Guía de operaciones para el desarrollo del asistente

The ‘Preferred reporting items for systematic reviews and meta-analyses’ checklist[47] applied to the CoSMIC systematic review.

PRISMA-P 2015 Checklist

Title Item

Identification 1a Yes

Update 1b N/A

Registration 2 PROSPERO database 2016 CRD42016039163

Authors Contact 3a See title page

Authors Contributions

3b AKS conceived the idea; AKS and AC developed the systematic

review protocol (PRISMA-P); AC carried out the systematic review and extracted the data; AKS and AC analysed the data; All authors contributed to writing the final manuscript.

Amendments 4 N/A

Support

Sources 5a This review was supported by a small grant from the Dickinson Trust.

Sponsor 5b N/A

Role of Sponsor 5c N/A

INTRODUCTION

Rationale 6 Surgery can improve the 5-year survival of patients diagnosed with synchronous colorectal cancer and liver metastases to 40%. The traditional ‘bowel-first’ approach removes the colorectal primary in the first operation, followed by the liver metastasis in a subsequent operation. Recent advances in liver surgery have allowed the ‘liver- first’ approach and simultaneous bowel/liver resection to be safe alternatives. Simultaneous resection removes the macroscopic tumour burden in a single but more extensive operation that requires careful patient selection. Currently, there is no evidence- based criteria of either patient comorbid factors or oncological considerations that determine the suitability of patients for the simultaneous approach. This systematic review identifies these factors based on current practice, published in studies that also

report 30-day and long term 1-5-year outcomes. It is anticipated that the study population will be heterogeneous between studies. A homogenous subgroup meta-analysis will allow for a comparison of outcomes, thus identifying the critical factors in patient selection.

Objectives 7 For

[P] patients diagnosed with synchronous colorectal cancer and liver metastases

[I] undergoing simultaneous bowel and liver resection surgery, [C] comparing patient selection factors (co-morbid and oncological) [O] with 30-day and long term 1-5 year outcomes.

METHODS

Eligibility Criteria 8 All studies reporting original data for patients undergoing a simultaneous bowel and liver resection for synchronous colorectal cancer and liver metastases. The selection criteria must be stated in the study methodology, and 30-day and 1-5 year outcomes reported. Limited to the English-language publications over the last 10 years.

Information Sources

9 PubMed and OvidSP medical databases

Search Strategy 10 #1 (cancer AND (colon OR rectal OR colorectal))

#2 synchronous

#3 (liver AND metasta*)

#4 (surgery AND (simultaneous OR combined))

#5 metachronous OR lung

#6 #1 AND #2 AND #3 AND #4 NOT #5

#7 #6 (limited to English language, Jan 2006 to Feb 2016)

#8 #7 (limited to NOT reviews)

Study Records

Data Management 11a Studies and the data will be managed in electronic format throughout the review period.

Selection Process 11b Studies will be selected for the review, and data extracted, by two independent reviewers.

Data Collection Process

11c Data will be extracted from the full papers of the included studies,

and stored on a software spreadsheet

Data Items 12 • Patient demographics (age, sex) • Patient co-morbidity (Charlson index) • Cancer staging (TNM)

• Bowel primary site

• Minor / Major (>3 segments) liver resection

• Descriptive criteria/rationale for selecting to simultaneous resection

•

Outcomes and Prioritization

13 • Length of hospital stay

• 30-day mortality (Clavien-Dindo score) and morbidity • 1-5-year outcome / survival curves

Risk of bias in individual studies

14 Studies may not publish comprehensive details of their study

population, or their rationale for simultaneous resection

(“availability of information”). As such, case-mix differences may exist among within the subgroup analysis.

Publication bias may also affect the power of the study.

Data Synthesis 15 Demographic data from the study cohort as well as the rates of complications and survival comparing each of the three surgical strategies are reported. Details of systemic chemotherapy received by each group were also reported. Statistical pooling of proportional estimates was explored using fixed effect models.

Meta-bias(es) 16 The long recruitment periods introduce bias though changes in diagnostic imaging, standards of surgical care and the availability of bowel-cancer specific chemotherapy.

RESULTS

Study Selection 17 The search strategy identified 223 unique citations of which 23 provided comparative data. Of these, 3 cohort studies met the inclusion criteria by reporting outcomes separately for the three surgical treatment pathways

Study

Characteristics

18 156 patients by van der Pool and colleagues

57 patients by Brouquet and colleagues 1,004 patients by Mayo and colleagues.

Risk of bias within studies

19 MINORS score was 14/22 (van der Pool and colleagues), 14/ 22

(Brouquet and colleagues) and 13/22 (Mayo and colleagues).

Results of individual studies

20 Table 6. Demographic profile of patients with colorectal cancer

with synchronous liver metastases. Table 7. Treatment allocations Table 8. Chemotherapy protocols

Synthesis of results

21 Patients were allocated to bowel-first surgery (748 patients,

62.2%), liver-first surgery (75, 6.2%) or synchronous liver/bowel surgery (380, 31.6%). Minor complications were similar between procedures. Major complications were consistent with a pooled fixed estimate of 9.1% (95%CI: 7.6%-10.8%, I(2) = 48%). Post- operative death was rare and consistent with a pooled fixed effect estimate of 3.1% (95%CI: 2.2%-4.3%, I(2) = 0%). Median follow- up ranged from 25.1 to 40.0 months, with a pooled underlying 5- year survival fixed effect estimate of 44% (I(2) = 39%).

Risk of bias across studies 22 - Additional analysis 23 - DISCUSSION Summary of evidence

24 3 cohort studies were identified comprising a pooled population of

1203 patients who completed treatment protocols between 1982 and 2011. Patients were allocated to bowel-first surgery (748 patients, 62.2%), liver-first surgery (75, 6.2%) or synchronous liver/bowel surgery (380, 31.6%). Minor complications were similar between procedures. Major complications were consistent with a pooled fixed estimate of 9.1%. Post-operative death was rare and consistent with a pooled fixed effect estimate of 3.1%. Median follow-up ranged from 25.1 to 40.0 months, with a pooled underlying 5-year survival fixed effect estimate of 44%.

Limitations 25 The definitions of ‘synchronous’ are inconsistent between the studies included in the review. The long recruitment periods also introduce bias though changes in diagnostic imaging, standards of surgical care and the availability of bowel-cancer specific

chemotherapy.

Conclusions 26 This review assesses outcomes of patients with colorectal cancer

with synchronous liver metastases managed by either

synchronous, sequential liver-first or bowel-first surgery. Overall treatment-related mortality is low, and survival is similar among the three groups. These findings provide support for the continued use of all three pathways until better evidence to guide selection of an individual treatment option is available.

FUNDING

Funding 27 This review was supported by a small grant from the Dickinson Trust.