Introduction

Colorectal cancer is a common disease which is the fourth reason resulted to patients’ death [1]. Since the first laparoscopic colorectal surgery was operated successfully in the year 1991 [2], laparoscopic surgery is widely performed in the colorectal cancer, and the skill is becoming more and more mature. Its’ security, feasibility, and short-term curative effect have already been verified [3, 4]. Some randomized controlled trials (RCTs) have gotten the result that laparoscopic colorectal surgery (LCS) had the better short-term outcomes than open colorectal surgery (OCS), for example, less blood loss, better quality life, less pain, the shorter time of return to normal life and shorter length of hospital stay, and so on [5, 6]. But the post-operation recurrence is the most important problem which we should consider. And there are few reports about meta-analysis results of post-operation recurrence between laparoscopic and open surgery, while it is essential first-class evidence of evidence-based medicine, so several RCTs comparing LCS and OCS’s short- and long-term outcomes were selected to have been done meta-analysis. And the factors of 3 and 5 years following up period below were concluded to evaluate the long-term results of LCS.

Materials and methods

We looked up many materials about RCTs of colorectal cancer comparing LCS and OCS which were published from January 1991 to June 2013 and searched the major medical databases such as Pubmed, Embase, Ovid, ScienceDirect, Springer, Interscience, and so on. The search terms were used: “laparoscopy surgery,” “colorectal cancer,” “open surgery,” “randomized controlled trial,” and so on. Furthermore, we limited our search to those studies that involved a following up period of 3 or 5 years to evaluate the long-term outcomes of LCS. We conducted a meta-analysis for the short and long term. For the short-term analysis, we collected data of the operation time, blood loss, number of patients requiring blood transfusion, number of harvested lymph nodes, time of fluid intake, bowel movement, anastomotic leak, length of hospital stay, length of operation incision, complications, and 30 days death. For the long-term analysis, we used data of the rate of 3 years local recurrence, 3 years overall survival rate, 3 years disease-free survival rate, 5 years overall survival rate, 5 years disease-free survival rate, 5 years local recurrence rate, and 5 years distant recurrence.

Statistical analysis

Weighted mean difference (WMD) and odds ratio (OR) were used for the variables analysis of continuous and dichotomous, respectively. χ 2 test was used to evaluate heterogeneity among the studies, and I 2 was used to quantify the inconsistency (there were two models: fixed effect model and random effect model. The fixed effect model was used when the effects were deemed to be homogeneous (p > 0.1, I 2 < 50 %); otherwise, the random effects model was used). And Z test was used to compare the overall difference. The confidence interval (CI) was established at 95 %, and p values of less than 0.05 were considered to indicate statistical significance. Begg’s test and Egger’s test were performed in order to evaluate the publication bias (in Begg’s test p > 0.05 and in Egger’s test p > 0.05 and 95 % CI includes 1; it is thought that there was no publication bias). Statistical analyses were performed using the stata12.0 (meta module) software.

Results

At last 15 papers of RCTs that compared LCS and OCS for colorectal cancer [5–20] were selected. The characteristics of each RCT are presented in Table 1. This meta-analysis included 6,557 patients with colorectal cancer in all, of which 3,509 had performed LCS and 3,048 had OCS. The results of the short and long term are shown in Figs. 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, and 18, respectively, and the data are presented in Tables 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18 and 19.

Table 1 Characteristics of the randomized control trials
Full size table
Fig. 1
figure 1

Lymph node results of forest plot (fixed effect model)

Full size image
Fig. 2
figure 2

Blood loss result of forest plot

Full size image
Fig. 3
figure 3

Operation time result of forest plot

Full size image
Fig. 4
figure 4

Length of hospital stay result of forest plot

Full size image
Fig. 5
figure 5

Incisional length result of forest plot

Full size image
Fig. 6
figure 6

Bowel movement result of forest plot (fixed effect model)

Full size image
Fig. 7
figure 7

Fluid intake result of forest plot

Full size image
Fig. 8
figure 8

Complication result of forest plot

Full size image
Fig. 9
figure 9

Blood transfusion result of forest plot

Full size image
Fig. 10
figure 10

Thirty-day death result of forest plot

Full size image
Fig. 11
figure 11

Anastomotic leak result of forest plot

Full size image
Fig. 12
figure 12

Three-year overall survival result of forest plot

Full size image
Fig. 13
figure 13

Three-year disease-free survival result of forest plot

Full size image
Fig. 14
figure 14

Three-year local recurrence result of forest plot

Full size image
Fig. 15
figure 15

Five-year overall survival result of forest plot

Full size image
Fig. 16
figure 16

Five-year disease-free survival result of forest plot

Full size image
Fig. 17
figure 17

Five-year local recurrence result of forest plot

Full size image
Fig. 18
figure 18

Five-year distant recurrence result of forest plot

Full size image
Table 2 Number of lymph nodes compared LCS and OCS (p = 0.535) among four studies (n)
Full size table
Table 3 Blood loss compared LCS and OCS (p = 0.044) among six studies (in milliliters)
Full size table
Table 4 Operation time compared LCS and OCS (p = 0.000) among five studies (in minutes)
Full size table
Table 5 Hospital length of stay compared LCS and OCS (p = 0.003) among six studies (in days)
Full size table
Table 6 Incisional length compared LCS and OCS (p = 0.000) among four studies (in centimeters)
Full size table
Table 7 Bowel movement compared LCS and OCS (p = 0.000) among three studies (in days)
Full size table
Table 8 Fluid intake compared LCS and OCS (p = 0.001) during four studies (in days)
Full size table
Table 9 Number of complication compared LCS and OCS (p = 0.011) among 12 studies (n)
Full size table
Table 10 Number of blood transfusion compared LCS and OCS (p = 0.000) among three studies (n)
Full size table
Table 11 Number of 30-day death compared LCS and OCS (p = 0.011) among seven studies (n)
Full size table
Table 12 Number of anastomotic leak compared LCS and OCS (p = 0.924) among seven studies (n)
Full size table
Table 13 Three-year overall survival compared LCS and OCS (p = 0.298) among five studies (n)
Full size table
Table 14 Three-year disease-free survival compared LCS and OCS (p = 0.487) among five studies (n)
Full size table
Table 15 Three-year local recurrence compared LCS and OCS (p = 0.270) among five studies (n)
Full size table
Table 16 Five-year overall survival compared LCS and OCS (p = 0.966) among five studies (n)
Full size table
Table 17 Five-year disease-free survival compared LCS and OCS (p = 0.356) among five studies (n)
Full size table
Table 18 Five-year local recurrence compared LCS and OCS (p = 0.649) among five studies (n)
Full size table
Table 19 Five-year distant recurrence compared LCS and OCS (p = 0.838) among five studies (n)
Full size table

Short-term outcomes

The blood loss for LCS was significantly less than for OCS, by an average volume of 91.06 ml (WMD = −91.06; 95 % CI = −179.66 to −2.46; p = 0.044); six of the 15 RCTs included data of blood loss. Operation time for LCS was significantly longer than for OCS, by 49.34 min (WMD = 49.34; 95 % CI = 29.57 to −69.12; p = 0.000); five of the 15 RCTs included data of operation time. The length of hospital stay for LCS was significantly shorter than for OCS, by 2.64 days (WMD = −2.64; 95 % CI = −4.41 to −0.87; p = 0.003); six of the 15 RCTs included data of the length of hospital stay. The incisional length for LCS was significantly shorter than for OCS, by an average of 9.23 cm (WMD = −9.23; 95 % CI = −13.77 to −4.68; p = 0.000); four of the 15 RCTs included data of incisional length. The bowel movement time for LCS was significantly shorter than for OCS, by an average of 0.95 day (WMD = −0.95; 95 % CI = −1.18 to −0.73; p = 0.000); three of the 15 RCTs included data of bowel movement. The fluid intake for LCS was significantly shorter than for OCS, by 0.70 day (WMD = −0.70; 95 % CI = −1.11 to −0.29; p = 0.001); four of the 15 RCTs included data of the fluid intake. There were no significant differences in lymph nodes between the LCS group and the OCS group for treatment of the colorectal cancer. The rate of perioperative complications for patients in the LCS group was significantly lower than for those in the OCS group in this analysis of the pooled data for colorectal cancer treatment (OR = 0.86; 95 % CI = 0.77–0.97; p = 0.011). Twelve of the 15 RCTs included data of perioperative complications. The number of blood transfusion in the LCS group was significantly lower than that in the OCS group in this analysis of the pooled data for colorectal cancer treatment (OR = 0.46; 95 % CI = 0.32–0.65; p = 0.000). Three of the 15 RCTs included data of blood transfusion. There were no significant differences in anastomotic leak between the LCS group and the OCS group for the treatment of the colorectal cancer.

The rate of 30 days death in the LCS group was significantly lower than in the OCS group in this analysis of the pooled data for colorectal cancer treatment (OR = 0.58; 95 % CI = 0.38–0.88; p = 0.01). Seven of the 15 RCTs included data of 30 days death.

Long-term outcomes

We found no significant differences in the rate of 3 years local recurrence between the surgery groups when we pooled data for the treatment of the colorectal cancer. Our analysis of the 5 years of local and distant recurrence between the LCS group and the OCS group for the treatment of the colorectal cancer indicated no significant difference. There were also no significant differences between the surgery groups for the overall survival in the 3 and 5 years. We also found no significant differences in the 3- and 5-year disease-free survival rates between patients who underwent LCS and OCS.

Heterogeneity

In the short-term period, significant heterogeneity was detected among studies with respect to the following six factors: blood loss, the length of hospital stay, operation time, time of fluid intake, the rate of perioperative complications, and the number of blood transfusion. In the long-term period, significant heterogeneity was detected among studies with respect to the following factors: 3 years disease-free survival, 3 years local recurrence, 5 years overall survival, 5 years disease-free survival, and 5 years distant recurrence. Random effect model was used in the above given factors. Fixed effect model was used in the rest factors. Begg’s test and Egger’s test were performed, respectively. And all factors below passed the tests (p > 0.05 and 95 % CI includes 1).

Discussion

The biggest advantage of LCS than OCS lies to its minor injury. And many studies concluded that LCS had lower complications, less pain, shorter hospital stay, and less time to return to normal life than ORS in short-term period [9, 21, 22]. But the recurrence is the focus of debate laparoscopic approach and conventional open approach for the treatment of colorectal cancer. Therefore, we examined the results of LCS and compared to those of OCS in short- and long-term periods by a meta-analysis of 15 RCTs.

From the data meta-analysis, it is indicated that in short-term period, LCS has less blood loss, lower length of hospital stay, lower incisional length, less time bowel movement, lower rate of perioperative complication, lower number of blood transfusion, and lower number of 30 days death than OCS. It fits to the LCS’s consistent advantage. LCS is prior obviously to OCS in post-operation recovery. The length of hospital stay and time of bowel movement can be shown. But the operation time of LCS is longer than OCS because laparoscopic approach is more difficult than conventional open approach. And with surgeons’ richer and richer experience, the operation time will decrease. LCS is similar to OCS with no significant differences in lymph nodes and anastomotic leak, while the number of lymph nodes is one of the most important factors of prognosis of colorectal cancer patients.

Long-term effectiveness is the basic criterion to evaluate the tumor radical operation. Long-term survival and recurrence are acknowledged standard criterion to detect if it is radical surgery-based disease free. There are no significant differences between LCS and OCS in 3 and 5 years overall survival and disease-free survival. There are also no significant differences in 3 and 5 years local and distant recurrence between two groups. So it can be concluded that there are similar long-term effectiveness between LCS and OCS.

It is considered that the hospital charges of LCS are higher than those of OCS [23, 24]. The use of disposable surgical instruments, the high cost of intraoperative anesthesia, and the higher technical operation requirements made the charges of LCS higher than those of OCS. But YS Choi et al. ever separated charge from cost. Cost encompassed anesthesia, laboratory, radiology, pharmacy, nursing, medical therapy, and consumables charges, so total hospital charges should be evaluated by cost-effectiveness analysis. JS Park et al. also provided that total hospital charges for laparoscopic surgery were higher than those of open surgery only during the early learning period and became similar during the experienced period. So it is hoped the emergence of reusable materials which can reduce the costs and the shortening of the learning period to achieve cost-effective. It is also expected to increase the intensity of insurance of consumables.

In conclusion, this meta-analysis shows that LCS has the advantage of less blood loss, lower length of hospital stay, earlier bowel movement, and lower rate of complications than OCS in the short-term period. And LCS is similar to OCS with no significant differences in the long-term results. LCS can safely cure colorectal cancer; anyway, this article also has shortcomings, due to the lack of relevant data reported on the application condition of LCS. We at least put forward a bold attempt; at the same time, it is also hoped that more scholars and researchers can come together to explore and apply laparoscopic surgery routinely to the treatment of colorectal cancer.