2.1 Guideline Questions and Outcomes

The use of neoadjuvant chemoradiation has become increasingly common in the treatment of patients with rectal cancer. However, the rates of neoadjuvant chemoradiotherapy across institutions, as well as countries in Europe is unknown, and is likely to vary. Thus, we will evaluate the subgroup effect of the use of these techniques in patients that have received neoadjuvant therapy (including the use of chemotherapy or radiation) versus upfront resection. Additionally, given that the location of rectal tumors (or distance from the anal verge) can greatly impact the technical difficulty of resection, we will compare clinical outcomes between patients with low- versus mid-rectal cancers.

2.2 Guideline Development Process

Given the absence of structured guidance for guideline protocols, this document is informed by best applicable reporting standards from the Preferred Reporting Items for Systematic reviews and Meta-Analysis Protocols (PRISMA-P) [8], and addresses key points in the Appraisal of Guidelines for Research and Evaluation (AGREE) II extension for surgical interventions (AGREE-S) [9]. This guideline will follow rigorous methodological standards set by the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) framework [10]. We registered this guideline proposal on the International Guidelines Library here.

We will invite members of the EAES to provide input on the content of this protocol via social media as well as the EAES newsletter. We will also provide a link to the protocol on the EAES website. The steering group will consider all suggestions for amendments to the protocol. Any deviations from the study protocol will be explained in the final manuscript, as applicable.

2.3 Steering Group

The steering group will consist of three co-chairs: a supervising methodology chair who is a certified master guideline developer, and a junior methodology cochair with experience in five previous clinical practice guidelines. Neither methodology co-chairs have personal, financial, or intellectual conflicts of interest [11]. The third cochair is a gastrointestinal surgeon with expertise in TaTME and has co-authored previous clinical practice guidelines, as well as prior publications on the utility of TaTME for rectal cancer.

2.4 Guideline Methodologist

A certified master guideline developer and chair (International Guideline Development Credentialing & Certification Program - INGUIDE certificate number: 2022-L3-V1-00014), SAA, will act as a supervising cochair for the junior cochair BH. This member has actively contributed to the development of over 17 clinical practice guidelines. Additionally, one trainee methodologist completing INGUIDE certification for “guideline methodologist” will assist in the guideline planning and development.

2.5 Guideline Panel and External Advisors

The multidisciplinary, expert guideline panel will include six gastrointestinal surgeons. Three of these surgeons will have expertise in performing transanal total mesorectal excision (TaTME), while the other three surgeons will have experience with performing open, laparoscopic, or robotic total mesorectal excision (TME) for the treatment of rectal cancer. Additionally, the panel will include one pathologist, one radiologist, and two patient partners. These ten members will be recognized as voting members during the synchronous consensus meeting.

The panel will be completed by two gastrointestinal surgeons who are considered as opinion leaders and have authored previous landmark randomized controlled trials or comparative cohort studies on the topic of TaTME vs TME for rectal surgery. To align with standards from the Guidelines International Network, these members will be considered as having an intellectual conflict of interest and will provide their expert input as nonvoting members of the panel [12]. Finally, we will include one GRADE Auditor who will ensure that our guideline development processes adhere to the core tents of GRADE as a nonvoting member of the panel. We will apply the criteria outlined by the International Committee of Medical Journal Editors (ICMJE) criteria for authorship to consider any member of the expert panel for authorship in the peer-reviewed publication of our final guideline report [13].

2.6 Systematic Review Group

This group will perform all stages of the systematic review. The group will be comprised of an expert systematic reviewer who will coordinate all activities of the Evidence Synthesis Group, as well as two trainees who have completed a structured curriculum on systematic review methodology.

2.7 Statistical Analysis Group

This group will be led by an expert statistician (DM) with experience in evidence synthesis for several clinical practice guidelines. This senior lead analyst will guide a junior lead analyst in performing all data analysis for the systematic review and clinical practice guideline.

2.8 Outcomes

The steering group will draft a list of potential outcomes by performing a scoping review of recently published trials [14]. Panel members will then propose additional outcomes for consideration. All panelists will participate, including patient partners. To identify outcomes of interest, panel members will be asked to rate the importance of each outcome using the GRADE scale to identify important (score 4-6) and critical (score 7-9) items for inclusion [15]. If substantial variation exists among panel evaluations of outcomes, a synchronous panel meeting will be conducted to address conflicts and reach consensus. To guide this discussion, we will apply Cochrane guidance in focusing on patient-important outcomes relevant for clinicians and policymakers [16].

2.9 Objective Determination of Minimal Important Differences

The steering group will also prompt panel members to rate the utility of each outcome, with 0 being the worst possible health state (death), and 1 being the best possible health state. All panelists will participate, including patient partners. To accomplish this, we will pose questions in the following format:

“What do you consider is the utility value of major postoperative complications (Clavien-Dindo ≥ 3b; e.g., reoperation, the need for admission to ICU)? [17].

Utility represents the strength for an individual's preference for a given outcome, where zero reflects states of health equivalent to worst imaginable health (death), and 1 reflects perfect or best imaginable health.”

Due to the small sample, we anticipate non-normally distributed data and therefore will select the median utility value for each outcome. In the event that there is significant variation in responses, a synchronous panel meeting will be held to discuss conflicts and achieve consensus among panel members.

We will use evidence-informed anchors as coefficients indicating trivial-to-small effect threshold (0.0135), small-to-moderate effect threshold (0.0321), and moderate-to-large effect threshold (0.0625) [18, 19]. This will facilitate the determination of absolute effect thresholds for trivial/small, small/moderate, and moderate/large effects.

We will then aggregate both positive and negative values associated with each outcome that informs the comparison between our interventions TaTME, laparoscopic TME, and robotic TME.

During discussions for the evidence-to-decision framework, we will compare absolute effect differences to empirically derived absolute effect thresholds, as well as the net benefit or harm/burden using the aggregate coefficient compared to coefficient thresholds [19]. The outcomes will be considered within a fully contextualized for all important and critically important outcomes during the evidence-to-decision process [20].

2.10 Literature Search

The junior cochair developed a systematic literature search with the guidance of an academic health sciences librarian who is an expert in systematic reviews (Supplementary Appendix 1). The junior cochair is an experienced systematic reviewer and will oversee the systematic review process by leading a systematic review team.

The junior cochair will query online databases including PubMED, Embase via Elsevier, & Cochrane Central Register of Controlled Trials to identify eligible articles of interest published before October 17th, 2024. The team will use the inclusion & exclusion criteria listed in Table 1 to select articles for inclusion.

We will upload articles identified by the systematic literature search to Covidence (Covidence systematic review software, Veritas Health Innovation, Melbourne, Australia). Next, the Systematic Review Group will conduct calibration screening using a sample of 10 abstracts via an Excel spreadsheet. Once calibration is complete, two trainees from the Evidence Synthesis Group will complete two rounds of screening, first by title & abstract and then by full-text. The expert systematic reviewer will hold another calibration meeting between both rounds of screening. The expert systematic reviewer will hold a synchronous meeting to address conflicts after both rounds of calibration screening to discuss conflicts and resolve discrepancies. The supervising cochair will be available to resolve any disagreements or provide guidance as needed.

Once eligible articles are identified, two reviewers from the Systematic Review Group will extract the following data elements into Covidence listed in Table 2.

Once data is collected, the expert systematic reviewer will hold a synchronous meeting to address conflicts. The supervising cochair will be available to settle disagreements as needed.

2.11 Methodology for Data Analysis

We will apply a random and fixed-effect meta-analysis to synthesize the evidence for our key question [21]. We will conduct meta-analyzes of proportions to calculate the baseline risk for each outcome, for the purposes of estimating absolute effect differences [22]. We will calculate Mantel-Haenszel pooled odds ratios for binary variables and the standardized mean difference for continuous variables, with corresponding 95% confidence intervals. We will consider inconsistency by using the I² statistic to explore the percentage of variability of effect estimates that is due to metagenetic rather than sampling error with the inverse variance method [23]. Additionally, we will further explore heterogeneity by evaluating the Q-statistic and the 95% predictive intervals that show the plausible range of effect size values in a given study setting [23-25]. Furthermore, we will consider the degree of overlap across confidence intervals [23]. The independent data analysis group will perform analyzes using the R statistical package version 4.0.3 using the meta package [22, 26], without involvement of the steering group or panel members.

We will compute the fixed-effect pooled estimate as a sensitivity analysis. Per Cochrane, where results are similar, the random effects model will be favored. If results differ substantially, we will favor the random effects model if the quality of smaller studies is sufficient. Where means and p-values, or confidence intervals will be given, the standard error and the standard deviation will be obtained by using the formula suggested by the Cochrane Collaboration [27]. Time-to-event data, also known as survival data, arise when we are interested not only in the event of an outcome but also on the time elapsing before this event is observed. Participants for whom we did not observe the event (censored) but contributed some time will be part of the analysis. For trials including time-to-event outcomes, we will extract the hazard ratio and the standard error of the logarithm of the hazard ratio [28]. If instead of the standard error, the corresponding 95% confidence interval or p-value is reported, we can use them to estimate the standard error [27]. These are standard outputs from a Cox proportional hazards model. If these data are not given, we can approximately estimate the hazard ratio using statistics estimated from a long-rank analysis [27]. If some studies give the hazard ratio but other give risk ratios (or the number of events and sample size in each group), we will combine them in a sensitivity analysis. We will explore for small-study effects using funnel plots and statistical tests (e.g., Egger's test) [29]. Additionally, we will perform a sensitivity analysis for articles at serious risk of bias to evaluate the durability of results between studies at low or moderate risk of bias and those at serious risk of bias. If non-randomized studies at moderate risk of bias and randomized studies at low risk of bias or with some concerns provide similar summary effect estimates to randomized and non-randomized studies at serious risk of bias, we will consider the summary effect estimate across studies and we will downgrade the evidence certainty by 1 (to account for unknown residual confounding) or 2 levels, depending on the weight of studies and the risk of bias they carry. If effect estimates differ, we will either use the overall effect estimate across studies and we will downgrade by 2 levels, or we will use the summary effect estimate of randomized studies at low risk of bias or non-randomized studies with moderate risk of bias and we will downgrade by 1 level.

2.12 Risk of Bias Assessment

Two reviewers will independently assess the risk of bias for each outcome in related studies using the Risk of Bias (RoB) 2.0 tool [30], or the risk of bias in nonrandomized studies of interventions (ROBINS-I) tool, version 2 [31]. The lead systematic reviewer will hold a synchronous meeting to resolve conflicts once risk of bias appraisal is complete. We will use the Robvis tool to provide detailed judgments on the risk of bias assessments [32].

2.13 GRADE Certainty of Evidence

We will assess the certainty of the evidence by evaluating the risk of bias across studies, imprecision, inconsistency, indirectness, and publication bias [33, 34]. As we will include nonrandomised studies, we will follow GRADE guidance by applying ROBINS-I, and applying the judgment of high certainty of evidence at baseline and downgrading as appropriate [35].

We will construct GRADE evidence summary profiles in MAGICapp. We will assess imprecision by interpreting the confidence interval of effect estimates in the context of the empirically derived effect thresholds, as well as the sample size of patients across comparisons [34]. We will assess inconsistency as stated in the data analysis section. We will consider the indirectness of comparisons based on their correlation with the PICO elements according to those listed in our inclusion criteria including the population, intervention, comparison, and outcomes (Table 1) [34].

2.14 Evidence-to-Decision Framework

During this process, external advisors and GRADE auditors will not be permitted to make judgments about evidence-to-decision domains or final recommendations. Following the meeting, panel members will be prompted to anonymously vote on the direction and strength of the recommendation(s), with the opportunity to suggest revisions or additional recommendations in alignment with GRADE methodology [37]. All panelists will participate, including patient partners. We will allocate dedicated time for patient partners to voice their opinions at the time of decision-making with respect to completing the following domains: benefits and harms of the intervention, values and preferences of patients and healthcare providers, acceptability of the intervention, and equity. In the process of developing the evidence-to-decision framework, the guideline panel will discuss the importance of each domain for the decision on the direction and the strength of the recommendation. We will consider consensus as agreement among < 80% of panel members, and any issues where consensus was not achieved will be documented in the final guideline report. The GRADE auditor will ensure that this process is conducted in alignment with GRADE methodology.

2.15 Publication and Dissemination

A systematic review as well as a clinical practice guideline will be published separately in Surgical Endoscopy and Other Interventional Techniques as the official journal of EAES. We will present the guideline at the EAES and SAGES annual meetings. The steering group will also arrange for broader dissemination via peer-reviewed scientific meetings, social media, letters to the editor, and other channels. Following dissemination, we will gauge the uptake of the guideline recommendations through an internal survey of EAES members 2 years after publication of the guideline report.

2.16 Target Users

This guideline applies primarily to gastrointestinal, endoscopic, and general surgeons, any other healthcare professionals involved in the treatment of low- and mid-rectal cancers, as well as patients. This guideline also applies to policymakers and hospital managers. We will develop a patient-friendly version of the guideline.

2.17 Feedback

The steering group will gather feedback on the guideline report from social media, the EAES website, letters to the editor, and other feedback mechanisms. This feedback will be considered in future updates to the guideline.

2.18 Implications for Practice and Research

This clinical practice guideline will guide patients, surgeons, interventional radiologists, and hospital managers in the evidence-informed surgical management of patients with low- and mid-rectal cancers. Policymakers will be able to use this guideline to support policy-related decisions around this topic. The systematic review, meta-analysis, and evidence-to-decision framework will further identify evidence gaps for future areas of research.

2.19 Barriers and Facilitators to Implementation

Barriers exist to the implementation of this clinical practice guideline. These include the failure to reach relevant stakeholder groups with the recommendations produced in this rapid update, as well as potential culture-related issues such as the innate resistance to change practice. Furthermore, surgeons not currently performing taTME or roTME, or surgeons without access to a robot may not be able to implement recommendations from t his guideline.

To improve the uptake of the recommendations from this rapid update, we will ensure wide dissemination of this clinical practice guideline including at peer-reviewed national and international meetings. We will also pursue publication in a peer-reviewed journal, such as Surgical Endoscopy. To address the potential resistance to change innate to institutional culture, we will empower and support panelists to present the findings of this rapid update at their local institution as well as at local and regional meetings. This will improve the dissemination of the findings of this guideline update, as our panelists will span various countries across Europe. We will meet with panelists as needed and develop tailored presentations.

Finally, we will work within the EAES to develop opportunities for training in taTME and/or roTME pending the results of the rapid guideline update. We will do this through the form of “Masterclass” sessions which are held at the annual meeting. The topics vary, and we will hold these as needed depending on the level of need and interest, informed through internal society surveys of our members.

2.20 Strengths and Limitations

We will develop this clinical practice guideline according to robust methodological standards outlined by GRADE. The involvement of an expert health sciences librarian will improve the comprehensiveness of our search to ensure that all relevant evidence is captured. We will develop patient-centered recommendations, informed by the evidence and the perspectives of a diverse multidisciplinary panel.

Limitations will include general challenges such as the element of subjectivity present with interpreting evidence, which we will aim to mitigate by adhering to robust methodology. Moreover, it may be difficult to derive conclusive evidence should study design and outcomes be heterogeneous. Should the results be heterogeneous, we will apply the random effects model where small studies are of sufficient quality. Moreover, should this arise, new evidence gaps will be identified. Furthermore, patient partners may not reflect the opinions of all patients, and this will be considered during the evidence-to-decision framework. We will select patient partners by considering the nature and extent of their health experiences, to mitigate their impact on the development of strong opini ons for any intervention, in any direction among our patient partners.

This guideline will not be intended to dictate clinical practice, and clinicians should use this guideline as a guide while making patient-tailored decisions.

2.21 Research Ethics

All members of the Guideline Development Group will declare financial, personal, or intellectual conflicts of interest. Conflicts will be managed according to the principles of the Guidelines International Network (GIN) [12]. Any member with relevant conflicts will participate as external advisors and will not participate in discussions about the direction or strength of the recommendations, the voting procedure, or the Delphi process following the consensus meeting.