Side-by-side comparison of next-generation sequencing, cytology, and histology in diagnosing locally advanced pancreatic adenocarcinoma.

2020 
Abstract Background and Aims EUS-guided biopsy is the method of choice for obtaining pancreatic tissue. Next-generation sequencing (NGS) has been applied to EUS-guided biopsies and may classify patients based on specific molecular profiles. Our study aimed to compare side-by-side the diagnostic yield achievable by genetic identification of somatic mutations detected with NGS versus histological and cytological typing in locally advanced pancreatic carcinoma (LAPC) in samples acquired under EUS guidance. Methods We conducted a prospective comparative pilot study at Humanitas Research Hospital, registered on ClinicalTrials.gov (NCT03578939). The study included 33 patients referred for LAPC, who underwent EUS-guided tissue acquisition using a 22-gauge Franseen needle. Material was obtained for both pathological diagnosis and DNA extraction and targeted NGS analysis with the AmpliSeq Comprehensive Panel v3. Twenty-one genes were prioritized for somatic mutation detection. Results The final diagnosis was pancreatic ductal adenocarcinoma (PDAC) in all patients (100%). A macroscopic core was obtained in 30 patients (91%). In 3 lesions no cores adequate for histological analysis were obtained, but cytological analysis revealed tumoral cells from PDAC. DNA was extracted from 32 out of 33 samples (97%), most of which (27/32) carried at least 2 clearly pathogenic mutations in different genes. Detection of K-ras mutation allowed for molecular diagnosis of PDAC in most of the patients (30/32). Conclusions In our study we demonstrated that proper tissue specimens obtained under EUS guidance allowed DNA sample extraction and subsequent NGS analysis in 97% of cases. These results support the potential role of NGS as a complementary diagnostic test to be implemented in association with standard diagnostic modalities.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    20
    References
    4
    Citations
    NaN
    KQI
    []