Laparoscopic Versus Robotic Versus Open Surgery for Rectal Cancer

(a)

Patient population

Intervention

Comparator

Outcomes studied

Rectal cancer, post neoadjuvant chemo-radiotherapy

Laparoscopically performed TME

Open TME

Procedure related morbidity

Length of stay, complications – grade 3/4/5, anastomotic leak, reoperation

Oncologic

CRM involvement, distal resection margin involvement, distance to CRM, distance to distal resection margin, LN yield, completeness of TME

Disease specific survival, overall survival, local recurrence

(b)

Patient population

Intervention

Comparator

Outcomes studied

Rectal cancer, post neoadjuvant chemo-radiotherapy

Robotically performed TME

Laparoscopically performed TME

Procedure related

Length of stay, complications – grade 3/4/5, anastomotic leak, reoperation, cost, open conversion

Oncologic

CRM involvement, distal resection margin involvement, distance to CRM, distance to distal resection margin, LN yield, completeness of TME

Disease specific survival, overall survival, local recurrence

Compared with laparoscopic surgery, does robotic-assisted surgery result in better outcomes after rectal cancer treatment? (Table 46.1b)

Methods/Search Strategy

Studies reporting short- and long-term results for rectal cancer surgery in which a proportion of patients received neoadjuvant treatment were reviewed. Rectal cancer was defined as a tumor 15 cm or less from the anal verge. Laparoscopic surgery was defined as completion of the pelvic dissection using laparoscopic instruments. Non-conventional laparoscopic techniques were excluded (e.g., hand-assisted or single-port surgery). Robotic surgery was defined as completion of the pelvic dissection using a robotic platform. PubMed, Ovid, Web of Science and Cochrane databases were searched using terms “rectal cancer”, “laparoscopy”, “open”, “robot”, and “robotic” for studies up to December 1, 2015. We sought to review the highest quality evidence with emphasis on Level 1/2 data. For comparison (A), five multicenter RCTs have reported data pertinent to this question and therefore the meta-analysis focused on their results. For comparison (B), no prospective randomized data is available. If multiple reports were published from one institution, the most recent series was evaluated.

Results: (A): Laparoscopic Surgery Versus Open Surgery for Rectal Cancer

Description of Studies

Five multicenter RCTs have been undertaken to evaluate laparoscopic surgery versus open surgery in rectal cancer: the CLASICC, COREAN, COLOR II, ACOSOG Z6051 and ALaCaRT trials [1–5]. Trial characteristics and results are summarized in Table 46.2. Long-term outcomes are reported by CLASICC, COREAN and COLOR II. Each trial was designed with a slightly different rationale for power calculation and outcome reporting. CLASICC recruited 413 colon and 381 rectal cancer patients and was powered not by an outcome assessment; but on the need to evaluate laparoscopic colorectal surgery in a trial setting by examining differences between treatment arms for a range of endpoints [1]. The COREAN trial recruited 170 patients per arm with tumors ≤10 cm from the anal verge after neoadjuvant chemoradiotherapy [2]. The trial assessed non-inferiority of laparoscopic surgery based on 3-year disease-free recurrence. COLOR II recruited 1044 patients (699 laparoscopic vs. 345 open) with tumors ≤15 cm from the anal verge to assess non-inferiority of laparoscopic surgery based on 3-year local recurrence rates [3]. ACOSOG Z6051 recruited 462 patients (240 laparoscopic vs. 222 open) with tumors ≤12 cm from the anal verge after neoadjuvant treatment [4]. The trial was powered to detect non-inferiority of laparoscopic surgery based on a composite pathological endpoint: completeness of TME, negative circumferential margin (CRM) and negative distal resection margin (DRM). ALaCaRT had a similar design, recruiting 475 patients (238 laparoscopic vs. 237 open) with tumors ≤15 cm from the anal verge to assess non-inferiority of laparoscopic surgery based on a composite pathological endpoint [5].

Table 46.2

Perioperative and oncologic outcomes from published multicenter randomized controlled trials comparing laparoscopic versus open surgery for rectal cancer (laparoscopic/open)

Trial	CLASICC (Rectal cancers)		COREAN		COLOR II		ACOZOG Z6051		ALaCaRT
Design	Phase 3 multicenter RCT		Non-inferiority phase 3 multicenter RCT		Non-inferiority phase 3 multicenter RCT		Non-inferiority phase 3 multicenter RCT		Non-inferiority phase 3 multicenter RCT
Number of centers	27		3		30		35		24
Location	United Kingdom		South Korea		Europe		North America		Australia and NZ
Number (lap/open)	253/128 (ITT) 160/132 (ATG)		170/170		699/345		240^a/222		238/235
Time period	1996–2002		2006–2009		2004–2010		2008–2013		2010–2014
Site of tumors			Mid-low rectum <10 cm from AV		≤15 cm of AV		≤12 cm from AV		Within 15 cm of AV
Treatment group	Lap	Open	Lap	Open	Lap	Open	Lap	Open	Lap	Open
% receiving neoadjuvant treatment	58	60	100	100	59	58	100	100	50	49
BMI median (mean^b)	26	25^b	24.1	24.1	26.1	26.5^b	26.4	26.8^b	27	26
Perioperative outcomes
Conversion	34		1		16		11		9
LOS (days)	11	13	8	9	8	9	7.3	7.0	8	8^b
Mean operative time (min)			244	197	240	188	266	220	210	190^b
Complications Grade 3/4 (%) Grade 5 (%)			0.0	0.0	1.1	1.7	21.7 0.8	21.1 0.9	18.5 0.4	26.4 0.8
Anastomotic leak (%)	10	7	1.2	0	13	10	2.1	2.3	3	3
Reoperation (%)			1.8	1.8	6	6	5	2.3
Oncological outcomes
CRM negative ≤1 mm (%)	84	86	97.1	95.9	93	91	87.9	92.3	93	97
DRM negative (%)	No difference						98.3	98.2	99	99
Complete/nearly complete TME (%)			91.8	88.2	97	98	92.1	95.1	97	99
Composite endpoint							82	87	82	89
Mean LN yield			17	18	13	14	17.9	16.5
Distance DRM (mm)			20	20	30	30^b	32	31^b	26	30
Distance to CRM (mm)			9	8	10	10^b	10.5	12.8^b	10	12
3 year DFS OS LR	70.9 74.6 9.7	70.4 66.7 10.1	79.2 91.7 2.6	72.5 90.4 4.9	74.8 86.7 5.0	70.8 83.6 5.0

Data presented as median unless otherwise stated

RCT randomized controlled trial, NZ New Zealand, TME total mesorectal excision, AV anal verge, BMI body mass index, LOS length of stay, ITT intention to treat, ATG actual treatment group, CRM circumferential resection margin, DRM distal resection margin, LN lymph node, DFS disease free survival, OS overall survival, LR local recurrence

^aIncludes 34 Robot assisted procedures

^bMean

Short Term Outcomes

Length of hospital stay (LOS)

No differences in LOS were seen in the COREAN, ACOSOG Z6051, or ALaCaRT trials. In contrast, CLASICC and COLOR II reported lower LOS in the laparoscopic group (CLASICC: 11 vs. 13 days; COLOR II: 8 vs. 9 days). The COREAN trial demonstrated a trend towards reduced LOS in the laparoscopic arm (8 vs. 9 days P = 0.056), but unlike ALaCaRT and ACOSOG Z6051, consistently better short-term outcomes were also observed for the laparoscopic group (e.g. earlier passing of flatus, earlier defecation, and resumption of normal diet). The equivocal short-term outcomes for treatment groups in ACOSOG Z6051 and ALaCaRT may relate to the use of hybrid approaches permitted in the open arms.

Complications, Anastomotic Leak and Reoperation Rates

Clavien-Dindo grade 3/4/5 complication were comparable for open and laparoscopic surgery in the ACOSOG Z6051 and ALaCaRT trials. The COREAN and COLOR II trials reported similar rates of infectious and noninfectious complications and short-term mortality in each trial arm. CLASICC reported higher perioperative morbidity and mortality in patients converted from laparoscopic to open surgery. Complication rates for laparoscopic, open and converted rectal cancer operations were 32 %, 37 % and 59 %, respectively. No trial reported a difference in anastomotic leak or reoperation rates.

Short Term Outcomes Summary

Mean LOS after rectal cancer surgery ranged from 7 to 9 days across the treatment arms. Although two trials reported reduced LOS with laparoscopic surgery, we conclude that LOS is comparable for each approach. Rates of complications, anastomotic leaks and reoperation are also equivocal.

Conclusion

The assessed short-term outcomes are comparable for laparoscopic and open surgery.

GRADE: HIGH QUALITY

Oncologic Outcomes

Circumferential resection margin involvement

Excepting CLASICC, all trials reported non-involved CRM rates in excess of 87 %, underlining the technical skills of the surgeons participating in this study. No trial was powered based solely on CRM assessment. Clear CRM rates were comparable in all five trials (Table 46.2). In a subgroup analysis of anterior resections, CLASICC reported a nonsignificant trend towards higher CRM involvement for laparoscopy (12 % vs. 6 % based on 16 positive CRMs in 129 laparoscopic versus 4 positive CRMs in 64 open resections). Distance to CRM was comparable in COREAN, COLOR II and ALaCaRT. However, ACOSOG Z6051 reported reduced distance to CRM with laparoscopy (10.8 vs. 12.8 mm, P = 0.03).

Distal margin

DRM involvement was low (1–2 %) and incidence was equivalent where it was reported (Table 46.2) [1, 4, 5].

Complete/nearly complete TME

No differences in rates of complete/nearly complete TME were reported where this outcome was assessed (Table 46.2) [1–5].

Composite pathological outcomes

Both ACOSOG Z6051 and the ALaCaRT trials used a composite pathological assessment as their primary outcome measure. Both trials were powered based on the assumption that 90 % of rectal cancer resections are oncologically complete (CRM negative, DRM negative, and complete/nearly complete TME). ACOSOG Z6051 stated non-inferiority would be declared if the lower border of the 95 % CI for difference between groups was >6 %. ALaCaRT set a similar non-inferiority threshold of >8 %. In ACOSOG Z6051, a complete specimen was achieved in 81.7 % of laparoscopic versus 86.9 % of open resections (5.2 % difference, lower bound 95 %, CI −10.8). For ALaCaRT, a complete specimen was achieved in 82 % of laparoscopic versus 89 % of open resections (7.0 % difference, lower bound 95 %, CI −12.4). For each trial, non-inferiority was not established.

Only gold members can continue reading. Log In or Register to continue