8/
To establish how well 🩸🧬 cfDNA recapitulates 💉🧬 tumor tissue, we compared the prevalence of mutations between the BTC cfDNA samples (N=1,671) with metastatic BTC tumor NGS data from AACR Project GENIE (N=349).

Correlation was impressively high ⬆️ (R2=0.96)‼️

9/
🤝We next sought to characterize intra-patient cfDNA-tissue concordance.

💯All BRAF V600E mutations present in tumor were detected in cfDNA.

🎯87% of IDH1 mutations present in tumor were detected in cfDNA.

📢Timing of cfDNA collection is critical to optimizing detection!

10/
😕Unfortunately, cfDNA detected only 18% of FGFR2 fusions present in tissue.

🧐Detection rates strongly correlated with the identity of the FGFR2 fusion partner.

✅58% detection rate for BICC1 fusions (most common partner).

❌2% detection rate for non-BICC1 fusions.

11/
💪Leveraging the cohort size, we characterized the spectrum of FGFR2 muts in cfDNA – a key resistance mechanism to FGFR inhibitors in patients with FGFR2 fusion-positive BTC.

👋We identified 70 unique FGFR2 muts, including 13 previously unreported FGFR2 kinase domain muts.

12/
💎Clonality is important in cfDNA and provides insights into tumor biology…

▶️Clinically actionable alts in BTC – BRAF V600E, IDH1 muts, and FGFR2 fusions – were the most likely alts to be clonal in cfDNA, affirming that these alterations are early driver events in BTC.

13/
📄We collected clinical data to correlate outcomes with cfDNA VAF, a known prognostic factor.

🔢We divided patients into quartiles by the highest VAF for any alteration in a given sample (aka MaxVAF).

📉The highest quartile of MaxVAF (VAF >9%) had the worst OS.

14/
👉High pre-treatment cfDNA MaxVAF (VAF >9%) associated with shorter PFS to first-line chemotherapy in pts with advanced BTC.

📉Median PFS = 2.6 vs 7.6 months ▶️ HR: 1.9 (95%CI 1.1-3.2; P=0.035)

📉Median OS = 8.6 vs 18.3 months ▶️ HR: 1.5 (95%CI 0.80-2.5; P=0.21)

15/
👉High pre-treatment cfDNA MaxVAF (VAF >9%) associated with shorter PFS to FGFRi & worse OS in pts with FGFR2 fusion-positive CCA.

📉Median PFS = 3.2 vs 7.9 months ▶️ HR: 6.1 (95%CI 2.1-19; P=0.001)

📉Median OS = 19.8 vs 31.6 months ▶️ HR: 3.0 (95%CI 1.1-8.3; P=0.039)

16/
⭐️Conclusions⭐️

📢These findings strongly support the utility of cfDNA genomic profiling to inform clinical care for patients with BTC.

📢It is important that clinical trials incorporate cfDNA to further define its utility as a clinical biomarker for patients with BTC.

17/
👉A special thank you to Lipika Goyal who personifies excellence in mentorship. Her drive to improve outcomes for people diagnosed with BTC is relentless and infectious. Be on the lookout for more to come from this talented clinician-scientist & amazing person.

18/
👩‍⚕️👨‍⚕️This work would not have been possible without these wonderful collaborators at Mass General Cancer Center Mayo Clinic Siteman Cancer Center UCSF Helen Diller Family Comprehensive Cancer Ctr Gustave Roussy ➡️ @gioncdoc Rachna Shroff, MD, FASCO Daniel Catenacci Milind Javle Friboulet Luc Antoine Hollebecque Mitesh Borad, M.D. Aparna Raj Parikh Kabir Mody