Laparoscopy for rectal cancer reduces short-term mortality and morbidity: results of a systematic...

http://ueg.sagepub.com/United European Gastroenterology Journal

http://ueg.sagepub.com/content/1/1/32The online version of this article can be found at:

DOI: 10.1177/2050640612473753

2013 1: 32United European Gastroenterology JournalAlberto Arezzo, Roberto Passera, Gitana Scozzari, Mauro Verra and Mario Morino

review and meta-analysisLaparoscopy for extraperitoneal rectal cancer reduces short-term morbidity: Results of a systematic

Published by:

http://www.sagepublications.com

On behalf of:

United European Gastroenterology

can be found at:United European Gastroenterology JournalAdditional services and information for

http://ueg.sagepub.com/cgi/alertsEmail Alerts:

http://ueg.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

What is This?

- Feb 1, 2013Version of Record >>

by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from by guest on October 11, 2013ueg.sagepub.comDownloaded from

http://ueg.sagepub.com/

http://ueg.sagepub.com/content/1/1/32

http://www.sagepublications.com

http://www.ueg.eu/

http://ueg.sagepub.com/cgi/alerts

http://ueg.sagepub.com/subscriptions

http://www.sagepub.com/journalsReprints.nav

http://www.sagepub.com/journalsPermissions.nav

http://ueg.sagepub.com/content/1/1/32.full.pdf

http://online.sagepub.com/site/sphelp/vorhelp.xhtml


















Meta-analysis

Laparoscopy for extraperitoneal rectal cancerreduces short-term morbidity: Results of asystematic review and meta-analysis

Alberto Arezzo1, Roberto Passera2, Gitana Scozzari1, Mauro Verra1 andMario Morino1

AbstractBackground: The role of laparoscopy in the treatment of extraperitoneal rectal cancer is still controversial. The aim of the

study was to evaluate differences in safety of laparoscopic rectal resection for extraperitoneal cancer, compared with open

surgery.

Materials and methods: A systematic review from 2000 to July 2012 was performed searching the MEDLINE and EMBASE

databases (PROSPERO registration number CRD42012002406). We included randomized and prospective controlled clinical

studies comparing laparoscopic and open resection for rectal cancer. Primary endpoints were 30-day mortality and mor-

bidity. Then a meta-analysis was conducted by a fixed-effect model, performing a sensitivity analysis by a random-effect

model. Relative risk (RR) was used as an indicator of treatment effect.

Results: Eleven studies, representing 1684 patients, met the inclusion criteria: four were randomized for a total of 814

patients. Mortality was observed in 1.2% of patients in the laparoscopic group and in 2.3% of patients in the open group,

with an RR of 0.56 (95% CI 0.19–1.64, p¼ 0.287). The overall incidence of short-term complications was lower in the

laparoscopic group (31.5%) compared to the open group (38.2%), with an RR of 0.83 (95% CI 0.73–0.94, p¼ 0.004). Surgical

complications, wound complications, blood loss and the need for blood transfusion, time for bowel movement recovery, food

intake recovery, and hospital stay were significantly lower or less frequent in the laparoscopic group. The incidence of intra-

operative injuries, anastomotic leakages, and surgical re-interventions was similar in the two groups. Only operative time

was in favour of the open group.

Conclusions: Based on the evidence of both randomized and prospective controlled series, mortality was lower after

laparoscopy although not significantly so, while the short-term morbidity RR, including subgroup analysis, was significantly

lower after laparoscopy for extraperitoneal rectal cancer compared to open surgery.

KeywordsLaparoscopy, meta-analysis, rectal cancer, rectal neoplasms, systematic review

Received: 30 October 2012; accepted: 12 December 2012

Introduction

Laparoscopy for colon cancer, first described by Jacobset al. in 1991,1 aroused interest after the publication ofthe randomized trial by Lacy et al. in 20022 andobtained acceptance after the publication of the threetrials COST, COLOR and CLASICC.3–5 With theexception of the CLASICC trial, patients with rectalcancer were not included in these studies.

Excellence of surgical technique is of particular rele-vance in the treatment of rectal cancer. Routine exci-sion of the intact mesorectum during resection ofcancers of the middle and lower rectum has resulted

in a consistent reduction of local recurrences6 and inan increase in long-term survival rates.7

While awaiting long-term oncological results, differ-ent studies have reported on postoperative and short-term results, advocating similar advantages ofminimallyinvasive treatment as for many other procedures,

1Department of Surgical Sciences, University of Turin, Turin, Italy2Division of Nuclear Medicine, University of Turin, Turin, Italy

Corresponding author:Alberto Arezzo, Department of Surgical Sciences, University of Turin, corso

Dogliotti 14, 10126 Turin, Italy.

Email: [email protected]

United European Gastroenterology Journal

1(1) 32–47

! Author(s) 2013

Reprints and permissions:

sagepub.co.uk/journalsPermissions.nav

DOI: 10.1177/2050640612473753

ueg.sagepub.com

including less pain, shorter postoperative ileus, earlierreturn to oral intake, shorter hospital stay, and lowerblood loss.8,9 As the analysis of short-term benefits oflaparoscopy should be in some way a prerequisite for theanalysis of long-term results, with the present study weintended to evaluate in ameta-analysis whether there areclinically relevant short-term advantages of either lapar-oscopy or laparotomy for surgical treatment of extra-peritoneal rectal cancer in the published literature.While the assessment of laparoscopic advantages inrectal cancer treatment has been the topic of other sys-tematic reviews and meta-analyses in the past, this is thefirst meta-analysis that specifically focuses on extraper-itoneal rectal cancer.

Materials and methods

Themethods for the analysis and generation of inclusioncriteria were based on the recommendations of thePreferred Reporting Items for Systematic reviews andMeta-Analysis (PRISMA) Statement.10 According topopulation, interventions, comparators, outcomemeasures, and setting (PICOS) criteria, patients wereincluded if affected with extraperitoneal rectal cancerfor which laparoscopic or laparotomic treatmentwas indicated. The study methods were documentedProspero, an international prospective register ofsystematic reviews (registration number CRD42012002406).

Studies

Only randomized controlled trials (RCTs) or prospect-ive controlled clinical trials (noRCTs) were consideredfor this analysis, as suggested by the MOOSE group.11

Studies were excluded if the study population includedcolon cancers or intraperitoneal rectal cancers, unlessthe data were presented separately. When multiple stu-dies from the same institution were identified, the mostrecent or the most informative was selected. All andonly full-text papers in English were considered.

Participants

This meta-analysis compares laparoscopic and laparo-tomic resection for extraperitoneal rectal cancer withregard to possible benefits of laparoscopy or laparot-omy in the short-term postoperative period, defined asup to 30 days after surgery.

Intervention

All surgical procedures involving resection of rectalcancers defined either as of the mid and low rectum,or extraperitoneal or lower than 12 cm from the anal

verge, were considered, including partial or total mesor-ectal excision with mechanical of hand-sutured anasto-mosis, Hartmann procedures, and abdominoperinealresection. Type of interventions performed were notedin order to analyse separately those involving and thosenot involving a bowel anastomosis. For the laparoscopicgroup, any rectal resection performed by means of amini-invasive approach (i.e. in a space generated by aninsufflated pneumoperitoneum with operative field visu-alization obtained by a videolaparoscope and performedonly with laparoscopic trocars) was included, while allprocedures described as ‘open’ or ‘conventional’ andperformed by means of an abdominal laparotomic inci-sion were considered as open surgery.

Outcome measures

Primary endpoints were overall mortality and morbid-ity at 30 days after surgery. Intraoperative and early(<30 days) postoperative complications directly relatedto surgery, and early (<30 days) postoperative medicalcomplications were the subject of different sensitivityanalyses. Anastomotic leakage, bleeding and bloodloss, wound infection and/or wound dehiscence, pelvicand/or abdominal abscesses, and bowel and/or vascularand/or urological injuries were classified as surgicalcomplications. Paralitic ileum and/or non-surgicalbowel obstruction, respiratory events, cardiovascularevents, deep venous thrombosis and/or pulmonaryembolism, urinary infection, urinary retention, non-surgical infections, or sepsis were classified as medicalcomplications.

The secondary outcome measures were incidence ofanastomotic leakage, abscesses, blood loss, time to firstbowel movement, time for intake recovery, need fortransfusion, length of hospital stay, wound infections,internal organ injuries, need for re-intervention, andoperative time.

Results

Study selection

The search retrieved 5017 studies. Figure 1 illustratesthe PRISMA flowchart for study inclusion and exclu-sion criteria.

Characteristics of included studies

The characteristics of the 11 studies meeting theinclusion criteria are summarized in Table 1.12–22 All11 studies were reported as full papers and includeda total of 1684 patients. Four studies were RCTs fora total of 814 patients and seven studies were noRCTsfor a total of 870 patients.

Arezzo et al. 33

Table 2 shows the patients’ baseline characteristicscomparing open and laparoscopic procedures. Table 3shows characteristics of tumour location and stage,adjuvant therapy, and percentage of protectiveileostomy.

Risk of bias of included studies

Assessment of quality according to the CochraneCollaboration’s tool for assessing risk of bias forRCTs and to the Newcastle-Ottawa Scale for prospect-ive noRCTs are represented in Tables 4 and 5,respectively.

Quality analysis

Figure 2 shows the potential sources of heterogeneitywithin all studies by a L’Abbe plot on mortality out-come. Since in most of the analyses I2 was <50%, thefixed-effects and random-effects model results can beconsidered comparable.

Primary outcomes

The meta-analyses of the two primary outcomes inves-tigated mortality and overall complications. For thefirst primary outcome, the raw incidence of mortality,as reported in five studies, was lower in the laparo-scopic group (1.2%) compared to the open group(2.3%). The overall relative risk (RR) was 0.56 (95%CI 0.19–1.64, p¼ 0.287), showing no heterogeneity

(I2¼ 0%) and no differences between RCTs vs.noRCTs subgroups (RR 0.75 vs. 0.41, p¼ 0.590;Figure 3). Performing a cumulative meta-analysis withthese five studies (two RCTs and three noRCTs),adding one study at a time by publication date, theRR progressively raised from 0.33 to 0.59; performingan influential meta-analysis, by omitting one study inturn, the RR ranged from 0.50 to 0.60 over the wholetime frame.

For the second primary outcome, the incidence ofoverall 30-day morbidity, as reported in 10 studies,was significantly lower in the laparoscopic (31.5%)compared to the open group (38.2%). The overall RRwas 0.83 (95% CI 0.73–0.94, p¼ 0.004), showing verylow heterogeneity (I2¼ 8.2%) and no statistically sig-nificant differences between RCT vs. noRCT trials (RR0.84 vs. 0.83, p¼ 0.934; Figure 4). Performing a cumu-lative meta-analysis with these 10 studies (four RCTsand six noRCTs), the RR varied from 0.96 to 0.70,being almost constant in the period 2007–2011. In theinfluential meta-analysis, the RR resulted almost stableover the whole publication period, ranging between0.81 and 0.86.

Secondary outcomes

As secondary outcomes, the meta-analysis investigatedmedical and surgical complications in detail, i.e.patients with medical and/or surgical complications,surgical duration, mean blood loss, incidence of intra-operative injuries, bowel movement recovery, food

TitlesN = 5017

Non-relatedN = 4520

Exclusion based on abstract textN = 429

Manuscripts non-eligibleN = 57

noRCT studiesN = 7

AbstractsN = 497

ManuscriptsN = 68

Manuscripts eligible for dataextraction

N = 11

RCT studiesN = 4

Figure 1. Flow-chart diagram detailing the paper selection process.

34 United European Gastroenterology Journal 1(1)

Tabl

e1.

Su

mm

ary

of

the

stu

die

sin

clu

ded

inth

em

eta-a

naly

sis

Pu

blica

tion

Cou

ntr

yan

dst

udy

per

iod

Incl

usi

on

crit

eria

Exc

lusi

on

crit

eria

a

Eligib

le

pati

ents

LAP

surg

ery

pati

ents

OP

EN

surg

ery

pati

ents

Gen

der

(M/F

)A

ge

(mea

n�

SD

)B

MI

(mea

n�

SD

)

Con

vers

ion

rate

(%)

Lap

Open

Lap

Open

Lap

Open

Pro

spec

tive

con

trolled

clin

ical

tria

ls

Leu

ng

etal.

(200

0)16

Hon

g-K

on

g(J

an

1993

–Ja

n19

96)

Low

rect

al

can

cer

2,11

5925

3415

/10

21/1

362

.2�

13.3

63.5�

15.2

ND

ND

8.0

Hu

etal.

(200

3)20

Ch

ina

(Oct

2001

–Ju

l20

02)

Rec

tal

can

cer<

15cm

from

AV

2,3,

4,5,

9,13

,14

,

16,

19,

20

4520

259/

1116

/961

.6�

8.4

58.0�

10.7

ND

ND

0

Bre

uki

nk

etal.

(200

5)21

Net

her

lan

ds

(Lap:

Oct

2000

-Mar

2003

);

(Open

:A

pr

1996

–N

ov

2001

)

Pri

mary

rect

al

can

cer

aft

er

pre

oper

ati

vera

dio

ther

apy

9,19

8241

4125

/16

23/1

868

b70

b25

b25

b9.

8

Mori

no

etal.

(200

5)12

Italy

(Apr

1994

–Apr

2002

)R

ecta

lca

nce

r�

12cm

from

AV

2,9,

12,

1919

198

9359

/39

57/3

664

.961

.4N

DN

D18

.4

Sta

udach

eret

al.

(200

7)18

Italy

(Jan

1998

–Sep

2005

)M

iddle

an

dlo

wre

ctal

can

cer

218

710

879

65/4

342

/37

63.9�

12.2

64.7�

13.0

26.3�

3.8

25.8�

4.1

12.0

Baik

etal.

(201

1)19

USA

(Sep

2001

–Sep

2005

)R

ecta

lca

nce

r�

12cm

from

AV

4,5,

6,9,

1416

254

108

37/1

762

/46

60.0�

12.7

60.6�

13.6

27.3�

4.2

28.9�

5.2

11.1

Ses

hadri

etal.

(201

2)22

India

(Jan

2004

–Ja

n20

10)

Mid

dle

an

dlo

wre

ctal

can

cer

1,2,

9,11

,

15,

20

144

7272

47/2

545

/27

48b

48b

21b

22b

4.2

Ran

dom

ized

con

trolled

clin

ical

tria

ls

Zh

ou

etal.

(200

4)12

Ch

ina

(Ju

n20

01–Sep

2002

)R

ecta

lca

nce

rw

ith

low

est

marg

inu

nder

the

per

iton

eal

refl

ecti

on

an

d1.

5cm

above

the

den

tate

lin

e

1,2,

8,9,

1117

182

8946

/36

43/4

644

.045

.0N

DN

DN

D

Ng

etal.

(200

8)17

Hon

g-K

on

g(J

ul

1994

–Fe

b20

05)

Rec

tal

can

cer�

5cm

from

AV

2,7,

9,10

,11

9951

4831

/20

30/1

863

.7�

11.8

63.5�

12.6

ND

ND

9.8

Luja

net

al.

(200

9)14

Spain

(Jan

2002

–Fe

b20

07)

Mid

an

dlo

wre

ctal

can

cer

2,9,

1520

410

110

362

/39

64/3

967

.8�

12.9

66.0�

9.9

ND

ND

7.9

Kan

get

al.

(201

0)15

Sou

thKore

a(A

pr

2006

–Au

g20

09)

Rec

tal

can

cer�

9cm

from

AV

2,4,

5,9,

14,

1634

017

017

011

0/60

110/

6057

.8�

11.1

59.1�

9.9

24.1�

3.2

24.1�

3.2

1.2

a1,

Neo

pla

smoth

erth

an

ad

enoca

rcin

om

a(e

.g.

lym

ph

om

a);

2,em

ergen

cysi

tuati

on

s(e

.g.

acu

teob

stru

ctio

n,

haem

orr

hag

e,p

erfo

rati

on

);3,

con

train

dic

ati

on

sto

pn

eum

op

erit

on

eum

;4,

mali

gn

an

td

isea

ses

inth

e

past

5ye

ars

or

syn

chro

nou

sad

enoca

rcin

om

a;

5,p

reg

nan

cy;

6,ass

oci

ate

dgast

roin

test

inal

dis

ease

sn

eed

ing

surg

ical

inte

rven

tion

;7,

recu

rren

td

isea

se;8,

low

est

marg

inof

tum

ou

rw

ith

in1.

5cm

ab

ove

the

den

tate

lin

e;9,

Du

kes

stag

eD

or

T4TN

Mst

age;

10,

tum

ou

rla

rger

than

6cm

;11

,p

ati

ents

un

wil

lin

gto

take

part

inth

est

ud

y;12

,lo

cal

surg

ery

can

did

ate

s;13

,age<

18or>

80ye

ars

;14

,re

spir

ato

ryd

ysfu

nct

ion

,

card

iova

scu

lar

dys

fun

ctio

n,h

epati

cd

ysfu

nct

ion

,A

mer

ican

Soci

ety

An

este

siolo

gis

tsIV

;15

,fa

mil

ialad

enom

ato

us

poly

posi

s;16

,p

rese

nce

ofm

etast

ase

s;17

,on

goin

gin

fect

ion

s,lo

wp

lasm

an

eutr

op

hil

leve

ls;18

,B

MI

>30

kg/m

2;

19,

pre

viou

sco

lon

or

rect

al

surg

ery

an

d/o

rp

revi

ou

sn

eoad

juva

nt

chem

oth

erap

y;20

,p

revi

ou

sab

dom

inal

surg

ery.

bM

edia

nva

lue.

AV,

an

al

verg

e;La

p,

lap

aro

scop

ic;

ND

,n

od

ata

ava

ilab

le;

Op

en,

lap

aro

tom

ic.

Arezzo et al. 35

intake recovery, blood transfusions, incidence ofabscesses, incidence of wound complications, incidenceof anastomotic leakages, incidence of re-intervention,and length of hospital stay.

Surgical complications, as reported in 10 studies,were significantly less frequent in the laparoscopic vs.the open group (13.5 vs. 17.5%); the overall RR was0.78 (95% CI 0.62–0.97, p¼ 0.027; Figure 5), with no

differences between RCT vs. noRCT subgroups (RR0.65 vs. 0.86, p¼ 0.247).

Patients with medical complications, as reported innine studies, were 18.5% in the laparoscopic group and22.0% in the open surgery arm; the overall RR was0.85 (95% CI 0.70–1.04, p¼ 0.114; Figure 6), againwith no differences between RCTs and noRCTs (RR0.95 vs. 0.73, p¼ 0.201).

Table 4. Quality assessment of the randomized controlled studies included, based on the Cochrane Collaboration’s tool for assessing risk

of bias

Publication

Random

sequence

generation

Allocation

concealment

Blinding of

participants,

personnel,

and outcome

Incomplete

outcome

data

Selective

outcome

reporting

Other source

of bias

Zhou et al. (2004)12 No No Unclear Unclear Noa Yes

Ng et al. (2008)17 Yesb Yesc Unclear Yes Yes Yes

Lujan et al. (2009)14 Yesb Yesc Unclear Yes Yes Yes

Kang et al. (2010)15 Yesb Yesc Yesd Yes Yes Yes

In all cases, ‘Yes’ indicates a low risk of bias, ‘No’ indicates high risk of bias, and ‘Unclear’ indicates unclear or unknown risk of bias.aIn Zhou (2004), no clear follow-up data are provided and no comparison between groups with regard to follow-up length is available.bIn Ng (2008), Lujan (2009), and Kang (2010), randomization sequence was generated by a computer program.cIn Kang (2010), allocation concealment was done by telephone by the trial coordinator; in Lujan (2009) by means of sealed envelopes; in Ng (2008) by an

independent operating theatre coordinator.dIn Kang (2010), pathologists who examined the resected specimen were blinded as to allocation of patients.

Table 3. Comparison of tumour location, cancer stage, neoadjuvant therapy, and protective ileostomy

No. of

patients

Mean distance

from the anal

verge (cm)

Tumour

stage T0–T2aTumour

stage T3–T4aNeoadjuvant

therapy

Protective

ileostomy (%)b

Lap Open Lap Open Lap Open Lap Open Lap Open Lap Open

noRCT 418 452 6.1 5.4 171/332 187/352 161/332 165/352 224/275 236/300 48.0 56.0

RCT 404 410 5.6 5.7 164/322 141/321 158/322 180/321 243/322 247/321 60.0 52.2

Overall 822 862 5.8 5.5 335/654 328/673 319/654 345/673 467/597 483/621 54.6 54.3

aNumber of T0–T2 and T3–T4 are not equal to total number since data was not available in all studies.bPercentages of protective ileostomy are calculated, when available, on the number of patients undergoing a surgical procedure involving a bowel

anastomosis.

Table 2. Comparison of baseline patient characteristics

No. of patients Gender (M/F) Mean age (years) Mean BMI (kg/m2)

Lap Open Lap Open Lap Open Lap Open

noRCT 418 452 257/161 266/186 61.1 60.4 24.9 25.9

RCT 404 410 249/155 247/163 58.2 58.3 24.1 24.1

Overall 822 862 506/316 513/349 59.7 59.4 24.6 25.2

BMI, body mass index; Lap, laparoscopic; ND, no data available; noRCT, prospective controlled clinical trial; Open, laparotomic; RCT, randomized controlled

trial.


All but one study12 provided conversion to open sur-gery rate. Overall, 61 laparoscopic cases out of 740 wereconverted to laparotomy (8.2%): 4.7% (15/322) in theRCT studies and 11.0% (46/418) in the noRCT studies.

Conversion rates ranged between 1.2%15 and 9.8%17 inthe RCT studies, and between 0%20 and 18.4%13 in thenoRCT studies. In the RCT studies, reported conver-sion rates showed a time trend, which was not apparentin the noRCTs (Table 1).

The weighted mean operative time, as reported in10 studies, was significantly longer for laparoscopicsurgery (216 and 179 minutes, respectively); the overallmean difference (MD) was 36.4 minutes (95% CI25.1–47.6, p< 0.001; Figure 7), with no notable MDsbetween RCTs and noRCTs (32.4 vs. 39.4 minutes,p¼ 0.557), but with considerable heterogeneity(I2 83.4%).

The weighted mean blood loss, as reported in eightstudies, was almost halved in the laparoscopic group(226 vs. 412ml); the overall MD was �138ml (95%CI �195 to �81, p< 0.001; Figure 8), with borderlinesubgroup differences (MD �83 vs. �181, p¼ 0.056) andconsistent heterogeneity (I2 83.4%).

The overall incidence of intra-operative injuries, asreported in six studies, was 1.4% in both laparoscopicand open surgery patients; the overall RR was 1.10(95% CI 0.41–2.91, p¼ 0.851; Figure 9), with no differ-ences between RCTs and noRCTs (RR 1.65 vs. 0.74p¼ 0.428).

The weighted mean time for bowel movement recov-ery, as reported in six studies, was significantly shorterin the laparoscopic group (3.5 vs. 4.7 days); the overallMD was �1.3 days (95% CI �1.8 to �0.9, p< 0.001;Figure 10), with no significant MD between RCTs andnoRCTs (MD �1.2 vs. �1.3, p¼ 0.827) but in the pres-ence of extreme heterogeneity (I2 83.9%).

Table 5. Quality assessment of the included non-randomized controlled studies based on the Newcastle-Ottawa scale

Selections Comparability Outcome assessment

Score1 2 3 4 5 6 7

Leung et al. (2000)16 * – – * ** – * 5

Hu et al. (2003)20 * * * * ** * * 8

Breukink et al. (2005)21 * * * * ** * * 8

Morino et al. (2005)13 * * * ** * * * 8

Staudacher et al. (2007)18 * * * ** ** * * 9

Baik et al. (2011)19 * * * ** ** * * 9

Seshadri et al. (2012)22 * * * ** ** * – 8

Selections: 1, Assignment for treatment (if yes, one point); 2, How representative was the laparoscopic group in comparison to the general population

undergoing rectal resections (if yes, one point; no points if the patients were selected or selection of group was not described); 3, How representative was

the open group in comparison to the general population undergoing rectal resections (if yes, one point; no points if the patients were selected or selection

of group was not described).

Comparability: 4, Group comparable for age, gender, and American Society Anestesiologists (if yes, two points; one point if one of these three charac-

teristics was not reported even if there were no other differences between the two groups and other characteristics had been controlled for; no points were

assigned if the two groups differed); 5, Group comparable for neoadjuvant/adjuvant therapy, tumour location, stage, and procedure (if yes, two points; one

point if one of these four characteristics was not reported even if there were no other differences between the two groups and other characteristics had

been controlled for; no points were assigned if the two groups differed).

Outcome assessment: 6, Clearly defined outcome of interest (if yes, one point for information ascertained by medical records or interview; no points if this

information was not reported); 7, Follow-up equal between the two groups (if yes, one point; no points if follow-up not reported).

0.8

0.6

0.4

Eve

nt r

ate

(Lap

aros

copy

)

0.2

0.0

0.0 0.2 0.4 0.6 0.8

Event rate (Open)

Figure 2. L’Abbe plot for all trials on mortality outcome to inves-

tigate potential sources of heterogeneity. Event rate plotted on

vertical axis in for laparoscopy group and on horizontal axis for

open group; size of circle proportional to number of patients

enrolled; solid line represents the overall relative risk line, indi-

cating estimation of relative risk by pooling results of all studies.

Arezzo et al. 37

The weighted mean food intake recovery, asreported in nine studies, occurred after 4.0 days in thelaparoscopic group and 4.8 days in the open surgerygroup, significantly favouring the former; the overall

MD was �0.9 days (95% CI �1.4 to �0.4, p< 0.001;Figure 11), showing a significant MD between RCTsand noRCTs (MD �0.3 vs. �1.0, p¼ 0.031) andextreme heterogeneity (I2 75.9%).

Study

group = noRCTLeung 2000Hu 2003Breukink 2005Morino 2005Staudacher 2007Baik 2011

Zhou 2004Ng 2008Lujan 2009Kang 2010

5324341

8251

101170

11404746

8948

103170

group = RCT

Fixed effect modelRandom effects model



346 383 0.830.84

[0.69; 0.99][0.69; 1.03]

50.0%--

0.1 0.5 1 2 10

0.960.250.581.110.850.81

0.490.750.960.89

220

18213717

Heterogeneity: I-squared=34.4%, tau-squared=0.0209, p=0.1782

Heterogeneity: I-squared=0%, tau-squared=0, p=0.5056


25204098

10854

342

31183242

3725419379

108

[0.81; 1.14][0.01; 4.90][0.39; 0.86][0.63; 1.94][0.58; 1.23][0.51; 1.28]

[0.18; 1.36][0.59; 0.96][0.68; 1.27][0.62; 1.28]

9.5%0.8%

10.7%6.4%

12.8%9.7%

3.7%14.3%16.1%16.0%

1.3%18.3%12.3%9.2%

31.9%0.1%8.2%4.0%8.7%5.9%

--58.9%

0.840.84

[0.70; 1.00][0.69; 0.97]

50.0%--

--41.1%

0.830.85

[0.73; 0.94][0.76; 0.95]

100%--

--100%

404 410

750 793

LaparoscopyEvents Total Events

OpenTotal

Overall complicationsRR 95%-CI W(fixed) W(random)

Figure 4. Forest plot for overall 30-day morbidity.

CI, confidence interval; RR, relative risk; W, weight of single study.

Study

group = noRCTBreukink 2005Morino 2005Baik 2011

Ng 2008Lujan 2009

12

51101

13

48103

group = RCT




193 242 0.410.41

[0.08; 2.05][0.08; 2.06]

56.7%46.0%

0.750.75

0.1 0.5 1 2 10

0.940.68

0.330.470.40

010




419854

122

4193

108

[0.01; 7.95][0.04; 5.15][0.02; 8.15]

16.3%22.2%18.2%

11.9%21.0%13.1%

[0.06; 14.63][0.12; 3.98]

11.2%32.2%

15.9%38.2%

[0.17; 3.29][0.17; 3.31]

43.3%54.0%

0.560.57

[0.19; 1.64][0.19; 1.70]

100%--

----

--

----

100%

152 151

345 393


OpenTotal

MortalityRR 95%-CI W(fixed) W(random)

Figure 3. Forest plot for 30-day mortality.



Study

group = noRCTLeung 2000Hu 2003Breukink 2005Morino 2005Staudacher 2007Baik 2011


111186

8251

101170

3142216

8948

103170

group = RCT




346 380 0.860.88

[0.65; 1.14][0.66; 1.18]

58.5%--

0.1 0.5 1 2 10

1.040.250.471.370.771.05

0.490.750.960.89

1307

132210




25204198

10854

172

159

2119

3425419379

108

[0.63; 1.72][0.01; 4.90][0.21; 1.02][0.62; 3.05][0.45; 1.29][0.53; 2.10]

[0.04; 3.41][0.37; 1.47][0.48; 1.46][0.15; 0.94]

10.8%1.7%

11.3%7.0%

18.3%9.5%

2.2%10.9%16.4%12.0%

1.0%10.9%16.2%6.1%

19.8%0.6%8.3%8.0%

18.5%10.6%

--65.7%

0.840.84

[0.44; 0.96][0.46; 1.00]

41.5%--

--34.3%

0.780.81

[0.62; 0.97][0.64; 1.01]

100%--

--100%

404 410

750 790


OpenTotal

Surgical complications

RR 95%-CI W(fixed) W(random)

Figure 5. Forest plot for 30-day surgical complications.


Study

group = noRCTLeung 2000Breukink 2005Morino 2005Staudacher 2007Baik 2011


4212535

8251

101170

8262530

8948

103170

group = RCT




326 355 0.730.72

[0.53; 1.00][0.52; 0.98]

44.0%--

0.2 0.5 1 2 5

0.610.690.841.000.61

0.540.761.021.17

9118

157




254198

10854

20169

1123

34419379

108

[0.34; 1.11][0.36; 1.30][0.34; 2.09][0.48; 2.05][0.28; 1.33]

[0.17; 1.73][0.50; 1.16][0.63; 1.65][0.75; 1.81]

10.6%10.0%5.8%8.0%9.6%

4.8%16.8%15.5%18.8%

2.8%21.9%16.6%19.9%

10.9%9.6%4.6%7.4%6.3%

--38.8%

0.950.93

[0.74; 1.22][0.73; 1.20]

56.0%--

--61.2%

0.850.84

[0.70; 1.04][0.69; 1.02]

100%--

--100%

404 410

730 765


OpenTotal

Surgical complications

RR 95%-CI W(fixed) W(random)

Figure 6. Forest plot for 30-day medical complications.


Arezzo et al. 39

The percentage of laparoscopic patients requiringblood transfusions, as reported in three studies, was sig-nificantly lower than that for open surgery patients (4.5vs. 7.0%); the overall RR was 0.55 (95% CI 0.31–0.98,

p¼ 0.041; Figure 12), in the absence of subgroup differ-ences (RR 0.33 vs. 0.57, p¼ 0.749).

Abscesses, as reported in seven studies, wereobserved in 1.8% of patients in the laparoscopic

Study

group = noRCTLeung 2000Hu 2003Breukink 2005Morino 2005Staudacher 2007Seshadri 2012


8251

101170

18464575

8948

103170

106164173197

24435963

120214194245

group = RCT




35.4839.38

[28.98; 41.98][25.73; 53.04]

38.6%--

−50−100 0 50 100

50.0081.0020.0033.0033.0030.00

14.0050.0021.0048.00


Heterogeneity: I-squared=89.3%, tau-squared=330.2, p<0.0001


[27.63; 72.37][55.93; 106.07]

[-0.21; 40.21][22.92; 43.08][3.31; 62.69]

[17.72; 42.28]

[7.67; 20.33][32.47; 67.53][6.62; 35.38]

[33.28; 62.72]

3.3%2.6%4.0%

16.1%1.9%

10.8%

40.7%5.3%7.9%7.5%

12.6%10.0%10.8%10.7%

8.7%8.0%9.2%

11.9%6.9%

11.3%--

55.9%

22.1732.36

[17.02; 27.32][13.29; 51.43]

61.4%--

--44.1%

27.3036.36

[23.27; 31.34][25.08; 47.65]

100%--

--100%

410

768

404

25204198

10872

48465442

11640

342541937972

166146180165218240

363838289135

216227200198251270

344364

754


OpenTotal

Operative timeMD 95%-CI W(fixed) W(random)

Figure 7. Forest plot for mean operative time.

CI, confidence interval; MD, mean difference; W, weight of single study.

Study

group = noRCTLeung 2000Breukink 2005Staudacher 2007Baik 2011Seshadri 2012

Zhou 2004Ng 2008Lujan 2009

8251

101

19750113

8948

103

92556234

251180174

20322128

group = RCT




−172.29−181.21

[−211.55; −133.03][−277.90; −84.51]

2.7%--

--54.6%

−500−1000 0 500 1000

25.00−750.00−148.00−108.00−200.00

−72.00−234.00−106.00



Heterogeneity: I-squared=83.4%, tau-squared=3713, p<0.0001

[−339.92; 389.92][−1100.98; −399.02]

[−220.97; −75.03][−199.52; −16.48]

[−255.39; −144.61]

[−78.62; −65.38][−626.18; 158.18][−146.18; −65.82]

0.0%0.0%0.8%0.5%1.4%

2.2%2.4%

16.7%14.4%18.9%

94.7%0.0%2.6%

22.9%1.9%

20.6%−72.94−82.81

[−79.48; −66.41][−110.98; −54.64]

97.3%--

--45.4%

−75.62−138.17

[−82.07; −69.18][−195.35; −80.99]

100%--

--100%

240

534

234

2541

1085472

72575518226183

344179

10872

10001000356421400

681863292315225

1025250208313200

334300

574

LaparoscopyTotal Mean Total MeanSD

OpenSD

Blood lossMD 95%-CI W(fixed) W(random)

Figure 8. Forest plot for mean blood loss.



group and 1.9% of patients in the open group; theoverall RR was 1.08 (95% CI 0.50–2.36, p¼ 0.837;Figure 13), showing no subgroup differences (RR 1.00vs. 1.14, p¼ 0.871).

On the other hand, wound complications, asreported in 10 studies, were described for 6.9%

laparoscopic and 11.1% open patients; the overallRR was 0.63 (95% CI 0.46–0.86, p¼ 0.004;Figure 14), with no subgroup differences (RR 0.57vs. 0.66, p¼ 0.646).

Anastomotic leakages, as reported in eight studies,occurred in 6.9% of laparoscopic patients and 6.3%

Study

group = noRCTLeung 2000Breukink 2005Baik 2011

Ng 2008Lujan 2009Kang 2010

031

101

48103170

51101170

group = RCT




--53.5%

60.7%--

0.01 0.1 1 10 100




11.0%19.4%30.4%

0.317.141.00

2.720.330.28

[0.26; 28.36][0.01; 7.95][0.01; 5.41]

[0.01; 7.52][0.37; 136.45][0.06; 15.86]

39.3%--

20.0%6.4%

12.9%

24.5%13.4%15.5%

13.4%15.5%17.7%

--46.5%

100%--

1.101.05

1.651.38

0.740.83

[0.41; 2.91][0.33; 3.37]

[0.40; 6.78][0.24; 7.96]

[0.18; 2.99][0.17; 4.08]

--100%

442

200

113

3441

108

254154

120 183

504

322 321

LaparoscopyEvents EventsTotal

OpenTotal

InjuriesRR 95%-CI W(fixed) W(random)

Figure 9. Forest plot for incidence of intra-operative injuries.


StudyLaparoscopy

Total Mean SD Total Mean SDOpen Bowel movement

MD 95%-CI W(fixed) W(random)

group = noRCT

332 352Fixed effect modelRandom effects model



Morino 2005Staudacher 2007Baik 2011Seshadri 2012

981085472

3.83.93.64.0

1.20.81.80.7

9379

10872

4.75.14.76.0

1.81.12.50.8

133 137

8251

1.54.3

1.35.3

8948

2.76.3

1.52.8




group = RCTZhou 2004Ng 2008

−3 −2 −1 0 1 2 3

--75.7%

--24.3%

--100%

−1.52−1.33

[−1.68; −0.35][−1.89; −0.76]

[−1.34; −0.46][−1.49; −0.91][−1.77; −0.43][−2.25; −1.75]

[−1.62; −0.78][−3.66; −0.34][−1.66; −0.84][−1.66; −1.84]

18.7%20.8%14.9%21.3%

−0.90−1.20−1.10−2.00

−1.25−1.25

[−1.63; −1.32][−1.78; −0.90]

−1.48−1.34

-1.20-2.00

18.9%5.4%

85.7%--

14.3%--

100%--

12.4%28.9%5.2%

39.2%

134%0.9%

Figure 10. Forest plot for bowel movement recovery.


Arezzo et al. 41

of open patients. The overall RR was 1.05 (95%CI 0.68–1.61, p¼ 0.833; Figure 15), with no differencesbetween RCTs and noRCTs (RR 0.66 vs. 1.26,p¼ 0.192).

Patients who needed surgical re-intervention withinthe first 30 postoperative days, as reported in six

studies, were 5.3% in the laparoscopic group and6.9% in the open group; the overall RR was 0.75(95% CI 0.47–1.19, p¼ 0.217; Figure 16), again withno subgroup differences (RR 0.71 vs. 0.78, p¼ 0.838).

The weighted mean duration for hospital stay, asreported in 10 studies, was clearly shorter in the

Study MDReintervention

95%-CI W(fixed) W(random)Total Mean Total Mean SDSDLaparoscopy Open

group = noRCT

398 427

234 240




Leung 2000Breukink 2005Morino 2005Staudacher 2007Baik 2011Seshadri 2012

254198

1085472

4.07.04.85.94.06.0

4.84.32.02.72.71.0

34419379

10872

3.36.32.21.82.60.5

6.34.03.44.33.45.0



Heterogeneity: I-squared=75.9%, tau-squared=03235, p<0.0001

group = RCTZhou 2004Ng 2008Lujan 2009

8251

101

1.43.33.4

−4 −2 0 2 4

4.55.13.6

8948

103

1.15.84.4

4.34.32.8

72.5%67.5%

--

--

--

--−1.07−1.04

[−1.28; −0.85][−1.63; −0.45]

27.5%32.5%--

−0.28−0.28

[−0.63; −0.06][−0.63; −0.06]

100%100%--

−0.85−0.85

[−1.03; −0.67][−1.35; −0.36]

0.8%0.6%9.4%7.1%4.5%

50.1%

4.4%3.6%

15.0%14.0%12.1%18.4%

2.30−3.00−1.40−1.60−0.60−1.00

[0.23; 4.37][−5.33; −0.67][−2.00; −0.80][−2.28; −0.91][−1.46; 0.26]

[−1.26; −0.74]

23.6%1.0%2.9%

17.4%5.2%

10.0%

−0.20−0.80−0.80

[−0.58; 0.18][−2.65; 1.05][−1.88; 0.28]

Figure 11. Forest plot for food intake recovery.


Study RRTransfusions

95%-CI W(fixed) W(random)Events TotalTotal EventsLaparoscopy Open

group = noRCT


Morino 2005Staudacher 2007

314

9379

419

98108


group = RCT


Kang 2009 0 1701170

Heterogeneity: I-squared=NaN%, tau-squared=0, p=1

376 342Fixed effect modelRandom effects modelHeterogeneity: I-squared=0%, tau-squared=0, p=0.8975

94.6%96.8%

----

0.570.56

[0.32; 1.01][0.32; 1.00]

14.9%79.7%

14.9%82.0%

0.710.54

[0.16; 3.09][0.29; 1.01]

5.4%3.2%

----

0.330.33

[0.01; 8.13][0.01; 8.13]

100%100%

----

0.550.55

[0.31; 0.98][0.31; 0.98]

5.4% 3.2%0.33 [0.01; 8.13]

210.1 0.5 10

Figure 12. Forest plot for incidence of blood transfusion.



Study RRWound complications


group = noRCT


Leung 2000Hu 2003Breukink 2005Morino 2005Staudacher 2007Baik 2011Seshadri 2012

902353

11

3425419379

10872

16156

116

10

25204198

1085472


group = RCT


740 773

1062

48103170

119

11

51101170




322 321


210.1 0.5 10

62.8%67.6%

----

0.660.69

[0.45; 0.98][0.47; 1.02]

16.1%1.6%6.0%7.3%

15.1%4.8%

11.9%

25.4%1.0%4.1%5.5%9.8%5.6%

16.2%

0.760.410.400.470.331.001.10

[0.41; 1.44][0.02; 9.65][0.08; 1.95][0.12; 1.84][0.12; 0.92][0.26; 3.85][0.50; 2.43]

37.2%32.4%

----

0.570.58

[0.33; 0.98][0.26; 1.26]

100%100%

----

0.630.68

[0.46; 0.86][0.49; 0.93]

13.5%10.6%13.1%

17.6%10.2%

4.6%

0.860.680.18

[0.40; 1.83][0.25; 1.84][0.04; 0.81]

Figure 14. Forest plot for incidence of wound complications.


Study RRAbscesses


group = noRCT


Leung 2000Breukink 2005Staudacher 2007Baik 2011

1113

344179

108

0403

2541

10854


group = RCT


322 321

130

48103170

121

51101170




550 583


60.8%59.9%

----

1.141.27

[0.43; 3.02][0.40; 4.03]

3.7%34.7%5.0%

17.4%

7.3%15.8%

7.2%29.7%

4.060.252.202.00

[0.17; 95.61][0.03; 2.14]

[0.09; 53.26][0.42; 9.58]

39.2%40.1%

----

1.001.04

[0.27; 3.63][0.27; 3.99]

100%100%

----

1.081.17

[0.50; 2.36][0.50; 2.76]

8.9%17.2%13.0%

9.7%23.3%7.1%

0.941.530.33

[0.06; 14.63][0.26; 8.96][0.01; 8.13]

210.1 0.5 10

Figure 13. Forest plot for incidence of abdominal abscesses.


Arezzo et al. 43

laparoscopy surgery patients (10.2 vs. 12.4 days); theoverall MD was �2.2 days (95% CI �3.7 to �0.7,p¼ 0.005; Figure 17), with no differences in the sub-group analysis (MD �2.7 vs. �1.9, p¼ 0.651), butwith extreme heterogeneity (I2 79.8%).

Discussion

According to European registries, the prevalence ofrectal cancer exceeds 50 cases per 100,000 (womenand men). It accounts for almost 40% of all colorectal

Study RR

Anastomotic leakages


group = noRCT


Hu 2003Breukink 2005Morino 2005Staudacher 2007Baik 2011

02

1016

4

2531597982

143

106

203174

10839


group = RCT



310 316

582 592

152

8981

146

3100

8277

151

Zhou 2004Lujan 2009Kang 2010



210.1 0.5 10

64.7%74.1%

----

1.261.24

[0.76; 2.10][0.74; 2.10]

3.6%10.7%9.0%

31.0%10.4%

2.1%7.9%

13.4%36.5%14.2%

0.410.502.661.171.40

[0.02; 9.65][0.10; 2.53][0.77; 9.22][0.56; 2.44][0.42; 4.68]

35.3%25.9%

----

0.660.61

[0.28; 1.52][0.23; 1.62]

100%100%

----

1.051.03

[0.68; 1.61][0.65; 1.64]

17.7%26.2%1.4%

4.2%19.4%2.3%

0.360.534.83

[0.04; 3.41][0.19; 1.47]

[0.23; 99.86]

Figure 15. Forest plot for incidence of anastomotic leakage.


Study

group = noRCT

group = RCT

569 534

55.8%

RR

Reintervention

95%-CI W(fixed) W(random)

58.1%--

--0.780.77

[0.43; 1.42][0.40; 1.49]

20.8%8.0%

27.0%

21.0%12.1%25.0%

0.621.900.57

[0.22; 1.75][0.49; 7.37][0.22; 1.46]

44.2%41.9%

--

--

--

--

0.710.74

[0.35; 1.44][0.36; 1.53]

100%100%

0.750.75

[0.47; 1.19][0.47; 1.20]

10.7%25.7%7.8%

4.8%28.2%8.9%

0.240.821.00

[0.03; 2.03][0.34; 1.98][0.20; 4.88]

247 213

210.1 0.5 10

Events TotalTotal EventsLaparoscopy Open




Breukink 2005Morino 2005Staudacher 2007

567

419379

839

4198

108

322 321

183

48103170

410

3

51101170





Figure 16. Forest plot for incidence of re-intervention.



cancer cases23,24 representing one of the leading causesof cancer mortality, in which local recurrence of rectalcancer plays a major role. For mid and low rectal can-cers, total mesorectal excision remains the main-stay oftherapy. The feasibility of laparoscopic resection ofrectal cancer has been demonstrated for many yearswhen performed by expert laparoscopists, but whilethe laparoscopic approach in colon cancer has beenproved to be safe and feasible with equivalent long-term oncological outcome compared to open sur-gery,3–5 the safety of laparoscopic surgery for rectalcancer is still debated both for short-term and long-term outcomes.

While mid and long-term oncological results areawaited, the present study aimed to assess by meansof a systematic review and meta-analysis, the short-term safety of laparoscopic resection for extraperito-neal rectal cancer. Different meta-analyses have com-pared laparoscopic and open techniques forcolorectal25–29 and later for rectal cancers,30–33

although only including intra-peritoneal lesions. Thisis in fact the first meta-analysis that focuses only onextra-peritoneal rectal cancers. In this restricted fieldof mid and low rectal surgery, the issue of increasedtechnical challenges arises.

Since 2000, 11 studies12–22 have been published com-paring laparoscopic and open rectal resection in termsof safety. We restricted the beginning of the analysis tothe year 2000 in order to include studies performed with

a relative experience of the technique, as it had alreadybeen proposed around 8 years previously. This wasconsidered important in order to obtain results ashomogeneous as possible. Although a meta-analysisof only RCT studies might be considered preferable,the risk of bias analysis and the heterogeneity testshowed that extending the inclusion criteria to pro-spective non-randomized matched series would haveallowed a consistent level of evidence to be maintained.The heterogeneity of results was null or kept within areasonable range for most of the outcomes considered,in particular for the two primary ones, despite the factthat some of the study samples included in this analysiswere relatively small and that none of the studiesincluded had made an estimate of what sample sizewas needed to detect any differences between laparo-scopic and open surgery based on a well-defined pri-mary outcome. The sensitivity analyses showed that nostudy played an influential role on RR in the wholetime period. Where data were available, stage ofcancer disease, neoadjuvant chemoradiotherapy, pro-tective ileostomy rates, and type of surgery (partial ortotal mesorectal excision or abdominoperineal resec-tion) were all variables considered in the comparativeanalysis between open and laparoscopic groups. Thequality assessment for both RCT and noRCT studiesas measured by the Cochrane Collaboration’s tool forassessing risk of bias and the Newcastle-Ottawa scalewas interestingly high.

Study

group = noRCT

group = RCT

652 692 −3.18−2.21

100%

55.9%

44.1%

MDHospital stay

95%-CI W(fixed) W(random)

40.7%5.0%

10.3%

12.8%9.3%

11.1%

100%

33.2%

66.8%

----

----

[−3.80; −2.56][−3.74; −0.68]

−4.16−2.72

[−4.99; −3.33][−5.73; −0.29]

−1.93−1.93

[−2.87; −1.00][−3.51; −0.35]

−5.20−0.70−1.70

[−6.17; −4.23][−3.49; −2.09][−3.64; −0.24]

3.1%4.7%3.0%9.4%5.2%

10.4%8.3%

7.9%9.2%7.7%

10.9%9.4%

11.1%10.6%

2.000.30

−7.00−1.60−2.60−1.80−3.00

[−1.53; −5.53][−2.56; −3.16]

[−10.62; −3.38][−3.63; −0.43][−5.33; −0.13][−3.73; −0.13][−5.16; −0.84]

234 240

−10 −5 50 10

418

Total Mean SDSD Total MeanLaparoscopy Open

452Fixed effect modelRandom effects model



Leung 2000Hu 2003Breukink 2005Morino 2005Staudacher 2007Baik 2011Seshadri 2012

25204198

1085472

3425419379

10872

5.54.3

11.09.56.63.86.5

8.35.54.33.7

11.08.76.7

12.018.312.011.410.17.0

12.0

10.018.019.013.012.78.8

15.0

8251

101

8948

103

3.15.57.3

3.48.36.8

8.110.88.2

13.311.59.9

Zhou 2004Ng 2008Lujan 2009




Figure 17. Forest plot for length of hospital stay.


Arezzo et al. 45

An important aspect of the present analysis is thatall studies were performed at specialized centres byhighly experienced surgeons, and that patients wererigorously selected, particularly those staged T4 byTNM stage who were excluded in all but two studies.

In the light of the methodology of the present study,the results obtained appear highly significant. The mainresult of the meta-analysis was that the incidence ofoverall morbidity at 30 days showed a significant reduc-tion in the laparoscopic group compared to open sur-gery with an RR of 0.83. Furthermore the mortalityanalysis showed a trend in favour of laparoscopy,although it may not have reached statistical significancedue to the relative rarity of the event. The subgroupanalysis confirmed that surgical complications arealso significantly lower in the laparoscopic group,while medical complications showed a consistenttrend in favour of laparoscopy although with no stat-istical significance. Apart from the operative time,which was significantly in favour of open surgery, theanalysis of all the studies included showed a clearadvantage for laparoscopy also in the specific analysisof other comparative outcomes. Blood loss, time forbowel movement recovery, food intake recovery, needfor blood transfusion, wound complications, and dur-ation of hospital stay were all significantly in favour oflaparoscopy, while no significant difference wasobserved in terms of anastomotic leakages, intra-operative injuries, abscess formation, and the need forsurgical re-intervention at 30 days.

While most of the above-mentioned results wereexpected, as they confirmed the well-known advantagesof minimally invasive techniques shown in other surgi-cal procedures, the absence of any difference in anasto-motic leakage rates was particularly striking. In fact, inthe past few years, concern has been expressed aboutthe increase in anastomotic leakage risk associated withlaparoscopy. This risk was accounted for by the diffi-cult access of laparoscopic linear staplers to the distalrectum in a narrow pelvis, the oblique transectionfrom right to left or from anterior to posterior(depending on the position of the trocar used for thestapler), and the difficulty of cephalad traction on therectum. This had led the EAES Consensus panel ofexperts to affirm that ‘a cautionary note should becommented in relation to the anastomotic leak rate,in relation to the difficulty observed for low rectaltransection in the obese or in those with a narrowpelvis space’.34 In recent years, the advent of newtechnologies, such as the ultrasonic scalpel and articu-lated stapler, and the increase in surgical experiencehas resulted in a progressive optimization of the tech-nique which has recently led to an equivalence in leak-age rates and a lower incidence of surgicalcomplications with the laparoscopic approach.

Only the lack of sufficient data referring to sexualand urinary dysfunctions, postoperative quality of lifeand R0 achievement, and sphincter preservation ratesprevented further analyses which would also undoubt-edly have proved interesting from being performed.

The results obtained should in any case be inter-preted with caution as the present analysis shows cer-tain limitations. Despite the high quality assessment forlow risk of bias, most of the studies did not have short-term complications as primary outcome, while otherslacked important data with regard to secondaryoutcomes.

In spite of the above-mentioned limitations, it can beconcluded that, based on the evidence of both rando-mized and prospective matched series, laparoscopicresection appears to have clinically measurable short-term advantages in patients with primary resectableextraperitoneal rectal cancer. Although technicallydemanding, laparoscopic rectal resection for extraper-itoneal cancer in selected patients can be consideredsafe and guarantees faster recovery; it is therefore tobe recommended under optimal conditions.

Funding

The work described in this paper was funded in part by theEuropean Commission within the STIFF-FLOP FP7

European project FP7/ICT-2011-7-287728.

Conflict of interest

They authors have no conflict of interests to declare, includ-

ing financial interests and relationships and affiliations rele-vant to the subject of their manuscript.

References

1. Jacobs M, Verdeja JC and Goldstein HS. Minimally inva-

sive colon resection (laparoscopic colectomy). Surg

Laparosc Endosc 1991; 1: 144–150.2. Lacy AM, Garcia-Valdecasas JC, Delgado S, et al.

Laparoscopy-assisted colectomy versus open colectomy

for treatment of non-metastatic colon cancer: a rando-

mised trial. Lancet 2002; 359: 2224–2229.3. Guillou PJ, Quirke P, Thorpe H, et al. Short-term end-

points of conventional versus laparoscopic-assisted sur-

gery in patients with colorectal cancer (MRC CLASICC

trial): multicentre, randomised controlled trial. Lancet

2005; 365: 1718–1726.4. Clinical Outcomes of Surgical Therapy Study Group.

A comparison of laparoscopically assisted and open colec-

tomy for colon cancer. N Engl J Med 2004; 350:

2050–2059.5. Veldkamp R, Kuhry E, Hop WC, et al.; Colon cancer

Laparoscopic or Open Resection Study Group

(COLOR). Laparoscopic surgery versus open surgery for

colon cancer: short-term outcomes of a randomised trial.

Lancet Oncol 2005; 6: 477–484.


6. Heald RJ, Moran BJ, Ryall RD, et al. Rectal cancer: theBasingstoke experience of total mesorectal excision,1978–1997. Arch Surg 1998; 133: 894–899.

7. Ries LA, Wingo PA, Miller DS, et al. The annual reportto the nation on the status of cancer, 1973–1997, with aspecial section on colorectal cancer. Cancer 2000; 88:2398–2424.

8. Lacy A. Colon cancer: laparoscopic resection. Ann Oncol2005; 16(Suppl 2): ii88–92. Review.

9. Ortiz H, Armendariz P and Yarnoz C. Early postopera-

tive feeding after elective colorectal surgery is not a bene-fit unique to laparoscopy-assisted procedures. Int JColorectal Dis 1996; 11: 246–249.

10. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMAstatement for reporting systematic reviews and meta-ana-lyses of studies that evaluate health care interventions:

explanation and elaboration. Ann Intern Med 2009; 151:65–94.

11. Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis ofobservational studies in epidemiology: a proposal for

reporting. Meta-analysis Of Observational Studies inEpidemiology (MOOSE) group. JAMA 2000; 283:2008–2012.

12. Zhou ZG, Hu M, Li Y, et al. Laparoscopic versus opentotal mesorectal excision with anal sphincter preservationfor low rectal cancer. Surg Endosc 2004; 18: 1211–1215.

13. Morino M, Allaix ME, Giraudo G, et al. Laparoscopicversus open surgery for extraperitoneal rectal cancer: aprospective comparative study. Surg Endosc 2005; 19:1460–1467.

14. Lujan J, Valero G, Hernandez Q, et al. Randomizedclinical trial comparing laparoscopic and open surgeryin patients with rectal cancer. Br J Surg 2009; 96:

982–989.15. Kang SB, Park JW, Jeong SY, et al. Open versus laparo-

scopic surgery for mid or low rectal cancer after neoad-

juvant chemoradiotherapy (COREAN trial): short-termoutcomes of an open-label randomised controlled trial.Lancet Oncol 2010; 11: 637–645.

16. Leung KL, Kwok SP, Lau WY, et al. Laparoscopic-assisted abdominoperineal resection for low rectal adeno-carcinoma. Surg Endosc 2000; 14: 67–70.

17. Ng SS, Leung KL, Lee JF, et al. Laparoscopic-assisted

versus open abdominoperineal resection for low rectalcancer: a prospective randomized trial. Ann Surg Oncol2008; 15: 2418–2425. Erratum: Ann Surg Oncol 2009; 16:

229.18. Staudacher C, Vignali A, Saverio DP, et al. Laparoscopic

vs. open total mesorectal excision in unselected patients

with rectal cancer: impact on early outcome. Dis ColonRectum 2007; 50: 1324–1331.

19. Baik SH, Gincherman M, Mutch MG, et al.Laparoscopic vs open resection for patients with rectal

cancer: comparison of perioperative outcomes and long-term survival. Dis Colon Rectum 2011; 54: 6–14.

20. Hu JK, Zhou ZG, Chen ZX, et al. Comparative evalu-

ation of immune response after laparoscopical andopen total mesorectal excisions with anal sphincter

preservation in patients with rectal cancer. World J

Gastroenterol 2003; 9: 2690–2694.

21. Breukink SO, Pierie JP, Grond AJ, et al. Laparoscopic

versus open total mesorectal excision: a case-control

study. Int J Colorectal Dis 2005; 20: 428–433.22. Seshadri RA, Srinivasan A, Tapkire R, et al.

Laparoscopic versus open surgery for rectal cancer after

neoadjuvant chemoradiation: a matched case-control

study of short-term outcomes. Surg Endosc 2012; 26:

154–161.23. GLOBOCAN IARC. GLOBOCAN 2008 Fast Stats.

Available at: http://globocan.iarc.fr/factsheets/popula

tions/factsheet.asp?uno¼900#BOTH (2011, accessed

Mar 2011).24. Micheli A. Cancer prevalence in European registry areas.

Ann Oncol 2002; 13: 840–865.25. Li MZ, Xiao LB, Wu WH, et al. Meta-analysis of lap-

aroscopic versus open colorectal surgery within fast-track

perioperative care. Dis Colon Rectum 2012; 55: 821–827.

26. Kuhry E, Schwenk WF, Gaupset R, et al. Long-term

results of laparoscopic colorectal cancer resection.

Cochrane Database Syst Rev 2008; (2): CD003432.27. Ohtani H, Tamamori Y, Arimoto Y, et al. A meta-ana-

lysis of the short- and long-term results of randomized

controlled trials that compared laparoscopy-assisted and

conventional open surgery for colorectal cancer. J Cancer

2011; 2: 425–434.28. Sammour T, Kahokehr A, Srinivasa S, et al.

Laparoscopic colorectal surgery is associated with a

higher intraoperative complication rate than open sur-

gery. Ann Surg 2011; 253: 35–43.

29. Ma Y, Yang Z, Qin H, et al. A meta-analysis of laparos-

copy compared with open colorectal resection for colo-

rectal cancer. Med Oncol 2011; 28: 925–933.30. Huang MJ, Liang JL, Wang H, et al. Laparoscopic-

assisted versus open surgery for rectal cancer: a meta-

analysis of randomized controlled trials on oncologic

adequacy of resection and long-term oncologic outcomes.

Int J Colorectal Dis 2011; 26: 415–421.

31. Racheva G and Elmusharaf H. A systematic review and

meta-analysis of randomized and non-randomized stu-

dies comparing laparoscopic and open abdominoperineal

resection for rectal cancer. Colorectal Dis 2012 (Epub

ahead of print).

32. Trastulli S, Cirocchi R, Listorti C, et al. Laparoscopic vs

open resection for rectal cancer: a meta-analysis of ran-

domized clinical trials. Colorectal Dis 2012; 14:

e277–e296.33. Aziz O, Constantinides V, Tekkis PP, et al. Laparoscopic

versus open surgery for rectal cancer: a meta-analysis.

Ann Surg Oncol 2006; 13: 413–424.34. Siegel R, Cuesta MA, Targarona E, et al.; European

Association for Endoscopic Surgery (EAES).

Laparoscopic extraperitoneal rectal cancer surgery: the

clinical practice guidelines of the European Association

for Endoscopic Surgery (EAES). Surg Endosc 2011; 25:

2423–2440.

Arezzo et al. 47

Laparoscopy for rectal cancer reduces short-term mortality and morbidity: results of a systematic...

Documents

Transcript of Laparoscopy for rectal cancer reduces short-term mortality and morbidity: results of a systematic...