Role of fluorescence confocal microscopy for rapid evaluation of EUS fine-needle biopsy sampling in pancreatic solid lesions.

Abstract Background and aims Endoscopic ultrasound fine-needle biopsy (EUS-FNB) is the standard procedure for the diagnosis of pancreatic lesions. Fluorescence confocal microscopy (FCM) allows imaging of tissues in the fresh state, requiring minimal preparation without damage or loss of tissue. Up to now, no data exist on FCM in the field of microhistological specimens. We aimed at assess the diagnostic performance of FCM in predicting histological adequacy of EUS-FNB samples in pancreatic solid lesions and to assess the agreement between FCM evaluation and final histology. Methods Single-center prospective study on consecutive patients with pancreatic lesions receiving EUS-FNB. Obtained samples have been evaluated at FCM and classified as “inadequate” or “adequate” (benign; suspicious; malignant). The Kappa test was used to quantify agreement. The diagnostic accuracy of FCM was assessed. A p Results From April 2020 to September 2020, 81 patients have been enrolled. In all cases FCM showed the macro image of the sample and created a digital image. The 92.6% of sample was defined adequate at the FCM evaluation and it was confirmed at histopathology. The histological diagnosis was 8% benign, 17.3% atypical/suspicious and 74.7% malignant with satisfactory agreement with the FCM evaluation (kappa Cohen’s coefficient 0.95; 95% CI, 0.89-1.01; p = 0.001). The sensitivity of the FCM evaluation was 100 % (95% CI, 95%-100%), specificity 66.7% (95% CI, 22.3%-95.7%) accuracy 97% (95% CI, 90.7%-99.7%), PPV 97 (95% CI, 91.8%-99%) and NPV 100%. Conclusion FCM represents a new technique successfully applicable to microhistological specimens. It provides fast information about sample adequacy in small specimens with good agreement with the final histology.