A systematic review was conducted in accordance with the PRISMA 2020 guidelines (Fig. 1).

PRISMA 2020 flow diagram illustrating the study selection process for the systematic review on Transanal Endoscopic Microsurgery (TEM) for rectal neuroendocrine tumours (NETs)

A comprehensive search of PubMed/MEDLINE and Embase was performed (last updated August 1st 2025). The PubMed search combined free-text and controlled vocabulary (MeSH) for rectal NET/carcinoid and rigid transanal platforms. The free-text component was:

("rectal neuroendocrine tumour" OR "rectal neuroendocrine tumor" OR "rectal neuroendocrine neoplasm" OR “rectal carcinoid”) AND ("transanal endoscopic microsurgery" OR TEM OR "transanal endoscopic operation" OR TEO OR "transanal minimally invasive surgery" OR TAMIS OR "transanal endoscopic surgery" OR TES OR (robotic AND TAMIS)). For Embase, the following Emtree/free-text strategy (title/abstract/keyword) was used: ('rectum neuroendocrine tumor’/exp OR 'rectal neuroendocrine tumor’:ti,ab,kw OR 'rectal neuroendocrine tumour’:ti,ab,kw OR 'rectal neuroendocrine neoplasm’:ti,ab,kw OR ‘rectal carcinoid’:ti,ab,kw) AND ('transanal endoscopic microsurgery’/exp OR 'transanal endoscopic operation’:ti,ab,kw OR TEM:ti,ab,kw OR TEO:ti,ab,kw OR 'transanal minimally invasive surgery’:ti,ab,kw OR TAMIS:ti,ab,kw OR 'transanal endoscopic surgery’:ti,ab,kw OR TES:ti,ab,kw OR 'robotic transanal minimally invasive surgery’:ti,ab,kw).

No language or year restrictions were applied; bibliographies of relevant papers were screened manually.

Included were studies reporting outcomes of rigid-platform transanal approaches (TEM, TEO, TAMIS, robotic TAMIS) for rectal neuroendocrine tumours (rNETs), either as primary treatment or as completion after incomplete endoscopic excision. Excluded were case reports (< 5 patients), purely endoscopic resections (EMR, ESD, eFTR), radical rectal resections, and reviews without original data (Fig. 2).

Clinical decision-making algorithm for the management of rectal neuroendocrine tumours (NETs)

Two reviewers (CAA, GD) independently screened abstracts and full texts. Data were extracted into a structured form including: author, year, journal, country/setting, number of patients, indications (primary vs completion), tumour size, distance from anal verge, operative details, complications, margin status (R0), residual tumour at re-excision, follow-up duration, and recurrence/metastatic outcomes. Additional fields included tumour grade (according to the WHO classification) and the presence of lymphovascular invasion (LVI), when reported. Disagreements were resolved by discussion; if needed, a third author adjudicated.

Potential overlap between cohorts was assessed by centre, recruitment period, and author lists; when overlap was suspected, the most comprehensive or most recent dataset with the longest follow-up was retained to avoid double-counting.

Cohort and comparative studies were assessed using the ROBINS-I; case series were evaluated using the JBI checklist for case series. Results are presented descriptively given heterogeneity.

Due to heterogeneity, results were synthesised narratively with descriptive summaries of R0 resection, complications, residual tumour in completion excision, and oncologic events.

We performed a retrospective review of our prospective TEM/TEO database maintained at the Department of General Surgery, University of Turin, covering the period 1993 to 2025. This database includes all consecutive patients undergoing rigid-platform transanal endoscopic surgery (TEM, TEO) for rectal lesions. For this analysis, only patients with a final histopathological diagnosis of rectal neuroendocrine tumour (NET) were included.

This study was retrospective and based on routinely collected clinical data. All data were fully anonymised prior to analysis. According to applicable national regulations and institutional policy for retrospective studies using anonymised data, formal Ethics Committee approval and individual informed consent were not required.

Patients were eligible if they underwent primary local excision for a suspected NET or completion excision after an incomplete endoscopic removal (R1 or piecemeal EMR/ESD). Exclusion criteria were: (i) histology other than NET; (ii) NETs treated by radical rectal resection. For oncologic endpoints, we required a minimum follow-up of 6 months to allow at least one post-treatment assessment; perioperative outcomes were recorded for all eligible patients.

All patients underwent digital rectal examination and endoscopy. Tumour size and location (distance from anal verge) were recorded. Specifically, lesion diameter was recorded as a preoperative endoscopic estimate (diameter pre) and as the maximum diameter on the resected specimen (diameter histo); both are reported in millimetres (mm) after conversion from the original centimetre-based dataset (× 10). Staging evolved over decades; in recent years, this included endorectal ultrasound and pelvic MRI for lesions ≥ 10 mm or suspicious for muscularis propria invasion, in line with ENETS guidelines.

Procedures were performed with Buess’ TEM platform (Richard Wolf) until 2008 and with TEO (Karl Storz) thereafter. Both systems provide rigid access with CO₂ pneumorectum, magnified vision, and allow full-thickness surgical excision. Excision intent was a full-thickness, en bloc resection with a margin of healthy tissue surrounding the lesion. Defects were closed with running long-lasting absorbable sutures when feasible. Type of anaesthesia (general vs spinal), operative time, and intraoperative events (e.g., peritoneal opening, conversion) were recorded.

Specimens were oriented, pinned, and assessed by dedicated gastrointestinal pathologists. Tumour size, grade (WHO classification), depth of invasion, presence of lymphovascular invasion (LVI), and margin status were recorded. R0 resection was defined as negative lateral and deep margins.

Patients were monitored for complications during hospital stay. Complications were graded according to Clavien-Dindo classification.

Follow-up consisted of clinical examination, endoscopy, and imaging (pelvic MRI and/or endorectal ultrasound), scheduled according to ENETS guidelines and tailored to tumour size, grade, and LVI status. Local recurrence, nodal and distant metastases, and survival outcomes were documented.

The primary endpoint was the rate of R0 resection. Secondary endpoints included operative time, complications, hospital stay, recurrence/metastasis, and pathological risk factors (tumour size, grade, and lymphovascular invasion).