Combining techniques and technologies increases adenoma detection rates in colonoscopy: More is more

Abstract

Screening colonoscopy with adenoma removal is the gold standard strategy to reduce colorectal cancer (CRC) incidence. Nevertheless, it remains an imperfect tool as nearly Twenty-five percent of adenomas can be missed during inspection by experienced endoscopists. Missed lesions are one of the primary reasons for post colonoscopy CRC and are associated with a significant variability in adenoma detection rate (ADR), which is the most important quality indicator for colonoscopy. Increasing ADR unquestionably decreases carcinoma miss rate. Simple measures to improve ADR include among others slower withdrawal time and position change. The introduction of optical imaging innovations has improved mucosal visualization. Moreover, auxiliary devices attached to the colonoscope tip have been introduced, aiming to improve lumen visualization by flattening the folds and revealing lesions hidden in blind spots, thereby increasing ADR. Digital image analysis using artificial intelligence is the latest approach to polyp detection. All of the above approaches have been separately evaluated concerning their effect in ADR; however, it has not been thoroughly investigated whether any benefit exists from their combined use. We aim to review the available data on the efficacy of each technique/technology and whether their combination offers any additional benefit while remaining cost-effective.

Key Words: Screening colonoscopy; Adenoma detection rate; Optical imaging innovations; Endocuff; Computer-aided diagnosis systems

Core Tip: Nowadays, no single measure has uniformly proven its efficacy in increasing the adenoma detection rate (ADR). Optical imaging innovations, auxiliary devices and computer-aided diagnosis systems may be of value when used by trainees and low detectors. Unfortunately, none of these have been able to increase the advanced ADR (AADR). Recent data from studies combining devices and technologies have shown increased ADRs, though failed to provide solid evidence that they increase AADR, especially among expert endoscopists. Data evaluating the combined efficacy of simple measures, such as a second look of the right colon, together with more advanced methods, are needed.

INTRODUCTION

Colonoscopy is the most efficacious screening method for colorectal cancer (CRC)[1,2]. Colonoscopy not only has the highest combined sensitivity and specificity in detecting CRC but also has the ability to detect and treat precancerous lesions, such as adenomas, thus acting as a prevention tool[1,2]. Therefore, its quality as a detection method is of paramount importance. The mostly used, investigated and validated metric assessing the quality of screening colonoscopy is adenoma detection rate (ADR)–the percentage of screening colonoscopies where an adenoma is identified–since ADR is inversely associated with the risk of developing interval CRC[3].

Optimal colonoscopy conditions are critical for a good quality screening procedure. Effective bowel preparation[4-7], reaching the cecum[8-10], and allowing adequate time (> 6 minutes) for the examination are considered fundamental prerequisites for a successful screening examination.

Unfortunately, colonoscopy remains an imperfect tool, with even experienced endoscopists missing up to 25% of adenomas[11]. Over time, several strategies have been introduced to increase ADR, including simple maneuvers, insertion techniques, technological developments enhancing the projected image, auxiliary devices designed to expand the visual field, and more recently, artificial intelligence (AI). Data regarding the efficacy of each of these modalities in altering ADR exist in the literature with varying results; however, the evidence regarding their combined efficacy remains limited. The aim of our review is to evaluate the most recent evidence regarding the efficacy of each modality independently and to summarize all available data from studies investigating whether a combination of such measures offers an additional impact on ADR.

SIMPLE MEASURES

Numerous simple and cost-effective techniques and maneuvers have been proposed to increase ADR without the need for technological developments or auxiliary devices. Among these, Water-Assisted Colonoscopy (WAC), second look of the right colon, position changes and the use of antispasmodics, have all been investigated.

Right colon second look and/or retroflexion

The right colon is a common site for missed adenomas and malignancies mainly due to the presence of flat adenomas and serrated lesions[12]. Data suggests that performing a second look in the right colon, either through retroflexion or a repeated forward view can increase ADR by approximately 5%-20%[13-15].

Water exchange

Water exchange is the gold standard WAC technique as supported by current literature. Water irrigation enhances mucosal visualization and its use has been associated with an increase in ADR[16]. A meta-analysis including 10350 patients demonstrated that water immersion during colonoscope insertion can increase ADR during screening procedures[17]. Another more recent meta-analysis indicated that water exchange, compared to gas insufflation also results in a higher serrated polyp detection rate (SPDR)[18].

Position change

Position change can enhance lumen visibility by facilitating the elevation of gas to the highest point and allowing fluid to move away from the area of interest, thus improving luminal distension. The largest study evaluating the efficacy of position change on ADR was a multi-center randomized control trial (RCT) including 1072 patients, randomized to either a left lateral position or dynamic position changes on withdrawal. A higher ADR, as well as an increase in the number of adenomas detected, was observed in the dynamic position change group (P = 0.002 and P = 0.01, respectively). However, the benefit was mainly attributed to endoscopists with a lower baseline ADR (< 35%) (low detectors). Moreover, no statistically significant improvement in advanced ADR (AADR) was noted in the intervention group[19]. Two recent meta-analyses have further suggested that dynamic position changes during colonoscopy can increase ADR without prolonging mean withdrawal time[20,21]. In contrast, a parallel design RCT of 776 patients by Ou et al[22], failed to show any significant benefit of dynamic position changes over the endoscopist’s usual adopted position in improving ADR.

Antispasmodics

Antispasmodics have been proposed as means to improve mucosal visualization by reducing colonic spasms; however, the evidence[23] regarding their impact on ADR remains controversial[24-28]. Additionally, none of the four meta-analyses assessing hyoscine butylbromide, the most commonly used antispasmodic, demonstrated a significant benefit in increasing ADR[29-32].

OPTICAL IMAGING INNOVATIONS

Image enhancing modalities aim to improve ADR by enhancing visualization of tissue vasculature and surface structures through various technological approaches.

Narrow-band imaging

Narrow band imaging (NBI) (Olympus) was the first system introduced and the one most studied. Meta-analyses comparing NBI with white light endoscopy (WLE) has not provided conclusive evidence supporting its superiority in adenoma detection among average risk screening populations[33-37]. However, it must be underlined that in certain high-risk populations, such as patients with inflammatory bowel disease, NBI has now been shown to enhance the detection of dysplastic lesions[38-41].

Blue laser imaging

Comparative analyses of blue laser imaging (BLI) and high definition (HD)-WLE have yielded controversial results as the majority have failed to demonstrate a significant advantage of BLI in terms of ADR, taking into account that the available studies are limited in number and lacking meta-analyses[42-46]. Nevertheless, a recent study comparing BLI, NBI and WLE reported a lower proximal adenoma miss rate with BLI [difference in ADR of 8.3% (95%CI: 2.7-15.9)] in favor of BLI compared to WLE[47].

I-scan

Two meta-analyses have evaluated the efficacy of i-scan technology in adenoma detection compared to HD-WLE, with both providing evidence that i-scan improves ADR[48,49].

Linked color imaging

Two recent meta-analyses provide evidence and assessed the efficacy of linked color imaging (LCI) in adenoma detection compared to HD-WLE including 10 studies (5510 patients) and 17 studies (10624 patients), respectively[50,51]. Both analyses concluded that LCI increases ADR, although the benefit appears confined to specific categories of adenomas (< 10 mm) and among endoscopists with lower baseline detection rates[50,51].

Texture and color enhancement imaging

Tree studies have investigated the impact of texture and color enhancement imaging (TXI), all showing some benefit. Reported ADRs with TXI were 58.2%, 55% and 58.9% compared to 46.8%, 49.4% and 42.9% with HD-WLE[52-54]. However, it must be stated that only one study reported an increase in the detection of high-risk adenomas [17.6% vs 12.8%; odds ratio (OR): 1.45] and data from the only available randomized study, indicated that the benefit was primarily limited to adenomas < 10 mm[53,54].

Summary

In conclusion current evidence regarding the use of image-enhanced endoscopy for adenoma detection, regardless of the modality applied, does not consistently support a generalized benefit among average risk individuals undergoing screening endoscopy. These technologies appear to improve detection rates particularly when used by non-expert endoscopists, under optimal bowel preparation conditions and predominantly or non-advanced adenomas measuring less than 1 cm.

AUXILIARY DEVICES ATTACHED TO THE TIP OF THE COLONOSCOPE

One of the major causes for missed polyps during colonoscopy is the restricted visualization of blind spots, such as areas behind colonic folds. To address the limitation, several attachments have been developed, including systems permanently integrated into the scope, such as the G-EYE system, and removable devices such as the Endocuff, EndoRing and Transparent cap.

G-EYE

A recent meta-analysis assessing the use of G-EYE, encompassing 3868 colonoscopies, confirmed its efficacy in improving both ADR and SPDR. The pooled analysis showed a statistically significant improvement with G-EYE with an OR of 1.744 for ADR (95%CI: 1.534-1.984, P < 0.001); and 1.603 for SPDR (95%CI: 1.176-2.185, P = 0.003)[55].

Cap

Data regarding cap-assisted colonoscopy and its impact in ADR are mixed. Two meta-analyses failed to demonstrate a significant improvement in ADR with the use of cap. However, another meta-analysis, including four studies suggested that cap may increase ADR in the proximal colon as well as increase the detection of flat adenomas and serrated colonic lesions compared with standard colonoscopy (SC)[56-58].

Endocuff

Regarding the use of Endocuff-assisted colonoscopy (EAC), RCTs, including studies conducted within different national bowel cancer screening programs, have failed to provide definitive evidence that EAC consistently increases ADR[59-63]. Nevertheless, meta-analytic data suggest that EAC use can significantly improve ADR, particularly among endoscopists with lower baseline detection rates (ADRs < 30%) where an increase of 9.4%, in ADR (P = 0.03)[64-66]. However, similar to other auxiliary devices, no improvement in AADR has been demonstrated.

EndoRing

Apart from findings from an initial study, subsequent data derived from RCTs, retrospective studies and meta-analyses have failed to demonstrate a consistent benefit of EndoRing use in improving ADR[67-71].

Summary

Overall, the evidence regarding the impact of add-on devices on ADR remains controversial. Most positive results have been associated with the use of Endocuff. However, a recent multicenter RCT found no significant benefit in ADR when comparing the use of distal attachment devices (EAC or cap) with HD-WLE, particularly among high-detector endoscopists[72]. Similarly, a recent meta-analysis evaluating all three methods demonstrated that while add-on devices were associated with increased ADR compared to SC, no significant differences were observed in AADR[73].

When considering the available data collectively, it appears that, similar to image-enhancing modalities, the primary benefit of add-on devices–particularly Endocuff–may be limited to low-detector endoscopists and trainees. Moreover, their advantage seems to be on the detection of non-advanced adenomas raising questions regarding the clinical significance of their use among experienced endoscopists. In conclusion, the decision to use add-on devices may largely depend on their cost-effectiveness and the potential for modest gain in certain settings.

WIDE-ANGLE COLONOSCOPES AND RETROGRADE VIEWING DEVICES

Standard colonoscopes are limited by a relative narrow field of view (typically 140°-170°) which may contribute to missed polyp detection. Several innovations have aimed to address this limitation, either by increasing the endoscope’s viewing angle through the addition of lateral lenses or by the incorporation of removable devices equipped with supplementary cameras. However, at present, the routine use of wide-angle colonoscopes or retrograde viewing devices cannot be recommended primarily due to limited device availability and a paucity of robust evidence supporting their efficacy[74,75].

COMPUTER-AIDED DIAGNOSIS SYSTEMS

The application of AI in endoscopy has led to the development of computer-aided diagnosis (CADe) systems which utilize deep learning models to assist in the real-time detection of lesions during colonoscopy. The first CADe system received approval from the United States Food and Drug Administration in 2020[76] and since then, several CADe systems have become commercially available. The four most extensively studied CADe systems are CAD EYE (Fujifilm, Tokyo, Japan); EN-DOAID (Olympus, Tokyo, Japan); EndoScreener (Shanghai Wision AI, Shanghai, China), and GI-Genius (Medtronic, Minneapolis, Minn, United States).

CADe systems hold significant promise in reducing inter-operator variability and improving the ADR. A number of past and recent meta-analyses have demonstrated that AI-assisted colonoscopy increases ADR by approximately 8%-20%[77-79] with the most commonly investigated CADe systems exhibiting similar performance [relative risk (RR) estimates ranging from 1.14 to 1.27]. In addition to improving ADR the CADe appears to substantially lower the adenoma missing rate (AMR), a metric directly associated with the incidence of interval cancer and considered a more stringent measure of screening efficacy[80,81]. For example, Wang et al[82], reported an AMR of 13.89% with CADe vs 40% with WLC. A meta-analysis including six tandem studies demonstrated a significantly lower AMR with CADe compared to WLC (RR = 0.46, 95%CI: 0.38-0.55, P < 0.001)[83]. However, the benefit of CADe systems appears more pronounced among low-detectors endoscopists than among high performer[79]. This finding was further supported by a recent RCT in which CADe did not significantly improve ADR among high-performing endoscopists who routinely used Endocuff[84]. Additionally, a recent metanalysis suggests that the increase in ADR attributed to AI is predominantly related to the detection of diminutive polyps (< 5 mm) with no significant impact on advanced adenoma detection[85]. Consequently, while CADe technologies represent a promising advancement, their true clinical impact remains uncertain and further studies assessing patient-important outcomes, such as CRC incidence and mortality, are warranted.

STUDIES USING COMBINATION OF METHODS

To date no single measure has uniformly proved its efficacy in increasing ADR, especially among expert endoscopists, and with respect to AADR. Consequently, it is reasonable to hypothesize that the combination of different methods may offer additive benefits, even when using such stringent endpoints. However, only a limited number of studies evaluating combined approaches are available in the literature mainly involving the combination of EAC with either image-enhanced modalities or CADe systems.

An initial prospective, multicenter, RCT including 1905 participants assessed the efficacy of combining a transparent cap with virtual chromoendoscopy (CAP/CHROMO) in compared to SC in a screening population. The study demonstrated a significant increase in ADR (38.6% vs 31.2%) (P = 0.001) and higher proximal SPDRs in the CAP/CHROMO group, although this effect was only observed among female participants. However, there was no significant difference in AADR between the CAP/CHROMO and SC groups (9.3% vs 7.6%, P = 0.180)[86]. Subsequently, a randomized trial investigating the combination of EAC and LCI reported that their concomitant use significantly improved proximal ADR compared to either modality alone or to SC [ADRs of LCI + EAC, LCI, EAC, and HD colonoscopy: (1) 57.2%; (2) 52.8%; (3) 51.6%; and (4) 47.6%, respectively]. Nevertheless, the benefit was not maintained among high-ADR performers[87].

More recently Pattarajierapan et al[88] evaluated the combination of TXI with EAC finding that the ADR was significantly higher in the TXI + EAC group than in the TXI-only group (65.6% vs 52.1%, P = 0.007).

Regarding the combination of EAC with CADe, a pivotal study by Aniwan et al[89], including 1245 participants evaluated whether CADe combined with EAC would be superior to either modality alone or to SC. The authors concluded that the CADe + EAC combination significantly improved both ADR and AADR compared to SC (13.6% vs 7.7%, P = 0.02). However, when comparing the combination to CADe or EAC individually, no additional benefit in AADR was demonstrated, despite increased ADRs.

A more recent study similarly found that combining EAC with AI-assisted colonoscopy significantly improved ADR and SPDR compared to AI alone (P = 0.03 and P < 0.001 respectively). Both the EAC-AI and AI-alone groups demonstrated higher ADR, SPDR, and AADR compared to the HD colonoscopy group (P < 0.001 for all comparisons, EAC-AI (58.7%), AI (53.8%), HD:ADR (46.3%); EAC-AI (41.7%), AI (35.3%), HD:SPDR (21.5%); EAC-AI (17.4%), AI (17.2%), HD:AADR (11.7%), respectively[90].

Conversely, a small study with a sample size of 90 participants, while confirming that the combination of CADe and EAC increased ADR, reported that the benefit was primarily driven by the increased detection of diminutive adenomas, with no improvement observed for adenomas > 5 mm[91].

CONCLUSION

In conclusion, although the available data regarding the combination of EAC with either image-enhanced modalities or CADe systems suggest an increased ADR, once again a clear and universal benefit from combining methods cannot yet be firmly established. As with previous studies evaluating individual interventions, the observed improvements appear to be primarily attributable to the detection of diminutive polyps. Moreover, although two studies evaluating the addition of CADe to EAC reported an increase in AADR, neither study accounted for the level of endoscopist expertise, which is now recognized as a crucial determinant of ADR outcomes.

Consequently, although there are indications that the synchronous use of multiple methods used may improve not only ADR but also AADR and SPDR, further data stratified by endoscopist expertise are necessary. To date, no studies have evaluated the combined efficacy of simple measures–such as a second look at the right colon, or dynamic position changes–with more advanced modalities such as CADe systems or adjunctive devices. We believe it is of paramount importance that future research investigates the synergistic effects of combining simple procedural techniques with emerging technological innovations. The integration of new technologies should not lead to the abandonment of fundamental, simple measures; rather technological advancements should build upon established practices to improve further adenoma detection outcomes (Table 1).

Table 1 Different technological innovations and their benefit in adenoma detection rate, advanced adenoma detection rate.

	Increase in ADR	Increase in AADR	Comments
Optical imaging innovation
Narrow-band imaging	No clear benefit	No benefit	Benefit in detection of dysplastic lesions in patients with inflammatory bowel disease
Blue laser imaging	No clear benefit	No benefit	Lower proximal adenoma miss rate compared to white light endoscopy in one study
I-scan	Improves ADR	No benefit
Linked color imaging	Improves ADR	No benefit	The benefit was mainly attributed to selected categories of adenomas (< 10 mm) and among low detectors
Texture and color enhancement imaging	Improves ADR	No clear benefit	Increased AADR in 1/3 studies–benefit in ADR restricted in adenomas < 10 mm
Auxiliary devices attached to the tip of the colonoscope
G-EYE	Improves ADR	No benefit	Benefit in detection of serrated polyp detection rate–not commonly accessible, paucity of data compared to cap, Endocuff
Cap	No clear benefit	No benefit	Easily accessible, one meta-analysis showing increased ADR in the proximal colon as also to increased detection of flat adenomas and serrated colonic lesions
Endocuff	No clear benefit	No benefit	Easily accessible, increase in ADR among low detectors
EndoRing	No benefit	No benefit
CADe systems	Improves ADR	No benefit	Benefit is restricted mainly to low detectors and adenomas < 10 mm
Combining optical imaging innovation with Endocuff	Improves ADR	No benefit	Benefit is restricted mainly to low detectors
Combining CADe systems with Endocuff	Improves ADR	Possible benefit	No study comparing low with high detectors exist, 1/3 available studies failed to prove any benefit in adenomas > 5 mm

AADR: Advanced adenoma detection rate; ADR: Adenoma detection rate; CADe: Computer-aided diagnosis.