BACKGROUND AND AIMS

FNA is the primary method of EUS tissue acquisition. In an attempt to improve our yield of EUS-guided tissue acquisition, we compared fine-needle biopsy (FNB) sampling without rapid onsite evaluation (ROSE) with FNA with ROSE and assessed the concordance of FNA and FNB sampling.

METHODS

This was a retrospective review of prospectively collected data from consecutive patients. Patients underwent FNB sampling and FNA of the same single lesion, with the same needle gauge and number of passes. FNA with ROSE was performed with a standard FNA needle. FNB sampling was performed with a new dedicated core needle. FNA samples were assessed with ROSE, and a final interpretation was provided by cytopathology staff; FNB samples were analyzed by surgical pathologists, each not made aware of the other's opinion.

RESULTS

Thirty-three patients underwent 312 passes in 42 different lesions. A diagnosis of malignancy was more likely with FNB sampling than with FNA (72.7% vs 66.7%, P = .727), although statistical significance was not reached. FNA and FNB sampling had similar sensitivities, specificities, and accuracies for cancer (81.5% vs 88.9%, 100% vs 100%, and 84.8% vs 90.9%, respectively). FNB sampling provided qualitative information not reported on FNA, such as degree of differentiation in malignancy, metastatic origin, and rate of proliferation in neuroendocrine tumors.

CONCLUSIONS

FNB sampling without ROSE using a dedicated core needle performed as well as FNA with ROSE in this small cohort, suggesting that FNB sampling with this new core needle may eliminate the need for an onsite cytopathologic assessment, without loss of diagnostic accuracy.