Introduction

Globally, colorectal cancer (CRC) is the third most commonly diagnosed malignancy and the second leading cause of cancer death [1]. Morbidity and mortality associated with CRC can be mitigated through appropriate screening and surveillance. Colonoscopy, as a screening procedure, is a gold standard in detecting tumors at an earlier and more treatable stage and also facilitates the timely removal of precancerous lesions or adenomas [2]. The adenoma detection rate (ADR) is now the main quality indicator of colonoscopy because of its inverse correlation with interval cancer rate [3, 4]. ADR can be improved by technique or devices that improve mucosal exposure or by tools that highlight flat colonic lesions.

A number of distal attachments have been tested to improve ADR, including a transparent cap, cuff, or rings [5]. The cuff is attached to the tip of the colonoscope, and the fingers are used to flatten colonic folds, leading to increased mucosal visualization [6]. Endocuff (Arc Medical, Leeds, UK), which was granted United States Food and Drug Administration approval in 2012, is a soft plastic cap with rows of finger-like projections attached onto the colonoscope tip. In 2014, the next-generation Endocuff (Endocuff Vision) was released, consisting of a single row of finger-like projections which were 3 mm longer [7].

A number of studies have been published comparing the efficacy of Endocuff-assisted colonoscopy (EAC) to that of standard colonoscopy, but so far, the impact of EAC on ADR is conflicting with data suggesting equivocal benefit from its use. Aside from individual studies, several meta-analyses have been performed to assess the effect of Endocuff [8,9,10,11,12]. Fewer randomized controlled trials (RCTs) were included in the early meta-analyses, and the high heterogeneity noted called for careful interpretation of the results. The most recent meta-analysis [12], including 8 RCTs, found a significant improvement in ADR and mean number of adenomas per colonoscopy with shorter withdrawal times using the second-generation cuff device compared with standard colonoscopy. Six RCTs comparing Endocuff-assisted and standard colonoscopy have recently been published [13,14,15,16,17,18]. Therefore, we conducted a meta-analysis of published data to evaluate the efficacy and safety of EAC.

Materials and methods

This systematic review and meta-analysis have been registered at NIHR PROSPERO (CRD42021231865). It is reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [19].

Search strategy

A literature search was conducted using PubMed, Embase, Google Scholar and Cochrane Library up to June 8, 2021 without language restrictions. Relevant studies were identified using the following terms: “Endocuff”, “Endocuff vision”, “distal attachment” and “adenoma detection rate or ADR”. The search was restricted to human subjects. Additional studies were identified using a hand search of references of original or review articles and international conferences on this topic, primarily including United European Gastroenterology Week (UEGW) and Digestive Disease Week (DDW).

Inclusion and exclusion criteria

Studies were included if they met the following criteria: (1) RCTs that compared Endocuff-assisted and standard colonoscopy for adenoma detection, (2) presenting the detailed outcomes of Endocuff-assisted and standard colonoscopy or including such data for calculation in the article. Non-randomized prospective, retrospective, feasibility or pilot studies, meta-analysis, editorials, reviews, case reports/series, studies not reporting on ADR and duplicate publications were excluded.

Data extraction

Two investigators (Wang J, Ye C) independently extracted the data and reached a consensus for all items. If the investigators generated different results, they checked the data again and had a discussion to reach an agreement. If they were unable to reach an agreement, an expert (Fei S) was invited to join the discussion. Data extracted from the selected articles included the first author’s name, year of publication, country of origin, study period, device type, indications for colonoscopy, baseline characteristic of the patients, and primary outcomes.

Bias assessment

The risk of bias in individual studies was assessed using the Cochrane Collaboration tool [20]. This particular tool evaluates different domains of potential source of bias: random sequence generation and allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias) and other bias. The analysis results were defined as low, high or unclear bias (bias-related information is not clear or bias cannot be determined). Publication bias was assessed using subjective judgment based on funnel plots as well quantitatively using Egger’s regression analysis. A p value < 0.05 was indicative of substantial publication bias.

Outcome measures

The primary outcome of this study was to calculate a pooled ADR. The secondary outcomes were the polyp detection rate, detection rate of advanced adenomas, sessile serrated lesion detection rate, mean number of adenomas per patient, mean number of polyps per patient right-sided lesion detection rate, left-sided lesion detection rate, ileum intubation rate, cecal intubation rate, cecal intubation time, withdrawal time, and adverse events.

Statistical analysis

A meta-analysis was performed using the Cochrane Collaboration RevMan 5.4 and STATA package version 12.0. The analyses were performed by calculating pooled estimates of primary and secondary outcome. Relative risk (RR) with 95% confidence interval (CI) for each proportional outcome was calculated. Mean difference (MD) with 95% CI was calculated for continuous variables. The χ2-test-based Q statistic test was performed to assess the between-study heterogeneity. We also quantified the effect heterogeneity according to the I2 test. When a significant Q test (P < 0.05) or I2 > 50% indicated heterogeneity across studies, the random effects model was used; otherwise, the fixed effects model was used. An analysis of sensitivity was performed to evaluate the stability of the results. Additionally, we conducted subgroup analysis by the ADR for standard colonoscopy groups (baseline ADR), the device type (Endocuff or Endocuff Vision), adenoma size (≤ 5 mm, 6–9 mm, ≥ 10 mm) and indications for endoscopy (screening or mixed population). A p value of < 0.05 was regarded as being statistically significant.

Results

Study characteristics

Following the searching strategy, a total of 546 citations were identified. According to the inclusion criteria, 25 studies were selected and subjected to further examination [13,14,15,16,17,18, 21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39]. We excluded 2 studies [38, 39] because they are randomized tandem studies. Therefore, 23 RCTs with 17,999 patients were included in the meta-analysis [13,14,15,16,17,18, 21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37]. (Fig. 1). The characteristics of the selected studies are summarized in Table 1. Of the 23 eligible studies, 5 were from the United States [14, 23, 24, 28, 33], 3 each from United Kingdom [15, 21, 31] and Germany [22, 26, 36], 2 from mixed countries [13, 32], 2 from Thailand [17, 35], and 1 each from Spain [16], Italy [18], Portugal [25], Mexico [27], Australia [29], France [30], Netherlands [34], and Japan [37]; Nine studies were multicenter [15, 16, 18, 23, 24, 26, 31, 32, 34], 4 were two-center [13, 22, 33, 37] and 10 were single-center experiences [14, 17, 21, 25, 27,28,29,30, 35, 36]. All studies were published in English (18 full-text articles [13,14,15,16,17,18, 21, 22, 26, 27, 29,30,31,32,33,34, 36, 37] and 5 abstracts [23,24,25, 28, 35]). Eleven studies used the first-generation Endocuff [13, 14, 22,23,24, 26,27,28, 32, 34, 37], and 12 studies used the second-generation device (Endocuff Vision) [15,16,17,18, 21, 25, 29,30,31, 33, 35, 36].

Fig. 1
figure 1

PRISMA flow chart showing study selection procedure. PRISMA preferred reporting items for systematic reviews and meta-analyses, EAC endocuff-assisted colonoscopy, RCT randomized controlled trial

Full size image
Table 1 Characteristics of the included studies
Full size table

Risk of bias assessment using the Cochrane Collaboration tool is provided in supplementary table 1 and supplementary Fig. 1 (Fig.S1). The shape of the funnel plots did not reveal any evidence of asymmetry (Fig. S2). Egger’s test also showed no statistical significance in evaluation of publication bias (p = 0.427).

Quantitative data synthesis

Primary outcome

Adenoma detection rate

All 23 studies reported ADR in the Endocuff and standard colonoscopy groups. The pooled ADR was 44.9% (95% CI 37.6–52.1) for EAC and 39.1% (95% CI 32.3–45.9) for standard colonoscopy. The pooled RR was 1.16 (95% CI 1.08–1.24, p < 0.00001) (Fig. 2; Table 2).

Fig. 2
figure 2

Forest plots comparing endoscopic-assisted and standard colonoscopy in terms of ADR. ADR adenoma detection rate

Full size image
Table 2 Outcomes of meta-analysis comparing EAC and SC
Full size table

In subgroup analyses based on the ADR in the standard colonoscopy group, a significant difference was observed between Endocuff and standard colonoscopy groups (baseline ADR < 25%: RR = 1.47, 95% CI 1.13–1.93, p = 0.004; baseline ADR < 30%: RR = 1.39, 95% CI 1.21–1.58, p < 0.00001; baseline ADR < 35%: RR = 1.40, 95% CI 1.25–1.58, p = 0.03; baseline ADR < 40%: RR = 1.37, 95% CI 1.23–1.52, p = 0.004; baseline ADR < 45%: RR = 1.28, 95% CI 1.15–1.41, p < 0.0001; baseline ADR < 50%: RR = 1.24, 95% CI 1.13–1.36, p < 0.0001). In contrast, no statistical difference was found in the subgroup of baseline ADR > 50% (RR = 1.04, 95% CI 0.97–1.11, p = 0.50) (Table 2).

Eight studies reported adenoma size. The pooled results showed ADR did not differ between Endocuff and standard colonoscopy (size ≥ 10 mm: RR = 1.02, 95% CI 0.91–1.15, p = 0.47; 6–9 mm: RR = 1.10, 95% CI 0.96–1.27, p = 0.29; size ≤ 5 mm: RR = 1.03, 95% CI 0.95–1.11, p = 0.05) (Table 2).

In the subgroup analysis of device type, a significant difference was found in both subgroups (Endocuff: RR = 1.22, 95% CI 1.07–1.40, p < 0.00001; Endocuff Vision: RR = 1.12, 95% CI 1.05–1.20, p = 0.11) (Table 2).

When grouping by indications for colonoscopy (pure screening or mixed populations), a significant difference was found in both subgroups (pure screening: RR = 1.20, 95% CI 1.06–1.37, p = 0.001; mixed population: RR = 1.14, 95% CI 1.05–1.23, p = 0.0007) (Table 2).

Secondary outcome

Polyp detection rate

13 studies with 11,421 patients reported polyp detection rates. The pooled polyp detection rate in the Endocuff group was 54.5% (95% CI 44.6–64.4) and in the standard colonoscopy group was 46.5% (95% CI 37.2–55.9). The pooled RR was 1.17 (95% CI 1.09–1.25, p = 0.0008) (Fig. 3; Table 2).

Fig. 3
figure 3

Forest plots comparing endoscopic-assisted and standard colonoscopy in terms of second outcomes. A Polyp detection rate; B Serrated lesion detection rate, C left-sided lesion detection rate

Full size image
Sessile serrated lesion detection rate

Ten studies with 9914 patients reported Serrated polyp detection rates. The pooled serrated detection rate was 8.4% (95% CI 5.8–11.1) for Endocuff and 5.9% (95% CI 4.0–7.8) for standard colonoscopy. The pooled RR was 1.23 (95% CI 1.05–1.43, p = 0.46) (Fig. 3; Table 2).

Detection rate of advanced adenomas

Seven studies with 9243 patients reported Advanced ADR. The pooled Advanced ADR was 13.7% (95% CI 8.0–19.4) for Endocuff and 12.7% (95% CI 7.3–18.1) for standard colonoscopy. The pooled RR was 1.11 (95% CI 1.00–1.23, p = 0.45) (Fig. S3; Table 2).

Right and left-side lesion detection rate

The right- and left-sided lesion detection rates were reported in 9 and 6 studies, respectively. The pooled right-sided lesion detection rate was 28.7% (95% CI 23.3–34) for Endocuff and 25.2% (95% CI 19.7–30.7) for standard colonoscopy. The pooled RR was 1.21 (95% CI 1.00–1.46, p < 0.00001) (Table 2). The pooled left-sided lesion detection rate was 30.5% (95% CI 22.7–38.4) for Endocuff and 25.5% (95% CI 17.6–33.4) for standard colonoscopy. The pooled RR was 1.24 (95% CI 1.08–1.43, p = 0.08) (Fig. 3; Table 2).

Mean number of adenomas per patient

Ten studies with 10,178 patients reported the available data between Endocuff and standard colonoscopy and the pooled mean difference with a random-effect model were 0.17 (95% CI 0.08–0.26, p < 0.00001) (Fig. 4; Table 2). The results above indicate a significant difference between the two groups.

Fig. 4
figure 4

Forest plots comparing endoscopic-assisted and standard colonoscopy in terms of MAP. MAP mean number of adenomas per patient

Full size image

Adverse events

Sixteen studies reported the available data between Endocuff and standard colonoscopy. The most commonly reported complication was minor mucosal lacerations, which was reported in 66 patients in Endocuff group (4%), and in 7 patients in Endocuff Vision group (0.1%). The other complication was loss of Endocuff (0.4%), post-polypectomy bleeding (0.3%), bleeding (0.06%), and perforations (0.04%). The pooled RR with a random effects model was 2.60 (95% CI: 1.29–5.26, p = 0.01) (Table 2). Stratification based on device type, similar results were observed in the Endocuff group (77/1536, RR = 7.16, 95% CI 3.82–13.41, p = 0.09), but not in the Endocuff Vision group (31/5241, RR = 1.31, 95% CI 0.78–2.20, p = 0.50) (Fig. S4).

Other outcomes

For additional secondary outcomes, including ileum intubation rate, MPP, cecal intubation rate, cecal intubation time, and withdrawal time, there was no significant difference between the two groups except for ileum intubation rate (61% versus 68%, RR = 0.89, 95% CI 0.80–0.99, p = 0.0001) (Fig. S5; Table 2).

Discussion

In this study, we included 23 studies involving 17,999 patients to assess the efficacy and safety of the EAC for ADR. The results showed that the EAC was superior to standard colonoscopy in terms of adenoma, polyp, sessile serrated, and left and right-sided lesion detection rates as well as mean number of adenomas per patient. No significant difference was found in advanced ADR, mean number of polyps per patient, right-sided lesion detection rates, cecal intubation rate, cecal intubation time and withdrawal time between EAC and standard colonoscopy.

Colonoscopy with adenoma detection and removal is widely considered the gold standard for the prevention of CRC. Every 1% increase in ADR is associated with a 3% reduction of interval CRC and 5% reduction of fatal interval CRC. Moreover, there is an increased risk of interval cancer when the colonoscopy is performed by an endoscopist with an ADR below 20% [3]. Recent years have seen the rapid introduction of a range of new mechanical and optical endoscopic devices aimed at improving the ADR and minimizing miss rates [40,41,42,43].

Due to the use of different devices, we further analyzed the efficacy of Endocuff or Endocuff Vision separately, and the pooled result indicated significant differences in both groups, which was in accordance with a previous study [12]. Moreover, the pooled ADR was slightly higher in the Endocuff Vision group (46.3%) compared with the Endocuff group (43.3%). However, a recent meta-analysis conducted by Aziz et al. [44], evaluated colonoscopy outcomes among Endocuff Vision, Endocuff and high-definition colonoscopy groups, and reported that the Endocuff Vision did not significantly improve ADR compared to Endocuff and high-definition colonoscopy. Because of the different expertise of endoscopists, we conducted subgroup analysis for ADR based on ADR in standard colonoscopy group (< 25%, < 30%, < 35%, < 40%, < 45%, < 50% and > 50%). The results showed that operators with baseline ADR < 50% benefit from the use of EAC, whereas the very high baseline detectors (ADR > 50%) did not. Hence, it is clearly suggested that the expertise of the endoscopist is inversely correlated with the benefit of Endocuff in terms of ADR vs. standard colonoscopy.

In addition, we performed subgroup analysis for ADR based on indications for colonoscopy, which indicated significant differences in both groups. A higher ADR was observed in mixed population (47.4% vs 38.7% in screening group). More studies including pure screening population are needed to further confirm the effect of population on ADR. As for colorectal adenomas size, up to 15.5% of small adenomas and up to 3.4% of diminutive adenomas contain high-grade dysplasia, and the omission of those adenomas may contribute to the occurrence of interval cancers diagnosed between surveillance colonoscopies [45]. In this study, we also conducted subgroup analysis based on adenomas size and no significant difference was found. However, because only a few studies were included in the above analysis, the result should be interpreted with caution, and more studies are needed. With regard to adverse events, the results showed that the rate in the first-generation Endocuff, rather than the second-generation Endocuff Vision, was higher than in standard colonoscopy. The revised design of the Endocuff, with the removal of the distal row of arms and the creation of more rounded tips, might explain the absence of the adverse events.

Some limitations of this meta-analysis should be addressed. First, the quality of bowel preparation was reported incompletely using different scoring criteria in the included studies. Second, the endoscopists in both groups were not blinded, which is common to most endoscopic studies designed for assessment of external attachments. Third, we did not perform a comparative cost-effectiveness analysis. Ideally, adoption of interventions in clinical practice would be premised on incremental cost with each intervention per additional adenoma detected.

Conclusions

This meta-analysis showed that a significant improvement in adenoma and polyp detection rates as well as mean number of adenomas per patient using EAC compared with standard colonoscopy, especially for operators with a low ADR. Further studies are needed to confirm the value of Endocuff in improving the ADR.