PTH-074 Smsa score not an independent predictor of outcomes in a large series of endoscopic resections

    loading  Checking for direct PDF access through Ovid



A predictive score to stratify the difficulty of endoscopic resection (ER) of large colorectal superficial neoplastic lesions (CSNL) and predict outcomes would be valuable. The SMSA score assigns a numerical score based on size, morphology, site and ease of access and stratifies lesions into 4 groups of increasing complexity. It has been recommended in several guidelines on colorectal ER and SMSA level 4 is reportedly associated with incomplete resection and increased complications. Despite this, it has not been widely validated in large series of ER for large CSNLs using standardised techniques. We applied the SMSA score to a large series of ER of lesions≥2 cm at a tertiary centre.


ER of large (≥2 cm) CSNL were included. Surveillance colonoscopy was performed at 3 months and 12 months. SMSA score was calculated and SMSA level 4 compared with level 2–3 for outcomes including failed ER,≥2 ER to achieve clearance, complications and recurrence. Multivariate logistic regression was performed to determine independent predictors of complications and recurrence.


ER was performed for 542 lesions (mean size 53.7 mm). SMSA level 4 was not associated with an increased risk of failed ER (OR 0.06, 95% CI 0.007–0.44, p<0.001) or requirement of ≥2 ER to achieve clearance (OR 3.96, 95% CI 0.49–31.9, p=0.16). Although SMSA level 4 was associated with complications (OR 5.25, 95% CI 1.58–17.5, p=0.002), it was not associated with perforation and a simple assessment of either large lesion size (≥50 mm) or difficult access was similarly associated with complications (OR 4.27, 95% CI 1.47–12.4, p=0.004). Multivariate logistic regression revealed only age (p<0.001), prior heavy manipulation (p=0.03), and ESD/Hybrid ESD (p=0.02) were independently associated with complications but not SMSA level 4 (p=0.06).


SMSA level 4 was associated with recurrence (OR 10.8, 95% CI 2.59–45.2, p<0.001), as was either large lesion size or difficult access (OR 6.35, 95% CI 2.47–16.33, p<0.001). On multivariate logistic regression only large size/difficult access (p=0.03) and piecemeal resection (p=0.003) were associated with recurrence but not SMSA level 4.


SMSA level 4 was not associated with failed ER or multiple ERs to achieve clearance. It was not an independent predictor of complications or recurrence on multivariate logistic regression. Although relatively simple, the SMSA score is somewhat difficult to rapidly calculate and stratify lesions. For practical purposes when making a judgement during a procedure, it may be as useful to use a simple assessment of large lesion size or difficulty in accessing the lesion.

Related Topics

    loading  Loading Related Articles