Reply to Backes et al.

We thank Backes et al. for their comments on our systematic review and meta-analysis on the diagnostic accuracy of magnifying endoscopy with narrow-band imaging (M-NBI) compared with magnifying chromoendoscopy (M-CE) in the diagnosis of deep submucosal colorectal cancers (dSMCs) [1]. Here, we respond to the issues raised by Backes et al.

We politely disagree with Backes et al. on the first issue raised. M-NBI and M-CE were two totally different diagnostic modalities. The M-NBI modality uses optical filters to highlight surface structures and mucosal microcapillaries, which can be clearly observed with magnifying endoscopy, whereas M-CE uses pit pattern for diagnosis. In addition, surface structure under M-NBI is different from pit pattern under M-CE, as the surface structure appears larger than the pit pattern [2]. In included articles with comparison between M-NBI and M-CE, diagnosis by M-CE used pit pattern only and not a combination of M-NBI surface classification with pit pattern. In fact, bias did exist for the M-CE diagnosis, as all M-CE diagnoses in the study were made after M-NBI assessment, as we mentioned in the discussion section [1], but this does not equate to a comparison of M-NBI with M-NBI + M-CE.

Our main outcome was to pool the diagnostic efficacy of M-NBI and therefore we only included studies relating to M-NBI. In this setting, studies that only investigated the diagnostic efficacy of M-CE were inevitably excluded from our study, which is not the same as selecting studies according to inclusion criteria as stated by Backes et al. [3]. We discussed this situation in our discussion section [1]. Another difference is that we also included meeting abstracts relating to M-NBI, as most full-text studies were conducted in Japan, leading to large publication bias. To reduce this publication bias, the meeting abstracts were included, but this is different from the quality concerns considered by Backes et al. [4] [5]. We have also previously discussed the diagnostic criteria for M-CE with Backes et al. [5]. Type V indicates colorectal cancer and is divided into Vi and VN (easily detected). Vi could also be divided into two types, mainly Kudo pit-pattern type Vi (high risk) and Vi, as Backes et al. stated in their review [3]. The main problem is that some earlier studies used only the VN pattern for diagnosis of dSMCs. This actually underestimated the sensitivity of M-CE for diagnosing dSMCs. Bearing this in mind, we conducted a similar analysis, including only studies with direct comparison between M-NBI and M-CE, and found a trend of preference for M-CE over M-NBI in the diagnosis of dSMCs [5]. This is an important finding to emphasize.

For the third issue raised by Backes et al., it is better to use negative predictive value (NPV) for interpretation. A high NPV would enable an endoscopist to accurately diagnose superficial submucosal colorectal cancers (sSMCs) or intramucosal neoplasms, leaving dSMCs for surgery. In our study, a higher NPV could be gained for M-CE than M-NBI, thus M-CE allowed a higher percentage of sSMCs among the predicted sSMCs to undergo endoscopic resection [1].

Based on the above discussion, we still recommend the use of M-CE for the re-revaluation of invasion depth; however, following careful consideration of the comments made by Backes et al., it would be better to remove the qualifying clause “to avoid unnecessary surgical therapy” in order to avoid confusion.

^* These author contributed equally to this paper.

• References

• 1 Zhang QW, Teng LM, Zhang XT. et al. Narrow-band imaging in the diagnosis of deep submucosal colorectal cancers: a systematic review and meta-analysis. Endoscopy 2017; 49: 564-580
• 2 Yao K, Anagnostopoulos GK, Ragunath K. Magnifying endoscopy for diagnosing and delineating early gastric cancer. Endoscopy 2009; 41: 462-467
• 3 Backes Y, Moss A, Reitsma JB. et al. Narrow band imaging, magnifying chromoendoscopy, and gross morphological features for the optical diagnosis of T1 colorectal cancer and deep submucosal invasion: a systematic review and meta-analysis. Am J Gastroenterol 2017; 112: 54-64
• 4 Backes Y, Moss A, Reitsma JB. et al. Response to Zhang et al. Am J Gastroenterol 2017; 112: 514-515
• 5 Zhang QW, Zhang XT, Zhou Y. et al. A trend of preference of magnifying chromoendoscopy to narrow band imaging in diagnosis of T1 colorectal cancer with deep invasion could be observed. Am J Gastroenterol 2017; 112: 512-514