Ung thư đại trực tràng là căn bệnh phổ biến đứng hàng thứ 4 nguyên nhân dẫn đến tử vong của bệnh nhân. Kể từ ca nội soi đầu tiên phẫu thuật đại trực tràng được phẫu thuật thành công trong năm 1991, phẫu thuật nội soi hiện nay được thực hiện rộng rãi để điều trị ung thư đại trực tràng, và kỹ năng ngày càng trở nên hoàn chỉnh hơn.
Trang 1ORIGINAL ARTICLE
The short- and long-term outcomes of laparoscopic versus
open surgery for colorectal cancer: a meta-analysis
Chun-Li Wang&Gang Qu&Hong-Wei Xu
Accepted: 31 December 2013
# Springer-Verlag Berlin Heidelberg 2014
Abstract
Purpose The aim of the study was to compare short- and
long-term outcomes of laparoscopic surgery and conventional
open surgery for colorectal cancer
Methods Published randomized controlled trial (RCT) reports
of laparoscopic surgery and open surgery for colorectal cancer
were searched, and short- and long-term factors were
extract-ed to perform meta-analysis
Results A total of 15 RCT reports (6,557 colorectal cancer
patients) were included in this study Blood loss of
laparo-scopic surgery was less by 91.06 ml than open surgery
(p= 0.044) Operation time was longer by 49.34 min
(p= 0.000) The length of hospital stay was shorter by
2.64 days (p= 0.003) Incisional length was shorter by
9.23 cm (p=0.000) Fluid intake was shorter by 0.70 day
(p= 0.001) Bowel movement was earlier by 0.95 day
(p=0.000) Incidence of complications, blood transfusion,
and 30 days death were significantly lower in laparoscopic
surgery than in open surgery (p=0.011, 0.000, 0.01) But there
was no significant difference in lymph nodes (p=0.535) and
anastomotic leak (p=0.924) There was also no significant
differ-ence in 3 and 5 years overall survival (p=0.298, 0.966),
disease-free survival (p=0.487, 0.356), local recurrence (p=0.270, 0.649),
and no difference in 5 years distant recurrence (p=0.838)
Conclusions Laparoscopic surgery is a mini-injured approach
which can cure colorectal cancer safely and radically, and it is
not different from conventional open surgery in long-term
effectiveness, so laparoscopic surgery can be tried to widely
use in colorectal cancer
Keywords Colorectal cancer Laparoscopy surgery Open surgery Meta-analysis
Introduction
Colorectal cancer is a common disease which is the fourth reason resulted to patients’ death [1] Since the first laparo-scopic 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 colorec-tal 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 be-tween 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:
Chun-Li Wang and Gang Qu contributed equally to this work.
C <L Wang:G Qu:H <W Xu (*)
The First Affiliated Hospital of Dalian Medical University, Dalian,
Liaoning 116021, People’s Republic of China
e-mail: xuhongwei@ymail.com
C <L Wang
e-mail: wangchunli808@126.com
DOI 10.1007/s00384-013-1827-1
Trang 2“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
dichoto-mous, respectively.χ2
test was used to evaluate heterogeneity among the studies, and I2was used to quantify the
inconsis-tency (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, I2<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 Table1 This meta-analysis
includ-ed 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, and18, respectively, and the data are presented in Tables2,3,4,5,6,7,8,9,10,11,12,13,14,
15,16,17,18and19 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
Table 1 Characteristics of the randomized control trials
Trang 3Overall (I-squared = 0.0%, p = 0.980)
Study
Lacy
ID Schwandner
Liang Fujii
-0.29 (-1.19, 0.62)
0.00 (-2.03, 2.03)
WMD (95% CI)
-0.10 (-2.43, 2.23)
-0.40 (-1.53, 0.73) -1.10 (-8.38, 6.18)
100.00
%
19.92
Weight
15.02
63.52 1.54
0
Fig 1 Lymph node results of
forest plot (fixed effect model)
NOTE: Weights are from random effects analysis
Overall (I-squared = 99.1%, p = 0.000)
ID
Fujii
Lacy Liang Study
Jing Gong
Milsom
Braga
-91.06 (-179.66, -2.46)
WMD (95% CI)
-129.00 (-251.85, -6.15)
-88.00 (-132.02, -43.98)
-186.00 (-192.10, -179.90)
-32.20 (-43.99, -20.41)
0.00 (-176.97, 176.97)
-81.00 (-130.98, -31.02)
100.00
Weight
14.06
18.40
19.26
%
19.21
10.90
18.16
0
Fig 2 Blood loss result of forest
plot
NOTE: Weights are from random effects analysis
Overall (I-squared = 83.6%, p = 0.000) Liang
Milsom Study
Fujii
ID
Schwandner Lacy
49.34 (29.57, 69.12) 40.40 (31.24, 49.56) 75.00 (57.77, 92.23)
45.00 (4.03, 85.97)
WMD (95% CI)
72.00 (36.07, 107.93) 24.00 (11.13, 36.87)
100.00 25.85
22.66
%
12.58
Weight
14.38 24.53
Fig 3 Operation time result of
forest plot
Trang 4NOTE: Weights are from random effects analysis
Overall (I-squared = 82.1%, p = 0.000)
Lacy Study
Fujii
ID
Color
Color
Liang Schwandner
-2.64 (-4.41, -0.87)
-2.70 (-4.50, -0.90)
-4.70 (-9.48, 0.08)
WMD (95% CI)
-1.10 (-1.87, -0.33)
-0.20 (-1.62, 1.22)
-5.00 (-6.58, -3.42) -6.60 (-13.48, 0.28)
100.00
19.84
%
8.86
Weight
23.80
21.49
20.79 5.24
0
Fig 4 Length of hospital stay
result of forest plot
NOTE: Weights are from random effects analysis Overall (I-squared = 99.2%, p = 0.000)
Study
Liang
LAPKON Braga Milsom ID
-9.23 (-13.77, -4.68)
-7.40 (-7.99, -6.81)
0.40 (-1.47, 2.27) -12.20 (-12.66, -11.74) -17.50 (-18.89, -16.11) WMD (95% CI)
100.00
%
25.35
24.41 25.39 24.86 Weight
0
Fig 5 Incisional length result of
forest plot
Overall (I-squared = 0.0%, p = 0.767)
Color
ID Study
Color Schwandner
-0.95 (-1.18, -0.73) -1.00 (-1.27, -0.73) WMD (95% CI)
-0.80 (-1.27, -0.33) -1.00 (-1.79, -0.21)
100.00 69.11
Weight
%
22.67 8.22
Fig 6 Bowel movement result
of forest plot (fixed effect model)
Trang 595 % 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
−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
significant-ly lower than for those in the OCS group in this anasignificant-lysis 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 differ-ences 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 signifi-cantly 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
NOTE: Weights are from random effects analysis
Overall (I-squared = 63.5%, p = 0.042)
Study
Lacy
color color Schwander ID
-0.70 (-1.11, -0.29)
-1.20 (-1.83, -0.57)
-0.20 (-0.70, 0.30) -0.90 (-1.21, -0.59) -0.50 (-1.14, 0.14) WMD (95% CI)
100.00
%
20.92
25.50 33.09 20.49 Weight
0
Fig 7 Fluid intake result of
forest plot
Overall (I-squared = 41.2%, p = 0.067)
ID
Liang
JingGong
Lacy
Leung
color
color
LAPKON
CLASICC
SChwandner
Braga
Study
Milsom
cost
0.86 (0.77, 0.97)
ratio (95% CI)
0.68 (0.41, 1.15)
0.61 (0.18, 2.05)
0.37 (0.21, 0.64) 0.88 (0.60, 1.28)
0.82 (0.59, 1.14)
1.00 (0.79, 1.27)
1.03 (0.75, 1.42) 0.67 (0.43, 1.05) 1.00 (0.48, 2.07)
0.61 (0.37, 1.00)
odds
0.98 (0.40, 2.43)
1.06 (0.82, 1.39)
100.00
Weight
5.51
1.19
7.10 8.58
12.40
20.93
10.75 7.65 1.89
6.25
%
1.53
16.22
1
Fig 8 Complication result of forest plot
Overall (I-squared = 72.7%, p = 0.026)
ID LAPKON Braga Study
Fujii
0.46 (0.32, 0.65)
ratio (95% CI) 0.66 (0.42, 1.03) 0.55 (0.27, 1.10) odds
0.02 (0.00, 0.38)
100.00
Weight 48.51 23.48
%
28.01
Fig 9 Blood transfusion result
of forest plot
Trang 6Overall (I-squared = 0.0%, p = 0.996)
color cost Milsom
color
Lacy
Braga Schwandner
Study ID
0.58 (0.38, 0.88)
0.59 (0.22, 1.63) 0.49 (0.09, 2.67) 0.98 (0.06, 15.30)
0.66 (0.23, 1.89)
0.32 (0.03, 3.07)
0.59 (0.34, 1.05) 0.33 (0.01, 7.89)
odds ratio (95% CI)
100.00
18.38 7.38 1.85
14.68
5.56
49.40 2.74
% Weight
1
Fig 10 Thirty-day death result
of forest plot
Overall (I-squared = 0.0%, p = 0.543)
Leung
Jing Gong
LAPKON
Fujii
color
Braga Park
ID Study
0.99 (0.72, 1.34)
0.25 (0.03, 2.18)
1.06 (0.07, 16.60)
1.42 (0.47, 4.28)
0.59 (0.31, 1.14)
1.15 (0.73, 1.80)
0.80 (0.22, 2.91) 1.17 (0.51, 2.71)
ratio (95% CI) odds
100.00
5.54
1.33
7.28
20.17
45.93
6.87 12.88
Weight
%
1
Fig 11 Anastomotic leak result
of forest plot
Overall (I-squared = 0.0%, p = 0.856) Braga
Study ID
CLASICC
Schwandner Leung
1.03 (0.97, 1.10) 1.09 (0.93, 1.28)
odds ratio (95% CI)
1.02 (0.92, 1.14)
1.00 (0.88, 1.13) 1.02 (0.93, 1.12)
100.00 17.01
% Weight
45.84
5.80 31.35
1
Fig 12 Three-year overall
survival result of forest plot
Trang 7Overall (I-squared = 3.0%, p = 0.356)
ID Study
CLASICC
Schwandner Park
1.30 (0.82, 2.07)
ratio (95% CI) odds
1.10 (0.65, 1.86)
3.00 (0.13, 71.00) 2.64 (0.82, 8.53)
100.00
Weight
%
87.41
1.74 10.85
1
Fig 14 Three-year local
recurrence result of forest plot
Overall (I-squared = 76.5%, p = 0.014)
Study
CLASICC ID
Park cost
1.03 (0.95, 1.10)
odds
0.98 (0.89, 1.09) ratio (95% CI)
0.94 (0.86, 1.03) 1.29 (1.04, 1.59)
100.00
%
44.77 Weight
36.05 19.18
1
Fig 13 Three-year disease-free
survival result of forest plot
Overall (I-squared = 0.0%, p = 0.595) Fujii
cost
Braga
ID Study
Leung
CLASICC
1.00 (0.95, 1.05) 1.02 (0.89, 1.16)
0.96 (0.90, 1.03)
1.09 (0.93, 1.28)
ratio (95% CI) odds
1.04 (0.93, 1.17)
1.00 (0.88, 1.13)
100.00 3.41
43.22
10.63
Weight
%
17.77
24.97
1
Fig 15 Five-year overall
survival result of forest plot
Trang 8Long-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
signif-icant 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
detect-ed 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,
Overall (I-squared = 0.0%, p = 0.649)
Study
CLASICC
Fujii Braga Leung ID
0.97 (0.90, 1.04)
odds
0.94 (0.83, 1.07)
1.06 (0.91, 1.24) 1.00 (0.83, 1.20) 0.96 (0.86, 1.07) ratio (95% CI)
100.00
%
43.58
5.68 17.60 33.14 Weight
1
Fig 16 Five-year disease-free
survival result of forest plot
Overall (I-squared = 53.1%, p = 0.119) Leung
Study
CLASICC
Lacy ID
1.09 (0.76, 1.57) 1.55 (0.61, 3.91) odds
1.26 (0.80, 2.00)
0.49 (0.20, 1.16) ratio (95% CI)
100.00 13.64
%
58.92
27.44 Weight
1
Fig 17 Five-year local
recurrence result of forest plot
Overall (I-squared = 0.0%, p = 0.754)
ID
CLASICC
Study
Lacy Leung
1.03 (0.81, 1.30)
ratio (95% CI)
1.02 (0.76, 1.36)
odds
0.76 (0.29, 1.96) 1.14 (0.70, 1.85)
100.00
Weight
67.36
%
8.43 24.21
Fig 18 Five-year distant
recurrence result of forest plot
Trang 9Table 2 Number of lymph nodes compared LCS and OCS ( p=0.535)
among four studies (n)
studies (in milliliters)
Table 4 Operation time compared LCS and OCS (p=0.000) among five
studies (in minutes)
Table 5 Hospital length of stay compared LCS and OCS ( p=0.003)
among six studies (in days)
Table 6 Incisional length compared LCS and OCS ( p=0.000) among four studies (in centimeters)
three studies (in days)
Table 8 Fluid intake compared LCS and OCS (p=0.001) during four studies (in days)
Table 9 Number of complication compared LCS and OCS (p=0.011) among 12 studies (n)
Trang 10Table 10 Number of blood transfusion compared LCS and OCS
(p=0.000) among three studies (n)
Table 11 Number of 30-day death compared LCS and OCS ( p=0.011)
among seven studies (n)
Table 12 Number of anastomotic leak compared LCS and OCS
(p=0.924) among seven studies (n)
Table 13 Three-year overall survival compared LCS and OCS
(p= 0.298) among five studies (n)
Table 14 Three-year disease-free survival compared LCS and OCS (p=0.487) among five studies (n)
Table 15 Three-year local recurrence compared LCS and OCS (p=0.270) among five studies (n)
Table 16 Five-year overall survival compared LCS and OCS (p=0.966) among five studies (n)
Table 17 Five-year disease-free survival compared LCS and OCS (p=0.356) among five studies (n)