Molecular Profiling in Daily Clinical Practice: Practicalities in Advanced Cholangiocarcinoma and Other Biliary Tract Cancers

Abstract

Background: Molecular profiling is becoming increasingly relevant in the management of patients with advanced cancer; to identify targetable aberrations and prognostic markers to enable a precision medicine strategy. Methods: Eligible patients were those diagnosed with advanced biliary tract cancer (BTC) including intrahepatic (iCCA) and extrahepatic cholangiocarcinoma (eCCA), gallbladder cancer (GBC), and ampullary carcinoma (Amp) who underwent molecular profiling between April 2017 and June 2020 based on analysis of either tumour samples (FoundationOne CDx®/Oncomine® platforms) or ctDNA (FoundationOne Liquid® platform (Foundation Medicine, Cambridge, MA, USA)). Baseline patient characteristics and molecular profiling outcomes were extracted. The primary aim was to describe sample failure rate. Secondary aims included description of reason for sample failure, summary of findings derived from molecular profiling, and assessment of concordance between paired tissue and ctDNA samples. Results: A total of 149 samples from 104 individual patients diagnosed with advanced BTC were identified and eligible for this analysis: 68.2% iCCA, 100% advanced stage; 94.2% received palliative therapy. The rate of sample failure was 26.8% for tissue and 15.4% for ctDNA; p-value 0.220, predominantly due to insufficient (defined as <20%) tumour content in the sample (the reason for 91.2% of tissue sample failure). Of the 112 samples successfully analysed, pathological molecular findings were identified in the majority of samples (88.4%) and identification of pathological findings using ctDNA, was possible regardless of whether the patient was on active treatment at time of blood acquisition or not (p-value 1.0). The rate of targetable alterations identified was 40.2% across all successfully-analysed samples (39 iCCA; 6 non-iCCA): IDH1 mutations (19.1% of individual patients), FGFR2 alterations (10.1% and 5.6% of individual patients had FGFR2 fusions and mutations, respectively); 10.6% of all patients (12.4% of patients with successfully analysed samples) entered trials with matched targeted therapies as a consequence. Concordance of findings for paired tissue and paired tissue-ctDNA was high (3/3; 100% and 6/6; 100%, respectively). Twelve ctDNA samples were taken prior to palliative treatment initiation, median maximum mutant allele frequency (MAF) was 0.47 (range 0.21–19.8); no significant association between reported maximum MAF and progression-free survival (PFS) or overall survival (OS) (all Cox regression p-values > 0.273). A total of 15 patients (16.6%) harboured alterations in DNA damage repair (DDR) genes; when treated with platinum-based chemotherapy, there was a trend towards increased partial response rate (21.4% vs. 15.9%; p-value 0.653), radiological benefit rate (64.3% vs. 36.2%; p-value 0.071), and longer OS (median OS 20.4 months (95% CI 7.9–26.7) vs. 13.3 (95 CI 11.0–16.4); Cox Regression HR 0.79 (95% CI 0.39–1.61), p-value 0.527). Conclusions: Molecular profiling is of use for identification of novel therapeutic strategies for patients with advanced BTC (mainly iCCA). One in four archived tissue samples may have insufficient tumour content for molecular profiling; ctDNA-based approaches may overcome these obstacles.

A Pancancer Analysis of the Expression Landscape and Clinical Relevance of Fibroblast Growth Factor Receptor 2 in Human Cancers

“…A previous study reported that FGFR2 fusions are observed mostly in cholangiocarcinoma, occurring in 10-16% of patients (24,44,45). In our study, we found that cholangiocarcinoma had the highest frequency of FGFR2 alteration, and fusion accounted for the major proportion.…”

Practical considerations in screening for genetic alterations in cholangiocarcinoma

“…99 In this regard, a recent retrospective analysis of 149 tumor samples from 104 patients with advanced BTCs demonstrated a high sample failure rate of 26.8% in tissue biopsies, which was mainly due to inadequate tumor content in the sample (91.2%). 100 Although repeat biopsy might remedy sample failure due to inadequate tumor content, this is also accompanied by correspondingly increased morbidity risk. 101 In addition, FFPE processing and storage of tumor tissue for RNA-based assays may pose practical challenges due to the instability of RNA, which is easily degraded by omnipresent RNases.…”

“…Recent studies of NGSbased genomic profiling in patients with advanced tumors have shown variable concordance (<60% to 100%) in the detection of clinically relevant genetic alterations in plasmaderived ctDNA versus genomic DNA derived from tumor biopsies. 100,[104][105][106][107][108][109] Low concordance has been attributed, in part, to intratumoral genomic heterogeneity (e.g. between primary and metastatic sites).…”

“…Current research focused on the overall cost and cost-effectiveness of NGS-based molecular profiling has been reviewed extensively elsewhere. 99,100,[136][137][138][139][140][141][142] Overall, there is a paucity of real-world evidence for the cost-effectiveness of the use of NGS in routine clinical practice, upon which payers rely for making coverage decisions. 143,144 In keeping with this, a qualitative study published in 2015 which canvassed US payers for their opinions regarding challenges associated with coverage of NGS-based tests cited a lack of evidence-based support leading to a belief that NGS should be considered an investigational/experimental modality rather than a medical necessity.…”

Current and New Biomarkers for Early Detection, Prognostic Stratification, and Management of Gallbladder Cancer Patients

“…In terms of therapy, predictive and prognostic markers are urgently needed. In the setting of advanced disease, the access to good-quality tissue, sufficient for biomarker discovery and validation [14], together with limited translational research performed as part of prospective clinical trials, represent significant ongoing challenges [15]. In addition, the absence of actionable targets (excluding HER2) poses a major hindrance to the design of tailored pharmacological strategies beyond conventional therapy [16].…”