Aims
Controlling intraoperative bleeding is imperative for effective ESD. Flushing-assisted under-saline ESD (Flushing ESD) is a novel adjunct technique in saline immersion therapeutic endoscopy, designed to enhance bleeding prevention and improve haemostatic control. It provides concurrent, energy-synchronised flushing from the endoscopic knife, where effective vessel sealing and coagulation can be achieved with high radiofrequency forced coagulation. We aimed to compare haemostatic performance, bleeding events, and coagulation grasper utilisation between flushing and non-flushing rectal ESD in a UK tertiary referral centre.
Methods
We conducted a single-centre retrospective study comparing 10 rectal flushing with 5 non-flushing ESDs (3 CO₂ and 2 SITE). The Flushing Forced Method used an electrosurgical knife with integrated flushing (FlushKnife BTs, FUJIFILM Corp, Japan) connected to an endoscopic irrigation pump (EIP2, ERBE Corp, Germany). Videos were analysed to document visible vessels, minor bleeding (ooze), major bleeding (red-out), and coagulation grasper activations. Demographic, procedural and histological data were obtained from electronic records. Student’s t-tests were used for continuous variables, logistic regression for categorical variables, and Poisson regression for rate outcomes.
Results
Baseline demographics, lesion characteristics, and resection outcomes were comparable between groups. Flushing ESD demonstrated significantly higher oozing (rate ratio [RR] 1.55, 95%CI 1.09–2.23) when scaled per vessel, but not when scaled per lesion area. All oozing events were easily treated with further applications of Flushing Forced coagulation with the knife tip. It required markedly fewer coagulation grasper applications for active bleeding per vessel (RR 0.07, 95%CI 0.00–0.43, p=0.02) and per cm² (RR 0.04, 95%CI 0.00–0.25, p<0.01). Rates of major bleeding did not differ significantly.
| Flushing ESD (n=10) Median (IQR) | Non-Flushing ESD (n=5) Median (IQR) | Stat. | |
|---|---|---|---|
| Lesion (mm) | 55 (40-69.5) | 50 (35-50) | p=0.48 |
| Vessels/cm2 | 1.04 (0.6-1.7) | 1.27 (1.06-1.27) | RR:0.90,p=0.72 |
| Ooze/Vessel | 0.7 (0.6-1) | 0.5 (0.5-0.6) | RR:1.55,p=0.02* |
| Ooze/cm2 | 0.8 (0.6-1.2) | 0.6 (0.4-0.8) | RR:1.55, p=0.22 |
| Red Out/Vess. | 0.15 (0.11-0.26) | 0.06 (0-0.13) | RR:2.06, p=0.06 |
| Red Out/cm2 | 0.2 (0.11-0.23) | 0.05 (0-0.23) | RR:2.1, p=0.19 |
| Coag Grasp (Bleed)/Vessel | 0, 0 (0-0) | 0.03 (0.03-0.06) | RR:0.07,p=0.02* |
| Coag Grasp (Bleed)/cm2 | 0, 0 (0-0) | 0.07 (0.03-0.08) | RR:0.04,p<0.01* |
Conclusions
Although rates of major bleeding did not differ significantly between flushing and non-flushing ESD, Flushing ESD demonstrated improved haemostatic efficiency, with markedly fewer coagulation grasper activations. Minor bleeding was more frequent when normalised per vessel, but this did not translate into a greater clinical bleeding burden. All minor bleeding episodes were readily controlled with further Flushing Forced coagulation applications. Overall, Flushing ESD facilitated more controlled and efficient intraprocedural haemostasis for rectal ESD.