Aims Determine the most accurate endoscopic classification in comparison with forceps
biopsy(FB) for granular mixed laterally spreading lesions(LSLs), by comparing them
with histopathological findings [1]
[2]
[3]
[4].
Methods The study included 70 patients with granular mixed LSLs with a diameter≥20mm, type
0-IIa, 0-Is or IIa+Is Paris Classification. Patients with deep invasive cancer were
excluded. An expert endoscopist performed optical evaluations according to Kudo, JNET,
Modified Sano and Hiroshima classifications using chromoscopy and NBI, followed by
target forceps biopsy and endoscopic resection. The findings were compared to specimen
histopathology using Fisher’s exact test, Wilson confidence intervals and descriptive
statistics in Statistica 13.
Results Median age was 66±9.9 year; median lesion size was 40±16.7mm. 35.7% of all LSLs were
removed via piecemeal EMR, 32.9% via ESD, 17.1% by EMR en block and 14.3% by hybrid
ESD. Sensitivity was 67.39% (95%CI,53.8%-80.9%) for FB; 70.83%(95%CI,58%-83.7%) for
JNET, 57.14%(95%CI,38.8%-75.5%) for Hiroshima and 89.58%(95%CI,80.9%-98.2%) for Modified
Sano classification. Specificity was 95.83% (95%CI,87.8%-100%) for FB; 90.91%(95%CI,78.9%-100%)
for JNET; 95.24%(95%CI,88.8%-100%) for Hiroshima and 54.55%(95%CI,33.7%-75.4%) for
Modified Sano classification. P<0.001 in all cases. In the analysis using the Kudo
classification, the p-value was 0.103, indicating that the results were not statistically
significant.
Conclusions Among the classifications tested, JNET has slightly higher sensitivity and slightly
lower specificity compared to forceps biopsy. Given that the Sano classification demonstrates
significantly higher sensitivity but lower specificity, it may be beneficial to combine
both JNET and Sano classifications to enhance diagnostic accuracy. Comparable results
between optical assessment and forceps biopsy indicate that forceps biopsy is unnecessary
if endoscopic removal of the lesion is planned.
