Aims Subepithelial lesions (SELs) are common and their diagnosis is usually not possible
through regular endoscopy and biopsies. As deep biopsies may be associated with severe
bleeding, patients are usually referred for endoscopic ultrasound (EUS). EUS is performed
to characterize the lesion, its size, structure, location in the wall and vascularity.
EUS-fine needle aspiration or biopsy (FNA/FNB) may enable sampling and diagnosis.
However, EUS is highly operator dependent, expensive and is only performed in specialized
centers. EUS investigations of SEL represent a significant burden for EUS operations.
This causes long waiting lists for EUS and increases patient’s anxiety. Additionally,
diagnostic accuracy of EUS is usually dependent on tissue sampling, being poor in
small or polypoid lesions. There is the need for another approach for the assessment
of SELs. The ideal solution should be safe, easy to perform, cheaper, non-highly operator
dependent and available at first endoscopy (at any center). The doppler endoscopic
prove (DEP) can be inserted through a conventional endoscope and enables assessment
of vascularity beneath the mucosa in 3 different depth cathegories (superficial/middle/deep).
Some studies had proven its efficacy in the management of peptic ulcer.
Methods This was a prospective blind pilot study. Patients referred for EUS due to a gastrointestinal
SEL were invited to participate. First, endoscopy was performed and the lesion was
characterized using endoscopy and DEP. The results of DEP were classified as 4 negative/superficial/middle/deep.
Then, EUS was performed, being the EUSscopist blinded to the previous endoscopic findings.
If after EUS, the lesion was considered avascular or poorly vascular, unroofing with
biopsies was performed. The primary endpoint was the diagnostic accuracy of DEP (highly
vascular or non-highly vascular) and correlation between DEP and EUS findings in terms
of SEL vascularization. Secondary endpoints were the accuracy rate and adverse events
of unroofing in the histological characterization of SELs.
Results Twenty-eight SELs from twenty-seven patients were included, corresponding 1/14/10/3
SELs in the esophagus/stomach/duodenum/colon, respectively. Regarding the primary
endpoints, highly vascular/non-highly vascular for DEP and EUS is 3/25 and 4/24, respectively.
The positive predict value (PPV)/negative predict value (NPV) for DEP is 2/3 (67%)
and 23/25 (92%), respectively. Regarding the secondary endpoints, unroofing enabled
histological diagnosis in 24/26 (92%). Four patients (4/28 14%) reported pain in the
throat. 2/3 high vascular cases for DEP were liver and gall bladder compressions and
all 25 non-highly vascular were SELs.
Conclusions This method is safe, feasible and easy to perform and might be available at any center.
DEP has a good NPV, being a good tool for the exclusion or vascular lesions. Unroofing
is also safe and associated with high accuracy in the diagnosis of SEL. Prospective
RCT are warranted before the dissemination of DEP and unroofing.
