Subscribe to RSS
DOI: 10.1055/s-0038-1637226
PREOPERATIVE PREDICTIVE MODEL FOR EN BLOC COLORECTAL ESD (CR-ESD): A SPANISH PROSPECTIVE MULTICENTRE COHORT STUDY
Publication History
Publication Date:
27 March 2018 (online)
Aims:
To assess the preoperative factors that are able to predict en bloc CR-ESD.
Methods:
The demographic and clinical characteristics of the patients and lesions were prospectively collected. Morphological features of the tumors and technical factors were recorded.
Results:
We performed ESD in 209 lesions in 208 patients. The recruitment was done in 11 centres from January 2016 to June 2017. An en bloc resection was achieved in 165 lesions (78.9%). The procedure was aborted in 4 (1.9%). The remaining procedures were finished piecemeal (n = 40; 19.1%). In the multivariate analysis, after adjusting by age and sex, the independent predictors for en bloc resection are shown in the following table:
En bloc (n = 165) |
Aborted + Piecemeal (n = 44) |
OR (95% CI) |
p |
|
Experience > 10 cases |
153 (81.8) |
34 (18.2) |
5.8 (1.9 – 17.7) |
0.002 |
< 10 cases |
12 (54.5) |
10 (45.5) |
||
One endoscopist Yes |
160 (82.5) |
34 (17.5) |
12.2 (3.4 – 43.9) |
< 0.0001 |
No |
5 (33.3) |
10 (66.7) |
||
Size < 5 cm Yes |
148 (84.6) |
27 (15.4) |
9.2 (3.5 – 23.9) |
< 0.0001 |
No |
17 (50) |
17 (50) |
||
Absence of prior diathermy Yes |
156 (81.3) |
36 (18.8) |
5.1 (1.5 – 17.6) |
0.001 |
No |
9 (52.9) |
8 (47.1) |
The value of the area under the ROC curve of the predictive model for en bloc resection [logodds =- 6.05 + 2.22 x (size ≤5 cm) + 1.63 x (absence of prior diathermic injury) + 2.51 x (one endoscopist) + 1.76 x (centre case load > 10)] was 0.81 (95% CI: 0.74 – 0.88).
Conclusions:
A preoperative predictive model combining size, prior attempts of resection and the experience in ESD provided good diagnostic performance to predict a successful en bloc ESD.
#