Session VII: Biomarkers

Moderator Rachna Shroff, MD, MS, FASCO, with Mark Yarchoan, MD, Bruce Lin, MD, and Kristen Spencer, DO, MPH (not pictured, Pashtoon Kasi, MD, MS)

During Session VII, panelists discussed biomarkers in biliary tract cancer (BTC). Mark Yarchoan, MD, Johns Hopkins University School of Medicine, Baltimore, MD, discussed the immune microenvironment in cholangiocarcinoma (CCA). CCA is broadly characterized by a low tumor mutational burden (TMB) and an immune-resistant tumor microenvironment.¹ Research has suggested that the genomic landscape of CCA impacts the tumor immune microenvironment and survival outcomes when treated with immunotherapy.² An open-label, multicohort study of pembrolizumab (KEYNOTE-158) in multiple histologies showed an objective response rate of 29% in tissue TMB-high patients and 6% in non–tissue TMB-high patients.³ In addition, genomic and transcriptomic profiling of approximately 400 patients with BTC revealed 4 different clusters with distinct immune microenvironments in CCA.⁴ It was found that cluster 1, predominantly associated with TP53, KRAS, and ATM alterations, had the highest degree of inflammation among the tumor microenvironments, with increased B cells and plasma cells; cluster 2, associated with CDKN2A/B alterations, had decreased memory resting CD4 cells; cluster 3, associated with IDH1 alterations, had increased activated natural killer cells and M2 macrophages; and cluster 4, associated with FGFR2 rearrangements and BAP1 mutations, had low immune infiltration and decreased activated dendritic cells.⁴ Dr Yarchoan concluded that there is a rationale for exploring targeted inhibition of specific drivers to reprogram the tumor-immune microenvironment in combination with systemic immunotherapy.

Bruce Lin, MD, Virginia Mason Medical Center, Seattle, WA, presented on new developments in biomarkers. Many biomarkers have been investigated in BTC for a variety of clinical implications, some of which include cell-free DNA (cfDNA)/circulating tumor DNA (ctDNA), microRNA (miRNA), proteins, metabolites, extracellular vesicles, and circulating tumor cells.⁵ cfDNA/ctDNA may allow for early detection of disease and aid in the monitoring of patients at risk for cancer development and the evaluation of treatment response.⁶ miRNA is a useful biomarker for disease detection because it is highly abundant, stable in biofluids, more resistant to degradation, and easily detected and amplified.⁵ Cancer antigen 19-9 and carcinoembryonic antigen are 2 proteins that can be used as biomarkers for the detection of CCA; however, they are associated with low sensitivity and specificity⁵; additional novel diagnostic and prognostic protein biomarkers are currently being investigated. Metabolic profiling is a promising avenue in the identification of biomarkers.⁵ Algorithms combining multiple metabolites may help discriminate between hepatocellular carcinoma and intrahepatic CCA or primary sclerosing cholangitis and intrahepatic CCA.⁵ Extracellular vesicles are promising biomarkers because they can be detected in blood, saliva, urine, and bile, and they carry proteins, nucleic acids, lipids, and other metabolites.⁵ Detection of these extracellular vesicles is important because they mediate cell-to-cell communication and contribute to the growth of CCA.⁵ Circulating tumor cells can be used for in vitro drug screenings or in vivo treatment sensitivity testing.⁵ It has been shown that higher circulating tumor–cell counts are correlated with decreased overall survival.⁵ In general, the majority of biomarkers have not been validated and few are ready to be translated into clinical practice. Artificial intelligence may become an important tool to help analyze the vast array of biomarkers.⁵

Kristen Spencer, DO, MPH, NYU Perlmutter Cancer Center, New York, NY, presented on the use of biomarkers to predict treatment response and modify therapy. Dr Spencer first described the characteristics of a good biomarker—can be observed from outside the patient, can be measured accurately and reproducibly, can detect changes in disease or toxicity before other standard clinical indicators, has clinical actions available based on the results, and can be measured less invasively and less expensively compared with currently available indicators.⁷ Biomarkers can be used to predict treatment response, escalate therapy, de-escalate therapy, or change therapy.⁷ Dr Spencer presented a clinical study example for each of these biomarker uses. For predicting treatment response, 1 study found that mutated KRAS and chromosome-instability tumors have fewer tumor-infiltrating lymphocytes and are, therefore, resistant to PD-1/L1 blockade.⁷ For escalating therapy, one study found that mutations in IDH can predict response to gemcitabine-based treatment combinations.⁸ The study found that patients with mutated IDH BTC who received gemcitabine-based treatment as first-line chemotherapy had an overall response rate of 39% versus 0% when treated with non–gemcitabine-based first-line chemotherapy.⁸ For de-escalating therapy, CA 19-9 can be used to predict outcomes in patients with intrahepatic CCA prior to hepatectomy.⁹ Lymph node positivity is an important prognostic factor for overall survival, but it is difficult to detect preoperatively. CA 19-9 was used preoperatively to detect lymph node positivity, and results showed that CA 19-9 was significantly higher in the N1 group versus the N0 group (negative for cancer in lymph nodes).⁹ CA 19-9 was also predictive of N1 disease, with a sensitivity of 72% and specificity of 81%.⁹ Therefore, CA 19-9 may play a role in predicting who would be appropriate candidates for hepatectomy.⁷ Finally, regarding changing therapy, researchers found that polyclonal secondary FGFR2 mutations confer resistance to FGFR2 inhibition in patients with fusion-positive CCA.¹⁰ This study examined 3 patients, all of whom developed secondary FGFR2 kinase domain mutations in response to FGFR2 inhibition therapy and would, therefore, require a change in therapy.¹⁰ Overall, there are various ways in which biomarkers can be used to predict treatment response and modify therapy in patients with BTC, as demonstrated by these studies.

References

1. Yarchoan M, Albacker LA, Hopkins AC, et al. PD-L1 expression and tumor mutational burden are independent biomarkers in most cancers. JCI Insight. 2019;4(6):e126908.
2. Yarchoan M. Immune microenvironment in cholangiocarcinoma. Presented at: 5th Anniversary Cholangiocarcinoma Summit, October 19-21, 2023; Scottsdale, Arizona.
3. Marabelle A, Fakih M, Lopez J, et al. Association of tumour mutational burden with outcomes in patients with advanced solid tumours treated with pembrolizumab: prospective biomarker analysis of the multicohort, open-label, phase 2 KEYNOTE-158 study. Lancet Oncol. 2020;21(10):1353-1365.
4. Mody K, Jain P, El-Rafai SM, et al. Clinical, genomic, and transcriptomic data profiling of biliary tract cancer reveals subtype-specific immune signatures. JCO Precis Oncol. 2022;6:e2100510.
5. Lin B. New developments in biomarkers. Presented at: 5th Anniversary Cholangiocarcinoma Summit, October 19-21, 2023; Scottsdale, Arizona.
6. Rodrigues PM, Olaizola P, Paiva NA, et al. Pathogenesis of cholangiocarcinoma. Ann Rev Pathol. 2021;16:1,433-463.
7. Spencer K. Use of biomarkers to predict treatment response and modify therapy. Presented at: 5th Anniversary Cholangiocarcinoma Summit, October 19-21, 2023; Scottsdale, Arizona.
8. Wintheiser G, Zemla T, Shi Q, et al. Isocitrate dehydrogenase-mutated cholangiocarcinoma: natural history and clinical outcomes. JCO Precis Oncol. 2022;6:e2100156.
9. Yamada T, Nakanishi Y, Okamura K, et al. Impact of serum carbohydrate antigen 19-9 level on prognosis and prediction of lymph node metastasis in patients with intrahepatic cholangiocarcinoma. J Gastroenterol Hepatol. 2018. Epub ahead of print.
10. Goyal L, Saha SK, Liu LY, et al. Polyclonal secondary FGFR2 mutations drive acquired resistance to FGFR inhibition in patients with FGFR2 fusion-positive cholangiocarcinoma. Cancer Discov. 2017;7(3):252-263.