributions between the two groups in aspects of age, previous abdominal surgery history, comorbidity and primary tumor size. Therefore, PSM was performed by matching these four variables for selection of patients with comparable characteristics. Propensity score distributions of two groups before and after matching are shown in Fig. 2. One hundred and twenty one patients were matched in each group, with no significant difference in clinical and pathological characteristics (Table 1).

Before matching, there were 15 (5.0%) patients occurring conversion from laparoscopic to open surgery due to severe adhesion of tumor (5 cases, 33.3%), bulky tumor (8 cases, 53.4%) and technical difficulties (2 cases, 13.3%). And in the matched cohort, among 121 patients in LAC group, 11 converted to open surgery (9.1%) with the following reasons: severe adhesion of tumor in 5 cases (45.5%), bulky tumor in 4 cases (36.4%) and technical difficulties in 2 cases (18.2%).

Surgical findings and short-term outcomes for the propensity-matched cohort are summarized in Table 2. There was no statistically significant difference in mean operating time between the LAC and OC groups, whether for pT4a stage (LAC 148.71 min vs. OC 160.32 min, p = 0.153) or for pT4b stage (LAC 165.24 min vs. OC 180.13 min, p = 0.273). Although the proportion of patients with intraoperative blood transfusion in two groups was similar for both pT4a (LAC 5.3% vs. OC 8.1%, p = 0.533) and pT4b stages (LAC 6.5% vs. OC 8.5%, p = 0.716), patients received LAC experienced less blood loss than those received OC (for pT4a stage, LAC 91.80 ml vs. OC 135.68 ml, p = 0.002; for pT4b stage, LAC 89.24 ml vs. OC 168.09 ml, p = 0.020). Whether for pT4a or pT4b stage, R0 resection was achieved at similar rates between two groups without statistical difference (both p > 0.05). For postoperative recovery, the time to first flatus and first liquid intake in the LAC group were significantly shorter than that in the OC group (time to first flatus, LAC 3.35 days vs. OC 4.45 days, p < 0.001; time to first liquid intake, 4.28 days vs. 5.62 days, p < 0.001) for pT4b stage. However, no significant difference was examined between two groups for pT4a stage, in aspects of time to first flatus (LAC 3.53 days vs. OC 3.94 days, p = 0.074) and time to first liquid intake (LAC 4.08 days vs. OC 4.59 days, p = 0.070). Shorter hospital stay was observed in LAC group for both pT4a and pT4b stages. (both p < 0.05).

Two hundred and forty two patients were followed up with a mean duration of 43.7 months (range, 7-136 months). OS and DFS curves for pT4a and pT4b stages are shown in Fig. 3. For DFS and OS, no significant difference was examined between LAC and OC groups, whether for pT4a or pT4b stage. The 5-year DFS rate was 64.2% for pT4a stage and 35.5% for pT4b stage in LAC group, and 62.9% and 33.7% in OC group for pT4a (p = 0.374) and pT4b (p = 0.385) stage respectively. For 5-year OS rates, two groups were also similar in pT4a stage (LAC 69.2% vs. OC 66.0%, p = 0.151) and pT4b stage (LAC 37.5% vs. OC 38.1%, p = 0.510).