2018 年 4 巻 1 号 p. 6-10
Objectives: Full-spectrum endoscopy (FUSE) provides a 330° angle of view with two side-viewing cameras in addition to a forward-viewing camera and may improve the colorectal polyp detection rate in blind spots. We evaluated whether FUSE improved colonic polyp detection in a single-center, randomized, crossover trial.
Methods: Between July 2016 and May 2017, 55 participants (40 men, 15 women) aged 23 to 82 years were enrolled for colorectal polyp screening. Participants were randomly allocated to a forward-viewing (FV)-first group or a FUSE-first group, with stratification according to participant age, sex, and endoscopist. The right colon was examined twice in succession from the cecum to the splenic flexure. Colonic polyp miss rates (PMRs) and endoscopic observation times were compared.
Results: Five participants were excluded: three because of insertion difficulty and two because of reinsertion difficulty. In the per-protocol analysis, 23 participants (46.0%) were randomly assigned to the FUSE-first group and 27 (54.0%) to the FV-first group. In the per-lesion analysis, the PMR was significantly lower with FUSE (1/29, 3.4%) than with FV (12/39, 30.7%; P=0.025). However, by per-participant analysis, polyp detection rates were not significantly different (11/23 [47.8%] with FUSE vs 9/27 [33.3%] with FV; P=0.667). All missed polyps were less than 6 mm in size. The observation time was significantly longer with FUSE than with FV in the FUSE-first group, but not in the FV-first group.
Conclusions: This study provides evidence that FUSE improves colonic polyp detection, especially in the right colon, and might improve interval colorectal cancer screening efficacy.
According to the Projected Cancer Statistics for 2016 from the National Cancer Center of Japan, colorectal cancer is the second leading cause of death and the most common cancer in Japan.1 Therefore, screening and surveillance for colorectal cancer is crucial. The National Polyp Study of the United States has reported that endoscopic resection of all adenomatous polyps to achieve a “clean colon” can reduce colorectal cancer mortality.2 That study recommended that colorectal adenomas, precursor lesions of cancer, and early-stage colorectal cancer should be detected and resected as early as possible. Forward-viewing colonoscopy (170° angle of view) is generally used in practice. To improve the colorectal polyp detection rate with standard endoscopy in blind spots such as flexures and folds, the tip cap3 and blue-water infusion4 methods have been used. Recently, full-spectrum endoscopy (FUSE) has been developed, with a wider, 330° angle of view achieved with two side-viewing cameras in addition to a forward-viewing (FV) camera. In Japan, FUSE received pharmaceutical approval in 2015; however, there have been few reports on its use.5 In this study, we evaluated whether FUSE improved the colonic polyp detection rate in a single-center, randomized, crossover trial.
We enrolled participants aged 20 to 85 years who were referred for colorectal screening and polyp surveillance between July 2016 and May 2017 at Fujita Health University Hospital. All patients provided written informed consent. We excluded participants with suspected colorectal polyps or cancers who had undergone prior examinations such as computed tomography or barium enema and those with a history of colorectal resection, multiple laparotomy, inflammatory bowel disease, polyposis syndrome, lower gastrointestinal bleeding, colonic stricture, acute diverticulitis or toxic megacolon, or a history of radiation therapy to the abdomen or pelvis. This study was reviewed and approved by the Institutional Review Board of Fujita University School of Medicine and was registered with the University Hospital Medical Information Network (UMIN 000013857). The primary endpoint of this study was comparison of detection rates for proximal colonic polyps, including hyperplastic polyps, sessile serrated adenomas/polyps, adenomatous polyps, and carcinomas, between FUSE and FV colonoscopy. The secondary endpoint was comparison of examination time for FUSE versus FV colonoscopy.
ProceduresFigure 1 shows left, center, and right images of FUSE (EndoChoice, Alpharetta, GA, USA). FUSE is equipped with three cameras. A single center camera was designated as a substitute for an FV endoscope by masking the bilateral cameras. Patients were randomly assigned (1:1) to an FV-first group or a FUSE-first group by computer-generated randomization with stratification according to participant sex, age, and endoscopist. Colon preparation was performed with polyethylene glycol solution (either Moviprep or Niflec; EA Pharma, Tokyo, Japan). We assessed bowel cleanliness at the time of colonoscopy with the Boston bowel preparation scale; a score ≥5/6 points was considered good and ≥3/6 was considered observable.6 Colonoscopy was performed without sedation except in one patient. Participants underwent colonoscopy twice in the right colon from the cecum to the splenic flexure. In the FV-first group, we performed the first colonic observation with only the center camera of FUSE from the cecum to the splenic flexure, which we marked with a few drops of crystal violet solution. Next, we reinserted the colonoscope to the cecum and performed a second colonic observation with all three FUSE cameras to the marked spot. This order was reversed in the FUSE-first group. Polyp resection or biopsy was performed for all observed polyps, which were submitted for pathologic diagnosis. The colonic polyp miss rate (PMR)7 was defined as the proportion of polyps newly detected during the second observation. The polyp detection rate (PDR)8 was defined as the frequency of detecting one or more polyps per participant. PMR, PDR, and endoscopic observation time were compared between the FV and FUSE observations and between the two groups. Endoscopic observation time from the cecum to the splenic flexure was measured; the procedural time during biopsy and endoscopic polyp resection was not included in the observation time. The shape of polyps was classified according to the Japanese Society for Cancer of the Colon and Rectum.9
Full-spectrum endoscopy (FUSE) images. An arrow indicates a polyp that was observed only with the left camera.
Values are expressed as the mean±standard deviation. Comparisons of PMR and PDR were analyzed with Fisher’s exact probability test or χ2 test. Comparisons of endoscopic observation times were analyzed with the Mann-Whitney U test or Wilcoxon signed-rank test. Differences were considered significant at P values less than 0.05.
Table 1 shows the demographic characteristics of participants. Among the 55 enrolled participants, five were excluded from analysis: three because of insertion difficulty and two because of reinsertion difficulty. In the per-protocol analysis, 23 participants (46.0%) were randomly assigned to the FUSE-first group and 27 (54.0%) to the FV-first group. In the FV-first group, 27 polyps were detected during the first observation with FV; an additional 12 polyps were detected during the second observation with FUSE. In the FUSE-first group, 28 polyps were detected during the first observation with FUSE; only one additional polyp was detected during the second observation with FV (Table 2). The PMR was significantly lower with FUSE (1/29 [3.4%]) than with FV (12/39 [30.7%]; P=0.025). Table 3 shows the detecting camera and size, shape, and location of all polyps detected. Even after initial observation with FV, 58.3% of polyps were found with the center camera and 41.7% were found with side cameras during subsequent FUSE. All missed polyps were less than 6 mm in size and most were type Is. With respect to location, the polyps missed during the first observation with FV were most often found in the transverse colon (58.3%).
| n=55 | |
|---|---|
| Age, years | 67 (23–82) |
| Sex (men:women) | 40:15 |
| Indications | |
| Positive fecal occult blood | 24 |
| Post-polypectomy follow-up | 12 |
| Elevated tumor markers | 4 |
| Preoperative screening | 7 |
| Bloody stool | 3 |
| Anemia | 1 |
| Abdominal/lower back pain | 2 |
| Abnormal barium enema results | 1 |
| Participant’s request | 1 |
| Excluded cases | |
| Insertion difficulty | 3 |
| Reinsertion difficulty | 2 |
| Boston Bowel Preparation Scale (excluding dropouts, n=50) | |
| 6/6 | 31 |
| 5/6 | 13 |
| 4/6 | 5 |
| 3/6 | 1 |
| 2/6 | 0 |
| 1/6 | 0 |
| 0/6 | 0 |
Values expressed as mean (range) or number of participants.
| Number of polyps detected in first observation | Number of polyps detected in second observation | Number of polyps overall | PMR with FV | PMR with FUSE | |
|---|---|---|---|---|---|
| FV-first group (27 participants) | 27 | 12 | 39 | 12/39 (30.7%)* | |
| FUSE-first group (23 participants) | 28 | 1 | 29 | 1/29 (3.4%)* |
PMR: polyp miss rate, FV: forward-viewing colonoscopy, FUSE: full-spectrum endoscopy.
*P=0.025
| First observation in FV-first group (FV) | Second observation in FV- first group (FUSE) | First observation in FUSE-first group (FUSE) | Second observation in FUSE-first group (FV) | |
|---|---|---|---|---|
| Detecting camera | ||||
| Left | — | 4 (33.3%) | 2 (7.1%) | — |
| Center | 27 (100.0%) | 7 (58.3%) | 26 (92.9%) | 1 (100.0%) |
| Right | — | 1 (8.3%) | 0 (0.0%) | — |
| Polyp size | ||||
| 1–5 mm | 23 (85.2%) | 12 (100.0%) | 21 (75.0%) | 1 (100.0%) |
| 6–9 mm | 3 (11.1%) | 0 (0.0%) | 5 (17.9%) | 0 (0.0%) |
| ≥10 mm | 1 (3.7%) | 0 (0.0%) | 2 (7.1%) | 0 (0.0%) |
| Shape | ||||
| Is | 17 (63.0%) | 11 (91.7%) | 16 (57.1%) | 0 (0.0%) |
| Isp | 0 (0.0%) | 0 (0.0%) | 3 (10.7%) | 0 (0.0%) |
| IIa | 10 (37.0%) | 1 (8.3%) | 9 (32.1%) | 1 (100.0%) |
| Location | ||||
| Cecum | 5 (18.5%) | 1 (8.3%) | 6 (21.4%) | 0 (0.0%) |
| Ascending colon | 12 (44.4%) | 4 (33.3%) | 18 (69.2%) | 0 (0.0%) |
| Transverse colon | 10 (37.0%) | 7 (58.3%) | 4 (14.3%) | 1 (100.0%) |
Values are shown as number (percentage).
FV: forward-viewing colonoscopy, FUSE: full-spectrum endoscopy.
Table 4 shows that polyp detection rates in the first observation were not significantly different between groups (11/23 [47.8%] with FUSE vs 9/27 [33.3%] with FV; P=0.667), according to per-participant analysis. This finding suggests that the two groups did not have significantly different numbers of colonic polyps.
| Total number of participants | Number of participants with polyps detected | PDR | |
|---|---|---|---|
| FV-first group | 27 | 9 | 33.3%* |
| FUSE-first group | 23 | 11 | 47.8%* |
PDR: polyp detection rate, FV: forward-viewing colonoscopy, FUSE: full-spectrum endoscopy.
*P=0.677
Table 5 shows the endoscopic observation time during the first and second observations in the FV-first and FUSE-first groups. There was no significant difference in observation time between FV and FUSE for the first observation (P=0.915). Furthermore, in the FV-first group, there was no difference in observation time between FV in the first observation and FUSE in the second observation (P=0.290). However, the observation time with FV in the second observation was significantly shorter than observation time with FUSE in the first observation (P<0.001). For patients in whom new polyps were detected with FUSE during the second observation, the observation time was significantly longer than that for patients with no polyps detected (P=0.001).
| First observation in FV-first group (FV) | Second observation in FV-first group (FUSE) | First observation in FUSE-first group (FUSE) | Second observation in FUSE-first group (FV) | |
| Endoscopic observation time, s | 287.7±109.3*,** | 272.6±139.7*,** | 299.1±154.2*,*** | 159.6±76.6*** |
| Participants with newly detected polyps in the second observation with FUSE | Participants without newly detected polyps in the second observation with FUSE | |||
| Mean observation time, s | 373.5±126.0† | 222.2±119.5† | ||
Values are shown as the mean±standard deviation.
FV: forward-viewing colonoscopy, FUSE: full-spectrum endoscopy.
*P=0.915, **P=0.290, ***P<0.001, †P=0.001
There were no adverse events related to colonoscopy, preparation, or sedation in this study.
FUSE represents a technological advancement for colonoscopy and could improve the efficacy of colorectal cancer screening and surveillance.7 This study provides evidence indicating higher colonic polyp detection rates with FUSE, especially in the right colon. Gralnek et al. reported that the PMR and adenoma miss rate (AMR) were significantly lower with FUSE than with standard FV colonoscopy in an international, multicenter, randomized, tandem colonoscopy trial.7 Among the 185 patients randomly allocated to either a standard FV-first group or a FUSE-first group in that study, the PMR was 43% for the standard FV-first group vs 10% for the FUSE-first group (P<0.001). Furthermore, the colorectal AMR was 41% in the standard FV-first group vs 7% in the FUSE-first group (P<0.001). These findings are consistent with those of the present study. Gralnek et al. also reported that among 20 adenomas missed with standard FV colonoscopy, most were characterized as sessile polyps (90%), were smaller than 6 mm in size (70%), and were located in the right colon (70%).7 This finding that FUSE lowered the PMR was consistent with the present study. In contrast, Ito et al.5 and Hassan et al.10 reported no significant difference in either adenoma detection rate (ADR)11 or PDR between FUSE and standard FV colonoscopy. Hassan et al. performed random allocation of 672 patients to a standard FV group or a FUSE group and found no significant difference in ADR between the standard FV group (45.5%) and the FUSE group (43.6%); however, no tandem crossover study was performed.10 Ito et al. compared a FUSE group (n=130) and a matched-control group that underwent standard colonoscopy (n=260) and reported no significant differences between the groups in PDR (68.3% vs 71.2%, respectively; P=0.567), ADR (63.4% vs 58.5%, respectively; P=0.355), or mean number of adenomas per colonoscopy (1.4 vs 1.4, respectively; P=0.917).5 However, no tandem crossover study was performed. The PMR is associated with the PDR or ADR. Hassan et al. reported an ADR of 45.5% in the standard FV group and 43.6% in the FUSE group. Ito et al. reported a PDR of 68.3% and an ADR of 63.4% in the standard FV group, whereas the FUSE group had a PDR of 71.2% and an ADR of 58.5%. The present study found a PDR of 33.3% in the FV-first group and 47.8% in the FUSE-first group; these rates are lower than or similar to the former two reports. This lower PDR or ADR may be attributed to differences in participant populations or to the quality of colonoscopic procedures, which depends on endoscopists. However, the present study was a randomized crossover trial of high evidence level, equivalent to that of Gralnek’s study,7 suggesting that FUSE achieves a higher PDR than FV colonoscopy. The reason approximately 60% of missed polyps were found with the center camera of FUSE during the second observation is unclear, but might be ascribed to changes in the location and rotation of the proximal colon resulting from two insertions or to differences in scope position between the first and second observation. Approximately 60% of missed polyps were located in the transverse colon, which is not fixed by a retroperitoneal structure but is omentum-attached intraperitoneal mobile intestine.
Colonoscopy and sigmoidoscopy are associated with a reduced incidence of cancer of the distal colorectum; colonoscopy is also associated with a modest reduction in the incidence of proximal colon cancer. Screening colonoscopy and sigmoidoscopy are both associated with reduced colorectal-cancer mortality; only colonoscopy is associated with reduced mortality from proximal colon cancer.12 The National Polyp Study of the United States found that colonoscopic polypectomy resulted in a lower-than-expected incidence of colorectal cancer.2,13
Interval cancer has recently emerged as an emerging entity. It is defined as colorectal adenocarcinoma that was diagnosed between the time of screening colonoscopy and the scheduled time of surveillance colonoscopy. Corley et al. reported that each 1.0% increase in the ADR was associated with a 3.0% decrease in the risk of interval cancer and a 5% decrease in the risk of a fatal interval cancer.14 Blind spots in colonoscopy are at the back of folds and along the inner curve of flexures. Interval cancer tends to be proximally located,15 presumably because the proximal colon has numerous tall folds and deep haustra. FUSE detected more polyps than FV colonoscopy in the proximal colon in this study; removal of these polyps might lead to a decreased risk of interval cancer.
The observation time was significantly longer with FUSE than with FV colonoscopy in the FUSE-first group. Ito et al. also reported that cecal intubation time and total procedure time were longer in the FUSE group than in matched controls (cecal intubation time: 8.8 min vs 5.1 min, respectively, P<0.001; total procedure time: 21.6 min vs 17.3 min, respectively, P<0.001).5 Endoscopists watch three monitors during observation with three-camera FUSE; therefore, it requires a longer time to complete FUSE than FV colonoscopy. Additionally, the examination time during the first observation was longer than that during the second observation because endoscopists needed to clear the colon of residual feces and air bubbles. However, the possibility that the difference in the PDR between FUSE and FV colonoscopy resulted from the shorter observation time with FV cannot be eliminated.
This study has some limitations, such as small sample size and observation only in the proximal colon. To avoid the burden of two consecutive insertions of FUSE from the anus to the cecum, and because there are more and taller folds in the proximal colon than in the distal colon, only the proximal colon was evaluated in the present study. Further work is needed to confirm the usefulness of FUSE for colorectal cancer screening and post-polypectomy surveillance. In conclusion, FUSE decreased the colonic PMR in the right colon in this study.
The authors report no potential conflicts of interest.
We thank the staff of the Endoscopy Center at Fujita Health University Hospital for their endoscopic assistance and the staff of the Center for Research Promotion and Support at Fujita Health University for computer-generated randomization.
