Computer-aided diagnosis system using only white-light endoscopy for the prediction of invasion depth in colorectal cancer.

BACKGROUND AND AIMS: Endoscopic treatment is recommended for low-grade dysplasia (LGD), high-grade dysplasia (HGD), and colorectal cancer (CRC) with submucosal (SM) invasion <1000 µm. However, diagnosis of invasion depth requires experience and is often difficult. This study developed and evaluated a novel computer-aided diagnosis (CAD) system to determine whether endoscopic treatment is appropriate for colorectal lesions using only white-light endoscopy (WLE). METHODS: We extracted 3442 images from 1035 consecutive colorectal lesions (105 LGDs, 377 HGDs, 107 CRCs with SM <1000 µm, 146 CRCs with SM ≥1000 µm, and 300 advanced CRCs). All images were WLE, nonmagnified, and nonstained. We developed a novel CAD system using 2751 images; the remaining 691 images were evaluated by the CAD system as a test set. The capability of the CAD system to distinguish endoscopically treatable lesions and untreatable lesions was assessed and compared with the results from 2 trainees and 2 experts. RESULTS: The CAD system distinguished endoscopically treatable from untreatable lesions with 96.7% sensitivity, 75.0% specificity, and 90.3% accuracy. These values were significantly higher than those from trainees (92.1%, 67.6%, and 84.9%; P < .01, <.01, and <.01, respectively) and were comparable with those from experts (96.5%, 72.5%, and 89.4%, respectively). Trainees assisted by the CAD system demonstrated a diagnostic capability comparable with that of experts. CONCLUSIONS: The CAD system had good diagnostic capability for making treatment decisions for colorectal lesions. This system may enable a more convenient and accurate diagnosis using only WLE.

Comparison of endoscopic visibility and miss rate for early gastric cancers after Helicobacter pylori eradication with white-light imaging versus linked color imaging.

BACKGROUND AND AIM: We aimed to investigate whether linked color imaging (LCI) improves endoscopic visibility of early gastric cancers (EGC) after Helicobacter pylori eradication, which are often difficult to detect, and reduces the miss rate when compared with white-light imaging (WLI). METHODS: The visibility study used two images, one each with WLI and LCI, from 84 consecutive EGC after H. pylori eradication. Endoscopic visibility was evaluated using a visibility score and color difference (CD) value. To analyze miss rates, we studied a library of recorded videos using both WLI and LCI for 70 other consecutive patients after H. pylori eradication, among whom 19 had EGC. Endoscopic screening was done using the same protocol to map the entire stomach. Six endoscopists reviewed the videos in a randomized order. Miss rates of EGC were compared among the modalities. RESULTS: Mean [(±standard deviation) visibility scores with LCI were significantly higher than those with WLI (3.19 ± 0.84 vs 2.52 ± 0.98, P < 0.001), as were mean CD values (26.3 ± 9.1 vs 13.6 ± 6.3, P < 0.001). Miss rates of the six endoscopists were significantly lower with LCI than with WLI (30.7% vs 64.9%, P < 0.001). Both expert and trainee endoscopists had significantly better results with LCI than with WLI. CONCLUSIONS: Linked color imaging significantly improved the visibility of EGC after H. pylori eradication compared with WLI using both subjective and objective criteria. Furthermore, LCI significantly reduced miss rates of these lesions compared with WLI.

Skill-up study of systemic endoscopic examination technique using narrow band imaging of the head and neck region of patients with esophageal squamous cell carcinoma: Prospective multicenter study.

OBJECTIVE: To evaluate the usefulness of a training program on endoscopic head and neck surveillance for beginner endoscopists. METHODS: This prospective multicenter study included 13 beginner endoscopists from 10 institutions who received training in systematic observation techniques and diagnostic criteria, and the training involved hands-on learning. Between May 2016 and February 2017, enrolled patients with current or previously diagnosed esophageal squamous cell carcinomas underwent head and neck surveillance using narrow band imaging (NBI) endoscopy, and histologically confirmed head and neck squamous cell carcinoma (HNSCC) detection rates, endoscopic image quality, and examination times were compared before (group A) and after (group B) the training program. Maximum possible score for the endoscopic images was 30 points. RESULTS: A total of 330 patients, comprising 181 in group A and 149 in group B, were enrolled. Three patients with HNSCC were detected in group A (1.7%) and in group B (2.0%; P = 1.000). Mean ± standard deviation (SD) examination times were 157 ± 71 s and 174 ± 109 s in groups A and B, respectively, (P = 0.073). Mean ± SD scores of the endoscopic images were 25.04 ± 5.47 points and 27.01 ± 4.35 points in groups A and B, respectively, (P < 0.001). CONCLUSION: The HNSCC detection rate based on the use of NBI on patients with ESCC did not improve after the training program for beginner endoscopists; however, endoscopic image quality improved significantly after the training program.

Magnified endoscopic observation of small depressed gastric lesions using linked color imaging with indigo carmine dye.

BACKGROUND AND STUDY AIMS: Magnifying linked color imaging with indigo carmine dye (M-Chromo-LCI) enables sterically enhanced and color image-magnified observation of the superficial gastric mucosa. This study investigated the usefulness of M-Chromo-LCI for the differential diagnosis of gastric lesions. PATIENTS AND METHODS: 100 consecutive small depressed lesions were examined with conventional white-light imaging (C-WLI), magnifying blue-laser imaging (M-BLI), and M-Chromo-LCI. Endoscopic images were reviewed by three experts and three non-experts. Diagnostic accuracy and interobserver agreement were compared among the modalities. RESULTS: For experts, M-BLI showed a significantly higher diagnostic accuracy than C-WLI (82.7 % vs. 67.0 %; P < 0.001). The diagnostic accuracy of M-Chromo-LCI was not different from M-BLI (87.7 % vs. 82.7 %; P = 0.31). For non-experts, M-BLI showed a significantly higher diagnostic accuracy than C-WLI (69.3 % vs. 52.3 %; P < 0.001). M-Chromo-LCI additionally showed a significantly higher diagnostic accuracy than M-BLI (79.7 % vs. 69.3 %; P = 0.005). M-Chromo-LCI had the highest interobserver agreement for each group. CONCLUSIONS: M-Chromo-LCI is expected to become a useful modality for the accurate diagnosis of gastric lesions.

Usefulness of close observation with non-magnified blue laser imaging for determining cold polypectomy indications.

PURPOSE: To examine the usefulness of non-magnified close observation with blue laser imaging (BLI) using a colonoscope with close observation capability in determining indications for cold polypectomy. METHODS: We conducted an image evaluation study on 100 consecutive colorectal lesions of 10 mm or less which were observed endoscopically without magnification using BLI mode prior to treatment. Two experts and two non-experts reviewed the images using the Japan NBI expert team (JNET) classification and the diagnostic accuracy was analyzed. RESULTS: The final pathological diagnoses of the 100 lesions were hyperplastic/sessile serrated polyp (HP/SSP), low grade dysplasia (LGD), high grade dysplasia (HGD) and deep submucosal invasive cancer (dSM), respectively, in 12, 79, 9 and 0 lesions. When JNET classification type 1 corresponds to HP/SSP; 2A to LGD; 2B to HGD; and 3 to dSM; the overall diagnostic accuracy was 84.3%. Accuracy was 90.5% for experts and 78.0% for non-experts. High confidence rate was 67.5% for experts and 48.0% for non-experts. In diagnostic accuracy for HGD, the sensitivity, specificity, PPV and NPV were, respectively, 77.8%, 98.9%, 87.5% and 97.8% for experts; and 66.6%, 92.3%, 46.2% and 96.6% for non-experts. CONCLUSIONS: The diagnostic accuracy of unmagnified close observation with BLI using a colonoscope with close observation capability is similar to that reported for magnifying endoscopy and is useful in predicting the histological diagnosis of colorectal polyps of 10 mm or less although the effectiveness may be limited for non-experts. This modality is a potentially useful tool in deciding indications for cold polypectomy.

Feasibility of endoscopic submucosal dissection for cecal lesions.

BACKGROUND: The treatment results of endoscopic submucosal dissection (ESD) for colorectal lesions have improved markedly, but some lesions remain difficult to treat. Hence the cecum is considered a technically challenging site for ESD. We examined the feasibility of ESD for cecal lesions. METHODS: Among a total of 708 colorectal ESD performed in our hospital between March 2006 and December 2016, 549 procedures performed after April 2012 were studied, at a time when the techniques of ESD had stabilized and the procedure was covered by health insurance in Japan. Among 549 cases, 61 were cecal lesions and 488 were noncecal lesions. The treatment outcomes were analyzed. RESULTS: For cecal lesions, the en bloc resection rate was 95.1%, R0 resection rate was 91.8%, perforation rate was 0%, delayed bleeding rate was 6.6%, median diameter of resected specimen was 32 mm (16-65 mm), median time of the procedure was 44 minutes (8-140 min). The corresponding results for noncecal lesions were 97.3%, 95.5%, 0.4%, 2.7%, 30 mm (10-109 mm), and 37 min (7-225 min). No significant differences were observed and the good treatment results were seen. When the outcomes were analyzed for cecal sites considered to be particularly challenging; proximity to appendiceal orifice, the ileocecal valve, and the bottom of cecum, the treatment results were not inferior to other sites. CONCLUSIONS: ESD is safe and effective even for cecal lesions considered challenging to treat. ESD is feasible for cecal lesions.

Linked-color imaging improves endoscopic visibility of colorectal nongranular flat lesions.

BACKGROUND AND AIMS: As a newly developed image-enhanced endoscopy (IEE) technique, linked-color imaging (LCI) provides very bright images with enhanced color tones. With the objective of improving the detection rate of colorectal flat tumor lesions, which are difficult to detect, we examined the usefulness of LCI from the viewpoint of visibility. METHODS: Fifty-three consecutive nongranular flat tumors were used in this study. Endoscopic images were acquired by white-light imaging (WLI), blue-laser imaging (BLI)-bright, and LCI modes. For each lesion, we selected 1 image each acquired by WLI, BLI-bright, and LCI modes. Six endoscopists interpreted the images. By using a previously reported visibility scale, we scored the visibility level on a scale of 1 to 4. RESULTS: The mean (± standard deviation) visibility scores were 2.74 ± 1.08 for WLI, 2.94 ± 0.97 for BLI-bright, and 3.36 ± 0.72 for LCI. The score was significantly higher for BLI-bright compared with WLI (P < .001) and again higher for LCI compared with BLI-bright (P < .001). When we compared between experts and trainees, the corresponding scores of experts were 2.83 ± 1.06, 3.17 ± 0.88, and 3.40 ± 0.74, with a tendency similar to the scores of all endoscopists. For the trainees, there was no difference between the scores for WLI (2.65 ± 1.10) and BLI-bright (2.71 ± 1.00), but the score for LCI (3.31 ± 0.69) was significantly higher than that for WLI or BLI-bright (P < .001). When only sessile serrated adenoma/polyp lesions were analyzed, LCI remained significantly higher than the other 2. CONCLUSIONS: The present findings suggest that LCI increases the visibility of colorectal flat lesions and contributes to improvement of the detection rate for these lesions.

Usefulness of cold polypectomy under linked color imaging.

Background and study aims Cold polypectomy is becoming popular for treatment of colon polyps due to its safety and convenience, but there is still the problem of tumor remnants. Because linked color imaging (LCI) improves polyp visibility, cold polypectomy under LCI is anticipated to reduce the tumor remnant rate. Therefore, we investigated the usefulness of this procedure. Patients and methods Fifty patients scheduled to undergo cold polypectomy for treatment of colon polyps < 10 mm and assumed to be adenomas were registered prospectively. After performing cold snare polypectomy (CSP) under LCI, biopsy was performed at two resection margin sites for each polyp to determine the tumor remnant rate. Results A total of 145 lesions were treated by CSP. Of the 139 lesions in which polyps were retrievable and diagnosed as adenomas pathologically, one lesion was recognized as a remnant adenoma on biopsy (remnant rate: 0.7 % [95 % CI: 0.0-4.4]). This remnant rate was extremely low. Treatment results were extremely promising given that en bloc resection, post-procedure bleeding, and perforation rates were 100 %, 0 %, and 0 %, respectively. Conclusion Cold snare polypectomy under LCI may be an effective treatment method capable of reducing the tumor remnant rate. This trial was approved by our Institutional Ethics Committee and registered at the University Hospital Medical Information Network (UMIN 000033690).

Successful Neurological Recovery with Multimodality Therapy without Surgery for Spinal Metastases from Advanced Gastric Cancer.

Advanced gastric cancer with bone metastasis has a very poor prognosis with short median survival. To the best of our knowledge, no reports in literature have described extensive recovery of paralysis with multimodality treatment without surgery in these cases. This report describes the case of a 52-year-old severely paralyzed female patient with spinal metastasis from advanced gastric cancer. She was inoperable, owing to a large thrombus in the inferior vena cava; alternative multimodality treatments, including chemotherapy and radiotherapy, were administered. The paralysis and the bladder and rectal dysfunction improved considerably. In addition, the performance status (PS) and Frankel grade also improved dramatically, from 4 to 1 and grade B to D, respectively. At 1 year after initiation of treatment, she is ambulatory. Patients with poor PS are often offered palliative therapy. However, this case demonstrates that poor PS solely due to paralysis from spinal metastasis may necessitate multimodality treatment instead of palliative care.

Early gastric cancer diagnostic ability of ultrathin endoscope loaded with laser light source.

BACKGROUND: Conventionally, the low luminous intensity, low image resolution, and difficulty in operation have been reported with the ultrathin endoscope. However, it has markedly advanced recently. The improvement of the diagnostic ability is expected. AIM: To compare the early gastric cancer diagnostic ability of an ultrathin endoscope loaded with a laser light source and that of the conventional endoscope. METHODS: The target subjects were 375 consecutive patients who underwent endoscopy at our hospital for post-endoscopic submucosal dissection follow-up of gastric cancer from January to August 2018. During endoscopy, the ultrathin endoscope was used in 140 patients (37.3%), and the conventional endoscope was used in 235 patients (62.7%). Patient background was adjusted using the propensity score matching method, and gastric cancer detection ability was evaluated in the two groups. RESULTS: The gastric cancer detection rate was 7.8% in the ultrathin endoscope group and 7.0% in the conventional endoscope group, and the mean intragastric observation time was 4.1 ± 1.7 min in the ultrathin endoscope group and 4.1 ± 1.9 min in the conventional endoscope group, showing no significant differences between the groups. Moreover, the biopsy implementation rate was 31.8% in the ultrathin endoscope group and 41.1% in the conventional endoscope group, and the biopsy prediction rate was 17.9% and 13.2%, respectively, showing no significant differences between the groups. CONCLUSION: The gastric cancer diagnostic ability of the ultrathin endoscope loaded with a laser light source was comparable to that of the conventional endoscope. The observation time was also comparable. Thus, endoscopy using the ultrathin endoscope loaded with the laser light source would be the first option in screening examinations of gastric cancer due to its low invasion.

Linked color imaging improves the endoscopic visibility of gastric mucosal cancers.

Background and study aims As a newly developed endoscopy technique, linked color imaging (LCI) provides very bright images with enhanced color tones. With the objective of improving the detection rate of gastric mucosal cancers, which are often difficult to detect, we examined the utility of LCI from the viewpoint of visibility. Patients and methods The current study used 100 consecutive gastric mucosal cancers ≤ 20 mm in diameter. For each lesion, we selected one endoscopic image acquired by white-light imaging (WLI), blue-laser imaging (BLI) -bright, and LCI modes. Four endoscopists interpreted the images; using a previously reported scale, we scored the visibility level on a scale of 1 - 4. Results The mean (±â€ŠSD) visibility scores were 2.54 ±â€Š1.10 for WLI, 3.02 ±â€Š1.07 for BLI-bright, and 3.28 ±â€Š0.97 for LCI. The score was significantly higher for BLI-bright compared with WLI ( P  < .001) and again higher for LCI compared with BLI-bright ( P  < .001). For the experts, the scores for BLI-bright and LCI were similar, but both were significantly higher than the score for WLI. For the trainees, there was no significant difference between the WLI and BLI-bright scores, but LCI score was significantly higher than those for WLI and BLI-bright scores. With regard to clinical characteristics, LCI particularly enhanced visibility of normochromic, flat and depressed lesions, which had the lowest visibility scores of all three modalities compared with those of the other lesions. Conclusion LCI increased visibility and may contribute to early detection of gastric mucosal cancers.

Enrichment technique to allow early detection and monitor emergence of KRAS mutation in response to treatment.

Sensitivity of cell-free circulating tumour DNA (ctDNA) assays is often hampered by the limited quantity of intact mutant nucleotide fragments. To overcome the issue of substrate limitation in clinical applications, we developed an enrichment method utilizing pyrrole-imidazole (PI) polyamides and their ability to bind the minor groove of B-DNA. We present here a proof-of-concept experiment to enrich specific mutant KRAS alleles with biotinylated PI polyamides. We investigated the clinical feasibility of incorporating PI polyamides to detect KRAS mutations in ctDNA from 40 colorectal cancer (CRC) patients, of whom 17 carried mutations in KRAS. After enriching ctDNA with those polyamides, we used digital PCR to detect several common KRAS codon 12 mutations. Enrichment by biotinylated PI polyamides improved the sensitivity of ctDNA analysis (88.9% vs. 11.1%, P < 0.01) in 9 non-metastatic mutation-positive patients. We observed no differences in performance for the 8 metastatic subjects (100% vs. 75%, P = 0.47). In the remaining 23/40 patients with wild type KRAS codon 12, no mutant alleles were detected with or without polyamide-facilitated enrichment. Enriching B-form of ctDNA with PI polyamides significantly improved the assay sensitivity in detecting KRAS mutations in non-metastatic CRC patient samples.