Outcomes of Knife-Assisted Full-Thickness Resection for Rectal Lesions: A Multicenter Study

Background

Minimally invasive, organ-preserving techniques have become the preferred approach for early rectal cancer resection, replacing total mesorectal excision due to its high morbidity (12.1%-13.4%) and mortality (1%-1.5%). Recent meta-analysis data indicate that deep submucosal invasion is not an independent risk factor for lymph node metastasis in colorectal cancer (CRC), with 37.9% of deeply invasive cancers lacking other histologic high-risk features, making them suitable for curative endoscopic resection. However, current endoscopic techniques such as endoscopic submucosal dissection (ESD), Transanal Minimally Invasive Surgery (TAMIS), endoscopic device-assisted full-thickness resection (FTRD), and endoscopic intermuscular dissection (EID) have limitations in specific scenarios.

Endoscopic knife-assisted full-thickness resection (kFTR) has emerged as a novel technique for local excision of rectal lesions but remains under-evaluated in large cohort studies. The objective of this study is to assess the technical and clinical outcomes of kFTR for rectal lesions, including technical success, en-bloc and R0 resection rates, and the incidence of complications.

Study Endpoints

The primary outcomes of this study include technical success, en-bloc resection, and R0 resection rates. Additionally, complications such as intraprocedural and delayed bleeding, post-polypectomy syndrome, and other procedure-related adverse events will be evaluated. Secondary outcomes include the length of hospital stay, the need for adjunctive surgery due to complications or non-curative resections, locoregional or non-salvageable recurrence, and mortality.

Methods

This study follows a retrospective, international, multicenter cohort design. Eligible participants include patients aged 18 years or older who underwent kFTR at the participating centers. Cases with missing data on kFTR technical outcomes are excluded. Due to the limited availability of previous data on rectal kFTR, all eligible cases will be included without a predefined sample size calculation.

Data will be collected from electronic medical records, endoscopic images, and videos. Extracted patient characteristics include age, sex, comorbidities, ASA classification, prior neoadjuvant treatment, and preoperative staging. Lesion characteristics such as size, location, morphology (JNET and Paris classification), previous resection attempts, and fibrosis grade will also be recorded. Procedure-related variables include anesthesia type, dissection technique, defect closure, procedural success, procedural time, complications, and hospital stay. Histopathological data will be analyzed for tumor staging (pT), radicality, and high-risk features for nodal metastasis. Clinical outcomes such as adjuvant therapy, follow-up duration, recurrence, and mortality will be documented.

Variable Definitions

Technical success is defined as the completion of the procedure with lesion removal without conversion to another technique. Procedural time is measured from submucosal injection to procedure completion. Bleeding is classified as intraprocedural (requiring intervention) or delayed (occurring within 30 days post-procedure and necessitating medical intervention). Post-polypectomy syndrome is characterized by abdominal pain within 24 hours of the procedure that resolves with conservative management. Tumor budding is categorized into low (0-4 buds), intermediate (5-9 buds), and high (≥10 buds) groups. Metastases are classified as locoregional (spread to surrounding tissues or regional lymph nodes) or distant (e.g., liver, lung, peritoneum).

Statistical Analysis

Categorical variables will be presented as frequencies and percentages, while continuous variables will be reported as means (standard deviation) or medians (interquartile range), depending on data distribution. Subgroup analyses will assess effectiveness and safety outcomes based on lesion location, size, fibrosis, and resection technique. Comparisons between categorical variables will be conducted using chi-squared or Fisher's exact tests, while t-tests will be used for continuous variables.

Recurrence after resection will be analyzed over time using Kaplan-Meier survival analysis, with log-rank tests applied for group comparisons. Univariable analyses will assess associations between independent variables and key outcomes, including en-bloc resection, recurrence, and complications. Multivariable regression models will incorporate relevant variables to evaluate their independent associations with these outcomes. A p-value of ≤0.05 will be considered statistically significant. All statistical analyses will be performed using R software.

Data Management

Data will be anonymized, with each center maintaining a separate key linking patient identifiers to unique study IDs. Clinical data will be stored securely on UZ Gent servers and retained for 10 years following study completion before being permanently deleted. Access to the data will be restricted to study investigators.