A novel three-dimensional endoscopic ultrasound technique for the freehand examination of the oesophagus

This study details the development and evaluation of a freehand radial three-dimensional endoscopic ultrasound (3D-EUS) technique for use in the upper gastro-intestinal tract. It comprised of a commercial EUS system, a custom acquisition system to simultaneous acquire radial B-mode images and corresponding incremental changes in position of the scope as it was withdrawn and a custom 3D-EUS package written in Matlab™, to reconstruct and analyse the volume. This technique was evaluated using an EUS phantom with embedded objects of known dimensions and volumes and with clinical images acquired during routine cancer staging. For the phantom measurements, average Z-dimensional error was <1% and volume errors were 1.4% (volume(1) = 48930 mm(3)) and 4.5% (volume(2) = 5100 mm(3)). Application of this technique to EUS acquired clinical images produced excellent characterisation of oesophageal structures and accurate dimension and volume measurements. It also enabled the endoscopist to review "off-line" the EUS examination in case important information was missed.

Is endoscopic ultrasonography indispensable in patients with early gastric cancer prior to endoscopic resection?

BACKGROUND

Endoscopic resection as curative treatment of early gastric cancer (EGC) requires accurate staging for depth of invasion (T) before therapy. This study aimed to compare T staging of EGC using a miniprobe with that of conventional endoscopy (CE).

METHODS

A total of 388 patients with suspected of EGC by CE were prospectively enrolled in the study. After miniprobe staging by an experienced endosonographer, CE staging was performed by another endoscopist who was blinded to the miniprobe assessment. Patients underwent either endoscopic resection (n = 325) or surgery (=63) according to staging results. Results of each method were compared with the histology of the resected specimen. Clinicopathological factors affecting accuracy of each test were also evaluated.

RESULTS

Overall accuracy of miniprobe and CE staging was 78.9% (306/388) and 81.4% (316/388), respectively (p = 0.052). Sensitivity for T1m was more than 95% (miniprobe 98.7%, CE 97.7%), whereas sensitivity for T1sm was lower at 6.6% for miniprobe and 23.7% for CE (p = 0.002). Among nine lesions identically considered submucosal invasion by both methods, three were T1sm cancer. Diagnostic accuracy of the miniprobe was lower than that of CE with respect to lesions located in the lower third of the stomach, differentiated histology, or massive submucosal invasion.

CONCLUSIONS

Overall accuracy of the miniprobe and CE for T staging in EGC was approximately 80% with no significant difference. Accurate prediction of both techniques for submucosal invasion was poor, even by using a combination of the two methods. The use of the miniprobe may not increase the positive predictive value for T staging over the use of CE alone.

The accuracy of endoscopic ultrasonography in differentiating mucosal from deeper gastric cancer

BACKGROUND AND AIM

With the advent of endoscopic mucosal resection and endoscopic submucosal dissection techniques, pretreatment differentiation between mucosal and deeper gastric cancer has become increasingly important. The aim of this study was to systematically review published data on the diagnostic performance of endoscopic ultrasonography (EUS) in differentiating between mucosal and deeper gastric cancer.

METHODS

A systematic search for relevant studies was performed of the PubMed/MEDLINE and Embase databases. Two reviewers independently assessed the methodological quality of each study. Sensitivity and specificity of EUS were calculated for each study and a summary receiver operating characteristic curve (sROC) was constructed. Subgroup analyses were performed if results of individual studies were heterogeneous.

RESULTS

The inclusion criteria were met by 18 studies. The studies had poor methodological quality. Sensitivity and specificity of EUS in detecting cancerous extension beyond the mucosa ranged from 18.2 to 100% (median 87.8%) and from 34.7 to 100% (median 80.2%), with an area under the sROC of 0.8924. There was significant heterogeneity in both sensitivities and specificities among the included studies (P < 0.0001). Studies which only included patients endoscopically suspected of having EGC and studies which only used transducer frequencies > or =15 MHz were homogeneous in sensitivities (P= 0.1828 and 0.4501, respectively).

CONCLUSION

It is unclear yet whether EUS can accurately differentiate between mucosal and deeper gastric cancer. Factors that may influence its diagnostic performance should be further explored.