Introduction
Colorectal cancer (CRC) is the third most common type of cancer worldwide and is considered curable in its early stages [1]. Colonoscopy is an endoscopic technique [2] with high percentages of CRC detection; however, it remains imperfect, since it entails a percentage adenoma miss rate, considered to be the key point in the detection of precancerous lesions. The reasons standard colonoscopy (SC) is subject to failure vary from poor bowel preparation to limited visualization of haustral folds and flexures. Zhao et al demonstrated in a meta-analysis that 26% of adenomas are not detected during SC [3].
The introduction of various add-on devices, attached to the tip of the endoscope, helps unfold the lumen while also providing the operator with a better view of the epithelium. These devices consist of various single-use components, such as cylinders, rings and wings, and aid the unfolding of the mucosa. Since they have an easier learning curve, especially for novice colonoscopists, add-on devices tend to be widely utilized in colonoscopies [4,5]. The first generation of Endocuff, and its successor Endocuff-Vision, are single-use devices mounted on the tip of the scope, consisting of a cylindrical core and 1 or 2 rows of flexible projections [6]. Likewise, Endoring (EndoAid Ltd., Caesarea, Israel) is another device that flattens the intestinal folds, incorporating a 2-layer silicon ring [7], while WingCap (A&A Medical Supply LLC, Seongnam, South Korea) is manufactured in 2 layers, each consisting of 6 wings [8]. However, there is lack of sufficient evidence regarding the prevalence and efficacy of add-on devices exclusively in the setting of screening colonoscopy.
In this context, we aimed to accumulate data on add-on devices with projections and compare their performance to SC, by assessing adenoma detection rate (ADR) as a primary outcome, and polyp detection rate (PDR), advanced ADR (AADR), and sessile serrated lesion detection rate (SSLDR) as secondary outcomes.
Materials and methods
This review was based on the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) 2020 guidelines [9] (Supplementary Table 1), and a predefined protocol was registered at the International Prospective Register of Systematic Reviews (PROSPERO), under registration number CRD42022363186.
Eligibility criteria
The main question was based on the validated PICO (population, intervention, control, and outcomes) framework for systematic reviews and included the comparison between add-on devices with projections and SC with regard to ADR [10]. Only randomized controlled trials (RCTs) were assessed for eligibility, when: (A) patients: adult patients underwent screening colonoscopy for CRC, without symptoms; (B) interventions: screening colonoscopy using add-on devices with projections, including Endocuff, Endocuff- Vision, Endoring and Wingcap; (C) comparators: subjects underwent screening colonoscopy with conventional endoscopes and without assisting techniques (e.g., artificial intelligence); and (D) outcomes: studies not written in the English language, missing substantial data for analysis, nonrandomized prospective or retrospective studies, reviews, editorials, case reports, case series, narrative reviews, and conference abstracts were excluded. Studies not including ADR as outcome were also excluded.
Definitions
ADR is defined as the number of colonoscopies with adenomas detected divided by the total number of colonoscopies, multiplied by 100. PDR is defined as the number of colonoscopies with polyps detected divided by the total number of colonoscopies, multiplied by 100. The other studied variables, AADR and SSLDR were calculated similarly.
Search strategy
Between September and November, 2022, 2 investigators (MM and GT) conducted a detailed literature search across the Medline (PubMed) database and Cochrane Central Register of Clinical Trials Studies using the terms “add-on-device”, “endocuff”, “endoring”, “g-eye”, “amplifeye” and “adenoma detection rate”, as medical subject heading (MeSH) and free-text terms. These results were combined using the Boolean set operator “AND” with the term “screening colonoscopy” as a MeSH and free-text term. The initial electronic search was followed by a manual search of references from retrieved studies to identify additional suitable bibliography. All retrieved articles were screened for eligibility, first by 1 reviewer (MM) and afterwards by 2 independent senior authors (PG and GT), using the predetermined inclusion criteria. Initially, the titles and abstracts of all results were reviewed; the full-text content of eligible studies was obtained and reassessed independently for eligibility. For studies with missing or unavailable data, we attempted to contact the corresponding author to provide the missing information. In cases of multiple publications from the same study, only the most recent and complete article was included. Additionally, when both parallel design and crossover arm trials were found, only the parallel group was studied.
Data collection process
All eligible studies were reviewed by 2 investigators and relative data were extracted in a predefined extraction form. Through this process, any discrepancy was resolved either by consensus or following the senior authors’ (PG and GT) judgment.
Data items
Data on study-, participant- and intervention-related parameters were retrieved into a standardized form by 2 investigators (MM and GT), and a third author (AP) checked the 2 independent datasets for any discrepancies. Disagreements were resolved after consulting a senior investigator (PG) to reach a consensus.
Risk of bias in individual studies
Two authors (MM and GT) assessed all the studies independently for risk of bias and any discrepancies were resolved after discussion with a third author (PG). We used the Cochrane collaboration’s risk of bias assessment tool to assess the studies. This particular tool evaluates different domains of potential sources 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), and selective reporting (reporting bias). All studies were classified by the reviewers as having high, low, or unclear risk of bias for each domain.
Statistical analysis
Extracted data were analyzed using the statistical software Review Manager (RevMan 5.4.1; Copenhagen, Denmark; The Nordic Cochrane Centre, The Cochrane Collaboration, 2020). For the primary and all secondary endpoints, relative risks (RRs) and 95% confidence intervals (95% CIs) were calculated. Forest plots were created for visual presentation of the results, and all outcomes were compared using either the fixed-effects model (Mantel and Haenszel method) or the random-effects model (DerSimonian and Laird method) in the absence or presence of significant heterogeneity, respectively. The presence of heterogeneity was calculated using I2 tests with I2<30% interpreted as low-level heterogeneity and I2 between 30% and 60% as moderate heterogeneity. Any potential publication bias was verified through the visual assessment of funnel plots. We repeated the meta-analysis excluding 1 study at a time to assess whether its exclusion altered the heterogeneity’s significance level. Funnel plots, constructed by plotting the log-ORs vs. the precision of individual studies per outcome, were assessed visually for symmetry to exclude potential publication bias.
Quality of evidence
The quality of the provided evidence was rated based on the GRADE criteria. Two independent researchers (MM and GT) graded inconsistency, risk of bias, indirectness, imprecision and publication bias. Overall quality was deemed very low, low, moderate, or high, using GRADEpro (GRADE Working Group) [11].
Results
Characteristics of included studies
The initial search identified 109 unique records; after application of the exclusion criteria 7 studies [7,8,12-16] were included in this meta-analysis. The PRISMA flowchart showing the study selection process is given in Fig. 1 and Table 1 summarizes the main characteristics of the included studies.
Figure 1 Flow diagram of assessment of the studies identified
Table 1 Study characteristics
Overall, 5785 patients were recruited in the meta-analysis. The female-to-male ratio was 1:1 and the mean age was approximately 60 years. Four studies were multicenter [7,13,15,16], whereas the remaining 3 were single-center [8,12,14]. The patient recruitment period was between 2014 and 2021 and for each study the timeline of study completion was from 1-3 years. All the studies had a unique origin (Denmark, Italy, Mexico, Germany, Greece, Italy, and South Korea). Three studies [12,13,15] included Endocuff Vision as the add-on-device mounted on the tip of the endoscope, while 2 used the first generation Endocuff [14,16]. Both Endorings (EndoAid Ltd, Caesarea, Israel) [7] and Wingcap (A&A Medical Supply LLC, Seongnam, South Korea) [8] were assessed in 1 study. None of the published studies using NaviAid G-EYE (SMART Medical Systems Ltd., Ra’anana, Israel) or Amplifeye (Medivators Inc., Minneapolis, MN) were confined exclusively to screening colonoscopies and ADR. All studies enrolled individuals undergoing colonoscopy for CRC screening.
Quality assessment
A summarized assessment of the risk of bias per study using the Cochrane Collaboration’s risk of bias assessment tool is illustrated in Fig. 2. Participating physicians were not blinded to the equipment used or the measured outcomes in any of the studies. Three [7,8,12] of the 7 studies did not report the exact method of allocation concealment (selection bias). A detailed assessment of the risk of bias is summarized in Supplementary Table 1.
Figure 2 Risk of bias of the studies included
Grade evidence estimate
Confidence in the effect estimates was considered very low. The quality of the body of evidence was downgraded by 3 levels for the primary outcome: 1 due to the risk of bias given that blinding of the endoscopists was not possible; 1 because of the presence of heterogeneity; and 1 because of the presence of evidence indirectness, since the technical intervention was implemented by expert, highly trained specialists in specialist centers and in different populations. A detailed assessment of the evidence grade is summarized in Supplementary Table 2.
Primary endpoint
Colonoscopy with the assistance of add-on devices with projections [7,8,12-16] yielded an ADR of 45.9%, compared to 41.1% for SC. A comparison of the ADR achieved by these techniques found superiority for add-on devices with projections, albeit with high heterogeneity: RR 1.18, 95%CI 1.02-1.37; P=0.03; I2=79% (Fig. 3). Visual assessment of the funnel plot showed no evidence of publication bias (Supplementary Fig. 1). In an effort to address heterogeneity, 2 sensitivity analyses were performed. During the step-by-step sensitivity analysis, in which 1 study was excluded at a time, no study was found that could explain this result. Based on the high percentage of patients undergoing screening colonoscopy with the use of Endocuff, a sub-group analysis was performed to compare ADR between the Endocuff group, either first or second generation, and the SC group [12-16]. Our analysis revealed superior ADR rates for the Endocuff-assisted procedures: 46.2% vs. 40.8%; RR 1.18, 95%CI 1.02-1.36; P=0.03; I2=73%] (Supplementary Fig. 2).
Figure 3 Forest plot assessing the adenoma detection rate of add-on devices vs. standard colonoscopy (SC)
CI, confidence interval
PDR
In terms of PDR, data from 5 studies [7,12,14-16] also favored the use of add-on devices with projections in screening colonoscopies, in comparison to SC, without reaching statistical significance: 55.1% vs. 50.8%; RR 1.10, 95%CI 0.96-1.26; P=0.17; I2=75% (Fig. 4A). The sensitivity analysis, excluding 1 study at a time, failed to identify a single study accountable for this effect. No evidence of publication bias was found (Supplementary Fig. 3).
Figure 4 Forest plot assessing effect of add-on devices vs. standard colonoscopy (SC) on A: polyp detection rate; B: advanced adenoma detection rate; C: sessile adenoma detection rate
CI, confidence interval
AADR
Data acquired by 4 studies [7,13,15,16] showed a lack of difference between the 2 groups in terms of AADR: 18.5% vs. 17.6%; RR 1.00, 95%CI 0.79-1.27; P=0.98; I2=56% (Fig. 4B). Sensitivity analysis did not detect any study responsible for the detected heterogeneity. There was no evidence of publication bias (Supplementary Fig. 4).
SSLDR
Similarly, only 4 studies reported the SSLDR [7,12,13,15]. The SSLDR of screening colonoscopies assisted by an add-on device with projections was not superior to that of SC: 6.8% vs. 5.8%; RR 1.17, 95%CI 0.95-1.44; P=0.15; I2=0% (Fig. 4C). Neither heterogeneity nor publication bias was detected (Supplementary Fig. 5).
Discussion
Our meta-analysis is the first to evaluate add-on devices with projections exclusively in the context of screening colonoscopies, revealing statistically significant higher percentages of ADR, compared to the SC technique. Based on unequivocal inclusion criteria, the outcomes of our meta-analysis are quite solid. Firstly, only prospective RCTs were studied to ensure low rates of bias. Secondly, the selected screening population ensured that the participants neither exhibited gastrointestinal symptoms, nor had a known history of CRC; these variables are associated with higher ADRs. Thirdly, both groups in each study were evenly classified, leading to diminished odds of misrepresentation. Apart from that, the large number of enrolled patients (n=5785) contributed to the validity of the study. The authors, therefore, believe that the findings of the present study may pave the way for more widespread use of add-on device-assisted colonoscopy in the screening population.
Our analysis showed a clear-cut benefit from the use of add-on devices with projections compared to SC in terms of ADR which is the core quality indicator of the examination and has been inextricably linked to CRC prevention and incidence reduction [17]. Since Endocuff has been widely reported in the bibliography, a further analysis of its usage solely in screening colonoscopies reaffirmed that the insertion of this specific device surpasses SC’s ADR. Individual studies [13,18] have also shown that Endocuff exhibits superior rates in adenoma detection compared to SC. Our study concurs with our previously published meta-analysis in 2019 about the importance of Endocuff [6]. Regarding Endoring, a plethora of reviews demonstrates its dynamic; still, the majority of them either skips the comparison to SC [19-21], or includes techniques which are not considered add-on devices, such as full-spectrum (FUSE) [22,23]. As for WingCap, only 1 published study is to be found by Hong et al [8] comparing this device to SC. Yet, this is the first systematic meta-analysis to investigate the majority of add-on devices in screening population.
While the use of add-on devices carries an extra cost for the healthcare system, the cost/benefit ratio is yet to be established. Yu et al used a decision-analytic Markov model to evaluate the clinical and economic consequences of Endocuff-assisted screening colonoscopies [24]. Although they highlighted the cost-effectiveness of the Endocuff intervention in endoscopy, data on the economic burden associated with these devices varies between regions and depends on the equipment; thus, further studies are needed. Moreover, endoscopists have to receive additional training before performing colonoscopies with add-on devices, adding to the overall cost. It is worth noting that add-on devices may not be suitable for all cases. For instance, in the presence of diverticular disease [12] the attachments had to be removed in order to complete the endoscopy. This is also the case if a polypectomy is needed, and the add-on device hinders the resection.
This systematic review has also some limitations. First, the absence of RCTs on every available tool limits the applicability of our results to those we assessed, even though some commercially available attachments may provide comparable results. A systematic review of different types of add-on devices in screening colonoscopies is needed to further establish the superiority of specific techniques. This was not feasible in our study, since 5 of the 7 chosen publications reported the use of Endocuff [12-16], with the remaining 2 making use of Endoring [7] and WingCap [8]. Undoubtedly, there are other devices attached to the tip of the endoscope which have not been mentioned in this meta-analysis. This is mainly due to the lack of comparison with the SC group, as well as the exclusion of screening populations, leading to selection bias [25,26]. Some of those devices include Amplifeye and G-eye [27,28].
Another drawback is the inability to assess potential confounders affecting ADR. For example, endoscopists’ personal ADR is based on experience and training, and may vary among studies. Moreover, bowel preparation, patients’ comfort and other parameters affecting the quality of colonoscopy were not assessed, as there were no data amenable to analysis. Another concern is the limited elicited data related to AADR and SSLDR, with only 4 of 7 studies providing such information. Hence, further analyses are needed to correlate the use of add-on devices with these parameters.
To conclude, this systematic review highlights the role of add-on devices with projections in CRC detection among a screening population. Although these devices displayed higher percentages in terms of ADR (primary outcome), compared to SC, more data are needed to reinforce these findings. Ideally, more adults undergoing screening endoscopies should be offered add-on device-assisted colonoscopy, depending on the availability of these technologies.
Acknowledgment
The authors express their gratitude to Hong Seung Wook, who kindly provided further information on his study.
References
1. Lotfollahzadeh S, Recio-Boiles A, Cagir B. Colon cancer. In:StatPearls [Internet]. Treasure Island (FL):StatPearls Publishing;2022. Available from:http://www.ncbi.nlm.nih.gov/books/NBK470380/[Accessed 7 July 2023].
2. Kim SY, Kim HS, Park HJ. Adverse events related to colonoscopy:global trends and future challenges. World J Gastroenterol 2019;25:190-204.
3. Zhao S, Wang S, Pan P, et al. Magnitude, risk factors, and factors associated with adenoma miss rate of tandem colonoscopy:a systematic review and meta-analysis. Gastroenterology 2019;156:1661-1674.
4. Tang Z, Zhang DS, Thrift AP, Patel KK. Impact of cap-assisted colonoscopy on the learning curve and quality in colonoscopy:a randomized controlled trial. Gastrointest Endosc 2018;87:723-732.
5. Choi J, Litten JD, Drozek D. Development of a colonoscopy add-on device for improvement of the intubation process. Med Devices (Auckl) 2011;4:197-208.
6. Triantafyllou K, Gkolfakis P, Tziatzios G, Papanikolaou IS, Fuccio L, Hassan C. Effect of endocuff use on colonoscopy outcomes:a systematic review and meta-analysis. World J Gastroenterol 2019;25:1158-1170.
7. Hassan C, Senore C, Manes G, et al. Diagnostic yield and miss rate of EndoRings in an organized colorectal cancer screening program:the SMART (Study Methodology for ADR-Related Technology) trial. Gastrointest Endosc 2019;89:583-590.
8. Hong SW, Hong HS, Kim K, et al. Improved adenoma detection by a novel distal attachment device-assisted colonoscopy:a prospective randomized controlled trial. Gastrointest Endosc 2022;96:543-552.
9. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement:an updated guideline for reporting systematic reviews. BMJ 2021;372:n71.
10. Eriksen MB, Frandsen TF. The impact of patient, intervention, comparison, outcome (PICO) as a search strategy tool on literature search quality:a systematic review. J Med Libr Assoc 2018;106:420-431.
11. Schünemann H, Brożek J, Guyatt G, Oxman A. GRADE handbook for grading quality of evidence and strength of recommendations [Internet]. 2013. Available from:https://gdt.gradepro.org/app/handbook/handbook.html [Accessed 7 July 2023].
12. Jaensch C, Jepsen MH, Christiansen DH, Madsen AH, Madsen MR. Adenoma and serrated lesion detection with distal attachment in screening colonoscopy:a randomized controlled trial. Surg Endosc 2022;36:1-9.
13. Zorzi M, Hassan C, Battagello J, et al;ItaVision Working Group. Adenoma detection by Endocuff-assisted versus standard colonoscopy in an organized screening program:the “ItaVision“randomized controlled trial. Endoscopy 2022;54:138-147.
14. González-Fernández C, García-Rangel D, Aguilar-Olivos NE, et al. Higher adenoma detection rate with the endocuff:a randomized trial. Endoscopy 2017;49:1061-1068.
15. Zimmermann-Fraedrich K, Sehner S, Rösch T, et al. Second-generation distal attachment cuff for adenoma detection in screening colonoscopy:a randomized multicenter study. Gastrointest Endosc 2023;97:112-120.
16. Triantafyllou K, Polymeros D, Apostolopoulos P, et al. Endocuff-assisted colonoscopy is associated with a lower adenoma miss rate:a multicenter randomized tandem study. Endoscopy 2017;49:1051-1060.
17. Kaminski MF, Wieszczy P, Rupinski M, et al. Increased rate of adenoma detection associates with reduced risk of colorectal cancer and death. Gastroenterology 2017;153:98-105.
18. Ngu WS, Bevan R, Tsiamoulos ZP, et al. Improved adenoma detection with Endocuff Vision:the ADENOMA randomised controlled trial. Gut 2019;68:280-288.
19. Thayalasekaran S, Alkandari A, Varytimiadis L, et al. To cap/cuff or ring:do distal attachment devices improve the adenoma detection?Expert Rev Gastroenterol Hepatol 2019;13:119-127.
20. Pamudurthy V, Lodhia N, Konda VJA. Advances in endoscopy for colorectal polyp detection and classification. Proc (Bayl Univ Med Cent) 2020;33:28-35.
21. Gkolfakis P, Tziatzios G, Spartalis E, Papanikolaou IS, Triantafyllou K. Colonoscopy attachments for the detection of precancerous lesions during colonoscopy:A review of the literature. World J Gastroenterol 2018;24:4243-4253.
22. Brand EC, Dik VK, van Oijen MGH, Siersema PD. Missed adenomas with behind-folds visualizing colonoscopy technologies compared with standard colonoscopy:a pooled analysis of 3 randomized back-to-back tandem colonoscopy studies. Gastrointest Endosc 2017;86:376-385.
23. Gkolfakis P, Tziatzios G, Facciorusso A, Muscatiello N, Triantafyllou K. Meta-analysis indicates that add-on devices and new endoscopes reduce colonoscopy adenoma miss rate. Eur J Gastroenterol Hepatol 2018;30:1482-1490.
24. Yu TM, Tradonsky A, Tang J, Arnold RJ. Cost-effectiveness of adding Endocuff® to standard colonoscopies for interval colorectal cancer screening. Clinicoecon Outcomes Res 2019;11:487-504.
25. Verheyen E, Castaneda D, Gross SA, Popov V. Increased sessile serrated adenoma detection rate with mechanical new technology devices:a systematic review and meta-analysis. J Clin Gastroenterol 2021;55:335-342.
26. Rex DK, Sagi SV, Kessler WR, et al. A comparison of 2 distal attachment mucosal exposure devices:a noninferiority randomized controlled trial. Gastrointest Endosc 2019;90:835-840.
27. Hocke M, Frieling T, Eckardt AJ, Teubner D, Kiesslich R. Discovery AI combination with G-eye colonoscopy significantly increases adenoma detection rate - results of a multicenter study. Gastrointest Endosc 2022;95:AB217.
28. Halpern Z, Gross SA, Gralnek IM, et al. Comparison of adenoma detection and miss rates between a novel balloon colonoscope and standard colonoscopy:a randomized tandem study. Endoscopy 2015;47:238-244.