One recently published Chinese study concluded that MCCE was able to detect GCs in a large population, but its role as a first line screening tool for GC remains to be further validated [9]. Since the risk of GC is closely related to H. pylori infection status, MCCE’s diagnostic accuracy on H. pylori infection status is of critical importance in risk stratification. Moreover, the morphological features of early GC or high-grade precancerous lesions also differ according to different H. pylori infection status, which further established the rationale for our study.
Yoshii et al demonstrated that overall diagnostic accuracy of three H. pylori infection status was 82.9% on white light endoscopy by the Kyoto classification of gastritis[8]. In this study, we found that most of the key findings documented in Kyoto classification of gastritis were well recognizable on MCCE, and H. pylori infection status could be accurately diagnosed via MCCE, and the overall diagnostic accuracy was 80.2% comparable with EGD.
Previous studies demonstrated that MCCE was capable of detecting various kinds of gastric lesions, including erosions, polyps, ulcers, and even superficial early gastric cancers [5, 6, 9, 13]. In our study, based on the results of phase one, we found that the Kyoto classification of gastritis generally applied well on MCCE in the diagnosis of the H. pylori infection status.
In the diagnosis of current infection status, the most reliable finding was mucosal swelling (sensitivity 76.5%, specificity 88.7%, PPV 80.2%), whereas in other recently published EGD studies, that diagnosis was established mainly based on observation of diffusive redness. This difference, we speculate, might have been the reason why MCCE had a much higher DOR for current infection compared with conventional EGD (77.2 vs 21.7) [7, 13, 14]. Additionally, when spotty redness and mucosal swelling were simultaneously observed, the specificity, PPV and DOR were 95.8%, 85.0% and 15.1, respectively; this combination of findings strongly indicates a diagnosis of current infection.
At the moment, in the midst of the COVID-19 pandemic, telemedicine has recently started to gain popularity[15, 16]. With high diagnostic accuracy for current H. pylori infection, besides inspecting the gastric mucosa, MCCE could also occasionally offers an ideal alternative for detecting active H. pylori infection in a safe and contactless fashion.
MCCE can reliably diagnose non-infection status, with sensitivity, specificity and PPV of 83.8%, 85.0% and 82.2%, respectively. This diagnosis is mainly based on observation of RAC; although FGP and streak redness were also of high specificity and PPV, these two findings were relatively uncommon. However, MCCE’s DOR for non-infection status in our study was much lower compared with that of Yoshii’s EGD study (30.7 vs 98.6) in which the authors made the diagnosis based on the same findings. Kyoto gastritis classification defines RAC as micro-vascular networks observed in lower part of the gastric corpus, mainly the lesser curve side[7, 12]. For MCCE, however, due to its steering pattern, sometimes it is difficult to confirm the location of micro-vascular network observed during real-time procedure or on still images. A number of current-infection individuals in this study might have been misdiagnosed as non-infection because RAC were thought to be observed in the gastric fundus or upper gastric corpus.
MCCE’s diagnostic performance on past-infection status was suboptimal in our study, largely due to the lack of specific findings. Besides, inter-observer variability might also have played role in its low diagnostic performance. Occasionally, intestinal metaplasia and map-like redness may resemble each other on MCCE, although these two findings could be differentiated at ease by a veteran on EGD, but it might have been difficult for a non-experienced physician to tell them apart on MCCE. That might have been the major cause of inter-observer variability. When map -like redness was used as the main predictor for past-infection, the sensitivity and PPV were 41.9% and 53.5% respectively. A new discovery in our study was that the combination of RAC and map like redness could be used as a highly specific predictor for past-infection, with specificity, PPV and DOR of 98.9%,86.7% and 39.4, respectively. This combination of findings is especially helpful for determining past-infection status when there is diagnostic ambiguity.
Our study had several strengths. First, in phase one, we made a direct comparison between EGD and MCCE using EGD findings as gold standard to validate Kyoto classification of gastritis’ applicability on MCCE. Second, this was a prospective study in which the reviewers were blinded to the final results and we used UBT results as the gold standard for the diagnosis of H. pylori infection, making the results reliable and robust. Third, we have found several combinations of findings to have high diagnostic value; this is of utility when the diagnosis was uncertain based on observation of a single finding. Further, we performed regression analyses in which the diagnostic performance of MCCE was assessed by combining 10 findings in Kyoto classification of gastritis. Fourth, in phase two, we had an expert and a non-expert reviewed MCCE images, and resolved inter-observer disagreement by a referee, thus making our results reproducible in future studies.
Our study had several limitations too. First, despite the fact that all participants in phase two were prospectively recruited, approximately half of the included participants were H. pylori non-infected (44.5%), while the proportion of past-infection participants was particularly low (15.9%), thus according to STARD (Standards for reporting of diagnostic accuracy studies), selective bias was inevitable[17]. Second, according to Kyoto classification of gastritis, sticky mucus and hyperplastic polyps are also key findings for H. pylori infection, but these findings were not found in phase one, nor could we rate the degrees of atrophy and intestinal metaplasia on MCCE, so the scoring system of Kyoto classification of gastritis described in previous studies [22, 23, 24] was not available in this study. Therefore the Kyoto classification of gastritis used in phase two was actually a modified version[7, 8, 9, 12, 18]. Third, spontaneous eradication of H. pylori might have been occurred in a tiny portion of the study participants, and that could have had an impact on the evaluation of the diagnostic accuracy, and this issue might have been underestimated in our study [7, 19]. Last but not least, we didn't directly compare the diagnostic performances between EGD and MCCE due to the relatively small sample size in phase one.
In future studies, more specific findings for past-infection are warranted, because using map-like redness as the predictor doesn’t appear to have sufficient diagnostic power22. Additionally, in recent years, the introduction of artificial intelligence (AI) has improved the diagnostic accuracy of GI neoplasms as well as EGD’s diagnostic accuracy on H. pylori infection status[18, 20]. Hopefully, our results could help establishing MCCE’s AI diagnosis on H. pylori infection status, thus improving GC’s early detection in a more comfortable way[20, 21]. Moreover, efforts to establish scoring models for atrophy and intestinal metaplasia on MCCE are required, and that may help us better stratifying GC risks via MCCE[22, 23, 24].