Next generation sequencing for biliary tract cancers

INTRODUCTION

Biliary tract cancers (BTC), consisting of cholangiocarcinoma (CCA) and gallbladder cancer (GBC), are rare and aggressive malignancies that carry a poor prognosis [1]. Gemcitabine and cisplatin-based regimens remain the standard of care in patients with advanced disease, and few treatment options exist upon progression. Although second-line oxaliplatin/5-FU was shown to modestly increase overall survival rates in patients who progressed on cisplatin and gemcitabine, chemotherapy is associated with low response rates and significant toxicity [2]. Increased utilization of next-generation sequencing (NGS) in clinical care has allowed for a greater understanding of the pathogenesis of BTC. NGS has enabled the identification of recurrent genomic alterations in BTC, including alterations in TP53, KRAS, IDH1, ERBB2, FGFR2, BRAF, and ARID1A, which has facilitated enrollment of patients in biomarker-selected clinical trials [1]. In this editorial, we discuss the mutational landscape of BTC, the efficacy signal of genomically-matched therapies, circulating tumor DNA (ctDNA) targeted genotyping, and the role of NGS in clinical decision making.

BODY

1. Mutational landscape of biliary tract cancers

Historically, CCA has been classified by anatomic location, with intrahepatic cholangiocarcinoma (IHCCA) affecting the bile ducts within the liver and extrahepatic cholangiocarcinoma (EHCCA) affecting the bile ducts outside of the liver. Furthermore, subclassification of EHCCA includes hilar (Klatskin) and distal common bile duct tumors. While these classifications can help guide clinical management, they have limited prognostic value and there is a pressing need for improved classification of these tumors. The affordability of NGS has enabled rapid profiling of multiple genes on one panel, which has helped identify molecular subsets of BTC and elucidate differences in genomic drivers between IHCCA, EHCCA, and GBC (Table 1) [1,3,4].

Table 1:

Frequency of selected genomic alterations among biliary tract cancer subtypes

Gene	IHCCA	EHCCA	GBC
ARID1A	18–23%	12%	13%
BAP1	8–19%	0–12%	0–2%
BRAF	5%	3%	1%
CDKN2A/B	27%	17%	19%
EGFR	1–4%	0–1%	4–18%
ERBB2	3%	11%	16%
FGFR1–3	4–13%	0%	3%
IDH1/2	20–29%	0–5%	0%
KRAS	7–25%	12–42%	0–11%
MET	2%	0%	1%
PI3KCA	5%	7%	14%
SMAD4	5%	30%	N/A
STK11	1%	11%	N/A
TP53	18–27%	40–49%	59%

Table 1: This table depicts selected genomic alterations across biliary tract cancer subtypes when stratified by anatomic location. These data were extracted from Javle et al. Cancer 2016, Lowery et al. Clinical Cancer Research 2018, and Nakamura et al. Nature Genetics 2015 [1,3,4].

IHCCA, intrahepatic cholangiocarcinoma; EHCCA, extrahepatic cholangiocarcinoma; GBC, gallbladder cancer.

Several studies over the past decade have investigated the molecular profiles of BTC. Nakamura et al. found that 40% of BTC harbored mutations in actionable genes, and FGFR2 fusions were more commonly seen in IHCCA in comparison to EHCCA, where mutations in PRKACA or PRKACB were identified [4]. Javle and colleagues utilized hybrid capture-based comprehensive genomic profiling to assess the mutational profile of 412 patients with IHCCA, 57 patients with EHCCA, and 85 patients with GBC [1]. Frequent genetic aberrations identified in IHCCA were TP53, CDKN2A/B, KRAS, and IDH1, while KRAS, TP53, CDKN2A/B, and SMAD4 were most prevalent in EHCCA [1]. Mutations in TP53, CDKN2A/B, ARID1A, and ERBB2 were frequently identified in GBC, and ERBB2 alterations were the most common in GBC in comparison to other BTC [1]. Pathogenic alterations in TP53 and KRAS were associated with poorer survival and FGFR2 mutations were associated with improved survival, which highlight the prognostic relevance of select mutations in BTC [1]. A study from Memorial Sloan Kettering Cancer Center identified recurrent mutations in IDH1, ARID1A, BAP1, TP53, and FGFR2 gene fusions in IHCCA tumors [3]. Of the 195 cases assessed, 47% of patients had potentially actionable genetic alterations, which led to enrollment in clinical trials or biomarker-directed therapy in 16% of cases [3]. In a recent study of GBC by Abdel-Wahab et al., at least 1 potentially actionable aberration was identified in 658 (87%) of 760 patients [5]. The most frequent actionable/potentially actionable alterations were CDKN2A, followed by ERBB2, PIK3CA, MDM2, CCNE1, STK11, ERBB3, ATM, and PTEN [5].

While CCA is a rare malignancy in the Western world, it is more prevalent in southeast Asia where Opisthorchis viverrini is endemic, and IHCCA accounts for nearly 70% of primary liver cancers in Thailand [6]. A study by Ong and colleagues sought to evaluate genomic profiles of liver fluke-associated CCA, and identified prevalent mutations in TP53, KRAS, and SMAD4 [6]. Furthermore, they identified mutations not previously implicated in CCA, which included genes involved in histone modification, G protein signaling, and genome stability [6]. The Thailand Initiative in Genomics and Expression Research for Liver Cancer (TIGER-LC) group performed targeted-exome sequencing on 197 patients from Thailand, identifying common mutations in TP53, ARID1A, ARID2, CSMD3, RYR2, NF1, PRKDC, PSIP1 in both IHCCA and HCC tumors [7]. These studies highlight the importance of both environmental exposures and ethnicity on the landscape of genomic alterations in BTC.

Whole-exome sequencing has also revealed genomic similarities within hepatocellular-cholangiocarcinoma (H-ChC), which shares pathologic features of both hepatocellular carcinoma (HCC) and IHCCA [8]. In a recent study from Wang and colleagues, nearly 80% of H-ChC were associated with EpCAM expression, suggesting a monoclonal origin of these tumors [8]. Additionally, these were characterized by substantial intratumoral heterogeneity, and mutations in VCAN, ACVR2A, and FCGBP were associated with differentiation of IHCCA and HCC within H-ChC [8]. Understanding the common genomic alterations within H-ChC may allow for better drug target selection in these tumors which are characterized by diverse molecular drivers.

2. Efficacy signal of genomically-matched therapies

Given the limited treatment options for patients with advanced BTC, there has been significant effort in the development of novel, molecularly-tailored therapeutic strategies (Table 2) [9]. IDH1/2 mutations are present in 10–23% of CCA and are more prevalent in IHCCA than EHCCA [1]. In the phase III ClarIDHy study (NCT02989857), Abou-Alfa and colleagues found 1.3 months of progression-free survival (PFS) benefit with ivosidenib in patients with IDH1-mutant CCA. While the improvement in median PFS appears modest, there were a significantly reduced risk of progression at 6 and 12 months, highlighting the potential benefit of targeting IDH1 in IDH1-mutant CCA [10]. Alterations in FGFR signaling are also prevalent in IHCCA, as 10–16% of tumors have FGFR2 fusions or rearrangements [11]. The phase II trial (NCT02924376) evaluating pemigatinib, a potent FGFR 1–3 inhibitor, demonstrated that 36% of patients with locally advanced or metastatic CCA and FGFR2 fusions or rearrangements had an objective response following treatment [11]. Another FGFR inhibitor in development for BTC patients harboring FGFR aberrations is futibatinib, an irreversible FGFR1–4 inhibitor, which is currently being evaluated in phase II/III trials across multiple histologies [12]. Preliminary analysis of FOENIX-CCA2 (NCT02052778), a single-arm multicenter phase II study evaluating futibatinib in patients with IHCCA and FGFR2 fusions or rearrangements, demonstrated a 37.3% objective response rate (ORR) and 82% disease control rate (DCR) [13].

Table 2:

Active genomically-matched trials in biliary tract cancers

Target	Therapy	Dedicated Trial or Basket	Biomarker Selection	Phase	Clinical Trial Registry #
ERK1/2	Ulixertinib	Basket	KRAS, NRAS, HRAS, BRAF (non-V600), MEK, and ERK Mutations	I	NCT04145297
FGFR	Infigratinib	CCA	FGFR2 Fusions/Translocations	III	NCT03773302
FGFR	Infigratinib	CCA	FGFR2 Mutation	II	NCT02150967
FGFR	Infigratinib	Basket	FGFR1–3 Fusions or Amplifications	II	NCT04233567
FGFR	Futibatinib	CCA	FGFR2 Rearrangement	III	NCT04093362
FGFR	Pemigatinib	CCA	FGFR2 Rearrangement	III	NCT03656536
FGFR	Derazantinib	CCA	FGFR2 Fusion, Mutation, or Amplification	II	NCT03230318
FGFR	RLY-4008	Basket	FGFR2 Fusion, Mutation, or Amplification	I	NCT04526106
FGFR	Pemigatinib	Basket	FGF/FGFR Alterations	I/II	NCT02393248
HER2	Zanidatamab	BTC	HER2 Amplification	II	NCT04466891
HER2	Tucatinib + Trastuzumab	Basket	HER2 Amplification or Overexpression	I/II	NCT04430738
HER2	Tucatinib + Trastuzumab	Basket	HER2 Mutation	II	NCT04579380
HER2	Trastuzumab deruxtecan	Basket	HER2 Expressing	II	NCT04482309
HER2	RC48-ADC	BTC	HER2 Overexpression	II	NCT04329429
IDH1	LY3410738	Basket	IDH1_R132 Mutation	I	NCT04521686
IDH1, FGFR	Ivosidenib, Pemigatinib	CCA	IDH1_R132 Mutation, FGFR Alterations	I	NCT04088188
MEK	Trametinib	BTC	KRAS Mutation	II	NCT04566133
NTRK	Entrectinib	Basket	NTRK 1/2/3 (Trk A/B/C), ROS1, or ALK Gene Rearrangements (Fusions)	II	NCT02568267
PARP	Olaparib	Basket	IDH1/2 Mutations	II	NCT03212274
PARP	Olaparib	BTC	DDR Pathway Mutations (Somatic or Germline)	II	NCT04042831
PARP	Niraparib	Basket	DDR Pathway Mutations	II	NCT03207347
PARP, ATR	Olaparib + Cerelasertib	Basket	IDH1/2 Mutations	II	NCT03878095

Table 2: This table highlights ongoing genomically-matched clinical trials in biliary tract cancers [9].

BTC, biliary tract cancers; CCA, cholangiocarcinoma.

HER2 pathway alterations are most commonly identified in EHCCA and GBC. A retrospective review of GBC patients at MD Anderson Cancer Center with HER2-amplification receiving HER2-targeted therapy found that 8 out of 9 patients had stable disease, partial response, or complete response, highlighting this promising treatment option [14]. The MyPathway trial (NCT02091141) showed promising activity in HER2-positive BTC following treatment trastuzumab and pertuzumab [15]. Zanidatamab, a HER2-bispecific antibody, was recently granted breakthrough therapy designation by the FDA for patients with HER2-amplified BTC based on the data from Phase I trial (NCT02892123) where zanidatamab treatment was well-tolerated and led to confirmed objective responses in 40% of patients [16]. A Phase II clinical trial in biliary cancers is ongoing (NCT04466891). HER2 is also being explored as a target in biliary tumors in ongoing antibody-drug conjugate (ADC) trials (NCT04482309, NCT04329429). BRAF is another promising therapeutic target in CCA, as 5–7% of BTC harbor BRAF mutations and BRAF V600E-mutant IHCCA is associated with worse survival [17]. The Rare Oncology Agnostic Research (ROAR) basket trial (NCT02034110) enrolled patients with BRAF V600E-mutant BTC for treatment with the combination of dabrafenib and trametinib and demonstrated a 47% ORR by independent review [17]. While present in <1% of BTC, NTRK fusions serve as another potential therapeutic target following the tissue agnostic approval of entrectinib and larotrectinib for solid tumors with NTRK fusions [18]. Immunotherapy is also being investigated in the treatment of BTC, including pembrolizumab as monotherapy or in combination with chemotherapy in MSI-high or MMR deficient CCA [15].

Several recent clinical trials, including MyPathway (NCT02091141), NCI-MATCH (NCT02465060), TAPUR (NCT02693535), and ROAR (NCT02034110), have successfully enrolled patients across a variety of indications based on genomic-alterations, including BTC. Considering the rarity of BTC and its diverse molecular drivers, signal-seeking basket trials will be crucial in evaluating responses to targeted agents for tumors with less prevalent mutations. A challenge in the treatment of patients with BTC is that many tumors have alterations that span multiple pathways, or mutations that are not considered targetable by currently approved agents. Rationale combinations that target multiple drivers may allow for more durable responses in these patients. Developing preclinical models that recapitulate human tumor biology, such as patient-derived xenograft models and organoids, may enable the testing of rational combination therapies.

3. Circulating tumor DNA targeted genotyping

The role of ctDNA in advanced malignancies is evolving and serves as a non-invasive tool to identify actionable mutations for therapy selection and monitor a tumor’s mutational profile throughout treatment. For patients with BTC, ctDNA has great potential, as tissue biopsies often do not provide adequate cytogenic material for molecular profiling [19]. A recent study from Ettrich et al. evaluated ctDNA from patients with CCA and found 74% concordance overall with tissue NGS [20]. Another study by Mody et al. assessed ctDNA profiles of 138 BTC patients and identified mutations with potential therapeutic implications in 76 (55%) patients, including BRAF mutations, ERBB2 amplifications, FGFR2 fusions, FGFR2 mutations, and IDH1 mutations [21]. A recent report from Yarlagadda et al. described a 71-year-old female with metastatic HER2-amplified CCA that was first identified by ctDNA, which then led to the initiation of dual anti-HER2 therapy and durable control of her disease for greater than 12 months [22]. There is also emerging data on genomic evolution in BTC with targeted therapy, such as development of polyclonal secondary FGFR2 mutations with FGFR inhibitors [23]. Further, there is evidence that FGFR inhibitors may have differences in their efficacy for different alterations, suggesting that therapies could be potentially tailored based on baseline and merging alterations [24]. Thus, it is likely that liquid biopsies may have clinical utility for the selection of targeted therapies, assessing mechanisms of acquired resistance, monitoring ctDNA dynamics in response to therapy, and assessing minimal residual tumor burden after surgery.

4. Role of next-generation sequencing in clinical decision making

Encouraging results from biomarker-selected clinical trials in BTC provides a strong rationale for a molecularly tailored approach to therapy for these patients. Recent FDA approval of pemigatinib for CCA with FGFR2 fusions/rearrangements highlight the importance of fusion testing in CCA for standard of care therapy. The 2017 National Cancer Care Network (NCCN) guidelines recommend MSI/MMR testing and consideration of additional molecular testing in patients with unresectable or metastatic IHCCA and EHCCA. More recently, the ESMO Scale for Clinical Actionability of molecular Targets (ESCAT) ranked genomic alterations in advanced CCA based on therapeutic implications, and level I alterations included IDH1 mutations, FGFR2 fusions, MSI-H, and NTRK fusions [25]. Level II and III alterations include BRAF V600E mutations, ERBB2 amplifications and mutations, PIK3CA hotspot mutations, BRCA 1/2 mutations, and MET amplifications [25]. Recent tumor agnostic approval of pembrolizumab for high tumor mutation burden (TMB) also supports the rational of genomic testing on large enough platform to report TMB.

Our strong recommendation is that all patients with unresectable and metastatic BTC undergo comprehensive genomic testing. The addition of routine genomic profiling to the management of BTC will allow clinicians and researchers to gain more insight into the biology of these rare tumors and potentially open the door for additional lines of personalized therapy. Furthermore, prospective collection of genomic data during routine care and clinical trials, as well as longitudinal collection of liquid biopsies, will allow for further understanding of the prognostic value of specific alterations and determinants of sensitivity/resistance to molecularly-targeted agents.