The first generation Endocuff
Endocuff (Arc Medical Design, Leeds, United Kingdom) is a single-use soft, radiopaque device that consists of a cylindrical polypropylene core and 2 rows of flexible thermoplastic elastomer-made projections. Each row counts 8 projections that emerge from gaps on the shaft of the device (Figure 2A and B). It is available in 4 different color-coded sizes to fit all scopes and its technical characteristics are presented in Table 1. Its designers were inspired through the practical difficulties that occur during a conventional colonoscopy, including the scope slipping back, difficulties in tip stabilization and inability to inspect the mucosa located behind folds, to mention a few. Endocuff was launched in 2012 and its use was reported for the first time in a small retrospective feasibility study where it facilitated endoscopic access for complex polypectomy and scar assessment in the sigmoid colon. The projections move independently from another in a passive way when in contact with the mucosa and during withdrawal, they extend radially manipulating colonic folds away from the field of view, allowing a more meticulous mucosa inspection (Figure 2C). Moreover, the device stabilizes the scope in the middle of the lumen and allows traction against sudden slippage around flexures. Moreover, the examiner’s visibility is not affected, since the device does not extend beyond the tip of the scope and thus does not interfere with suction, flushing or the working channel.
Figure 2 Endocuff (A) mounted on the tip of the scope (B) and the endoscopic view of the hinged projections during the withdrawal phase (C) (photos from the authors’ archive).
There are enough data regarding the effect of Endocuff on colonoscopy outcomes, since seven RCTs of parallel[37-43] and one of tandem design have been published. The first German studies[37,38]-each recruiting almost 500 patients who underwent colonoscopy for various indications (screening included)-showed a significant benefit of Endocuff-assisted colonoscopy (EAC) compared to the conventional one regarding ADR (35.4% vs 20.7%, P < 0.0001 and 36% vs 28%, P = 0.043, respectively)[37,38]. Similar results regarding PDR were also achieved (55.4% vs 38.4%, P < 0.0001 and 56% vs 42%, P = 0.001, respectively), while only the second study detected a difference in the mean number of adenomas detected per colonoscopy [2 (IQR: 1-3) vs 1 (IQR: 1-2), P = 0.002]. Regarding polyp location, both studies identified a superiority of EAC for the detection of polyps located in the sigmoid and the cecum. Moreover, no major adverse events related to EAC were reported and there were no differences between overall procedure and withdrawal times[37,38].
Similar positive results associated to Endocuff use were reported from Japan (477 patients, mixed indications for colonoscopy) and Mexico (337 screening individuals), where two single-centre RCTs demonstrated increase of ADR (55.2% vs 39.2%, P = 0.0002 and 22.4% vs 13.5%, P = 0.02, respectively), PDR (61.9% vs 49.2%, P = 0.003 and 29.9% vs 16%, P = 0.002, respectively) and MAC (1.11 vs 0.66, P < 0.01 and 0.29 vs 0.22, P = 0.04) in the device arms. An Italian single-centre study by De Palma et al enrolled 288 patients with mixed indications and reported that EAC increased ADR by 3.3% (29.6% vs 26.3%) compared to the conventional colonoscopy. However, use of Endocuff was associated with mucosal erosions in 7 (2.5%) cases, with one of them needing to be treated with adrenaline solution injection at the site of bleeding.
Additionally, in a recently published 4-arm multicenter parallel-group study comparing Endocuff, Endorings, FUSE and conventional colonoscopy, 299 and 295 patients underwent Endocuff-assisted and conventional high definition colonoscopy, respectively. EAC performed significantly better compared to conventional colonoscopy in terms of ADR (64% vs 56%, P = 0.003), PDR (83% vs 77%, P = 0.001) and MAC (1.82 ± 2.58 vs 1.53 ± 2.33, P = 0.014). However, Endocuff did not enhance the detection rate of sessile serrated polyps (11% vs 12%, P = 0.047). There were no differences between the mean insertion time (354 s ± 216 s vs 422 s ± 319 s) for Endocuff-assisted and conventional colonoscopy respectively and no adverse events were reported. On the contrary, a benefit regarding sessile serrated adenoma/polyp detection was shown in a retrospective veterans’ study which included almost 500 participants: Endocuff detected 50 sessile serrated adenomas/polyps compared to 8 detected by conventional colonoscopy (detection rate 15% vs 3%, P < 0.0001).
So far, the largest parallel RCT failed to confirm the positive results reported in the abovementioned studies. In this multicentre study from the Netherlands more than 1000 patients of various indications were randomized to undergo either Endocuff-assisted or conventional colonoscopy; MAC and ADR consisted the primary outcomes. ADR was the same in both groups (52%, P = 0.92), whereas the higher number of adenomas per patient in the Endocuff group (1.36 ± 2.10 vs 1.17 ± 1.65) did not reach statistical significance (P = 0.08). Interestingly, detection rates did not differ either according to indication or between academic and non-academic centres. Cecal intubation time was significantly shorter in the Endocuff arm [median (IQR) 7 min (5-10) vs 8.3 min (6-12), P < 0.001] and there were no Endocuff-associated adverse events.
Finally, a multicentre back-to-back study assessed Endocuff in terms of adenoma miss rates. Two hundred patients (86.5% were screening and surveillance cases) were randomized (1:1) to undergo either initial EAC followed by a conventional one or vice versa. EAC was associated with lower adenoma miss rates, both overall and in the proximal colon compared to conventional colonoscopy (14.7% vs 38.4% and 10.4% vs 38.9%, respectively). It is worthy to note that all examinations were performed by endoscopists with an historical ADR > 35%, suggesting that the device could enhance detection ability even of experienced and skilled endoscopists. Despite the fact that there were no serious adverse events, in three index Endocuff examinations cecal intubation failed to be achieved, compared to none with the conventional scope (P = 0.08).
The Endocuff Vision
Despite its revolutionary design, Endocuff was associated with a couple of drawbacks (mucosal erosions and difficulties in terminal ileum intubation) that paved to way for its descendant, namely Endocuff Vision (Norgine Pharmaceuticals Ltd, Uxbridge, United Kingdom). This single-use device is made of a polypropylene cylinder and a single row of 8-longer than in the first generation Endocuff-thermoplastic elastomer-made “spikes” (Figure 3). There are 4 different sizes with respective colors to fit in all scopes ranging from pediatric to adult ones (Table 1). Endocuff Vision is also mounted onto the tip of the scope before insertion and its “spikes” fold around the scope while it advances in the colon due to a hinge at the base of each spike that thins progressively. On the other hand, the “spikes” evert during withdrawal (Figure 3). This leads to an early and controlled view of the upstream surface of the large colonic folds in the right colon and prevents sudden scope slip-back. Moreover, when in the sigmoid colon, the device facilitates the opening of contracted folds, permitting a clearer view of the in-between mucosa. Similar to the first generation Endocuff it optimizes the tip’s position during endoscopically applied therapy (e.g., polypectomy).
Figure 3 Endocuff-Vision (A), illustration (B) and endoscopic view (C) of the opened-out projections during the withdrawal phase (photos from the authors’ archive).
Endocuff Vision has been evaluated only in two parallel multicenter RCTs from the United Kingdom[46,47]. The “ADENOMA” study recruited 1772 adult patients (45% screening). Of them, 884 underwent conventional colonoscopy and 886 Endocuff Vision-assisted colonoscopy. ADR was significantly higher with EAC compared to conventional colonoscopy (40.9% vs 36.2%, P = 0.02). The benefit of Endocuff Vision was even higher in patients participating in the screening program, where ADR was 61.5% for EAC compared to 50.9% (P < 0.001) for the conventional colonoscopy arm. Similar results in favor of EAC were also reported regarding PDR (54.1% vs 48%; P = 0.005 and 73.9% vs 63.3%; P < 0.001 for the whole and the screening cohorts, respectively). Of note, EAC showed a statistically significant increase in the detection rate in the left colon (26.1% vs 2.2%; P = 0.03), of small (10.6% vs 7.7%; P = 0.02) and of diminutive adenomas (34.6% vs 30.8%; P = 0.04). It should be underlined that in this study EAC detected significantly more cancers both in the whole cohort (4.1% vs 2.3%; P = 0.02) as well as in the screening participants (6.6% vs 3.7%; P = 0.03). Moreover, median insertion time was shorter with Endocuff Vision compared to conventional colonoscopy (8 min vs 9 min; P = 0.001). The investigators did not report any adverse event related to use of Endocuff Vision; however the device had to be removed in 4.1% of the cases mostly due to acute angulation in a fixed sigmoid colon.
On the contrary, the “E-cap” study failed to show any benefit in terms of ADR (60.9% vs 63%, P = 0.85), PDR (70.3% vs 69.8%, P = 0.93) and MAC (1.3 ± 1.8 vs 1.4 ± 1.5, P = 0.54). This single center study had PDR as the primary endpoint. Only patients attending the national screening program with a positive FOBT test were enrolled and all four participating endoscopists had an extremely high pre-study ADR (58.9%). All these reasons may attribute to the lack of any significant benefit deriving from application of Endocuff Vision and should be considered in the design of future “real-life” studies, which should possibly include both endoscopists with an average or even a low ADR and patients with various indications for colonoscopy.
Finally, a pilot evaluation study demonstrated that Endocuff Vision was associated with an improvement in endoscopists’ performance measured as increased ADR, increased MAC and decreased insertion time. In this non-randomized study, the investigators performed 410 screening colonoscopies in three periods (137 pre-Endocuff, 136 using Endocuff Vision and 137 post-Endocuff). Overall, an increase in ADR (16%, P < 0.03) and MAC (83%, P = 0.007) was noted between the pre-Endocuff and the Endocuff period; this benefit was maintained in the post-Endocuff period, where the device was not available. A potential explanation could be that during the Endocuff period the endoscopists had the chance to comprehend their flaws during the withdrawal phase, look for adenomas in more detail and improve their skills. Interestingly, insertion time was statistically lower during the Endocuff period compared to pre- and post-Endocuff one (7 min vs 8 min, P = 0.002 and 7 min vs 9 min, P = 0.002, respectively); no adverse events were reported.
Data from meta-analyses
To date, three meta-analyses attempting to summarize the impact of Endocuff devices on colonoscopy outcomes have already been published[35,50,51] (Table 2).
The earliest one meta-analyzed data from three published papers and six studies presented as abstracts, four of which with a prospective and five with a retrospective design. Eight studies (n = 4387) of mixed populations reported on ADR, which was measured to be higher for the Endocuff group [OR (95%CI): 1.49 (1.23-1.80), I2 = 50%]. In this pooled analysis, 27 patients (2.3%) in the Endocuff group experienced superficial mucosal lacerations.
A recently published meta-analysis updated these data by including only RCTs (7 published and 5 presented as abstracts). Regarding ADR, data from more than 8370 patients demonstrated a benefit of EAC compared to conventional colonoscopy [RR (95%CI): 1.20 (1.06-1.36), P = 0.003, I2 = 79%]. Of interest, this benefit was lower in the subgroup of studies with a mean conventional arm ADR > 45% [RR (95%CI): 1.01 (0.93-1.09), P = 0.087, I2 = 0], while it was maximized in the subgroup of studies with a respective ADR lower than 35% [RR (95%CI): 1.51 (1.35-1.69), P < 0.001, I2 = 43%]. These data imply a potential ancillary role of the device especially for lower detectors. Furthermore, a numerical higher MAC was detected in the Endocuff-assisted colonoscopy group, but this difference did not reach statistical significance [mean difference (95%CI): 0.11(-0.17 to 0.38]. Mean insertion times did not differ between the two groups and 4% of the Endocuff patients experienced adverse events (exclusively minor lacerations). This meta-analysis reported on additional outcomes such as advanced ADR and right colon ADR, with no difference detected between the two groups [RR (95%CI): 0.93 (0.76-1.13), P = 0.47 and RR (95%CI): 1.36 (0.80-2.34), P = 0.26, respectively]. However, the small number of studies included in the analysis regarding these outcomes warrants caution when attempting to generalize the respective results.
Finally, similar results were shown in a network meta-analysis investigating the comparative efficacy of distal attachments in increasing detection rates during colonoscopy. The mixed effect estimate (including both pairwise and indirect treatment effects) supported that ADR increased significantly with EAC compared to the conventional examination [RR (95%CI): 1.21 (1.03-1.41)]. Interestingly and contrary to the meta-analysis from Williet et al this network meta-analysis calculated a very modest benefit of Endocuff regarding low (baseline ADR 10%) detectors [anticipated ADR (95%CI): 11 (10-12)%] compared to a more considerable effect [anticipated ADR (95%CI): 48 (14-56)%] on ADR of high detectors (baseline ADR 40%).