BTC is a well-targeted cancer type addressed by precision medicine. The European Society for Medical Oncology precision medicine working group guidelines recommended that CGP should be investigated using NGS in this cancer ^[6][22]. Genomic findings differ for the three BTC types, including intrahepatic cholangiocarcinoma (IHCCA), extrahepatic cholangiocarcinoma (EHCCA), and gallbladder cancer (GBCA). Gene mutations of FGFR1–3, TP53, IDH1/2, ARID1/2, CDKN2A/B, and KRAS were found in IHCCA ^[1][2][1,2]. Changes in KRAS, TP53, and SMAD4 were detected in EHCCA ^[1][2][1,2]. Important alterations in TP53, ERBB2/3, CDKN2A/B, ARID1A, and KRAS have also been reported in GBCA ^[1][2][1,2].

2.1. IDH1

IDH is an enzyme that regulates energy metabolism in the citric acid cycle. Wild-type IDH1 contributes to the reaction, producing α-ketoglutaric acid from isocitric acid. Mutant IDH1 produces D-2-hydroxyglutarate (D-2-HG), an oncometabolite. D-2-HG competitively inhibits IDH1 enzyme activity that depends on α-ketoglutaric acid, causing carcinogenesis ^[7][23]. About 16% of patients with IHCC have IDH1 mutations [1]. Ivosidenib is a novel small molecule inhibitor of mutant IDH1 protein. ClarIDHy is an international phase 3 study that proved the efficacy of ivosidenib against BTC with IDH1 mutation. PFS, which was the primary study endpoint, was 2.7 months (95% CI: 1.6–4.2 months) and 1.4 months (95% CI: 1.4–1.6 months) for ivosidenib and placebo groups, respectively [8].

2.2. Fibroblast Growth Factor Receptor (FGFR)2 Fusion

FGFs are growth factors comprising 23 family members, which can bind to FGFR1–4 as ligands. FGFR2 fusion genes are known as driver oncogenes ^[9][10][24,25]. The prevalence of FGFR2 mutations in IHCCA was reported as 5–15%. At least 19 fusion partners were identified and contained functional domains related to the formation of a dimer or multimer. Generated from the fusion gene, FGFR2 forms a dimer which is activated without ligand stimulation, causing oncogenesis ^[9][10][24,25].

2.3. EGFR (HER)2 Amplification

HER2 (encoded by ERBB2 gene) is a membrane-penetrating receptor tyrosine-protein kinase. After the attachment of a specific ligand, the receptor forms a heterodimer and triggers growth-stimulating intracellular signaling. Overexpression or mutation of HER2 continuously activates the mitogen-activated protein kinase (MAPK) and PI3K/AKT pathways independently without a ligand. Therefore, HER2 overexpression/mutations are often associated with tumor proliferation or inhibition of apoptosis ^{[11][12][13][14][15][16]}[26,27,28,29,30,31]. ERBB2 amplification was rarely found in IHCCA(3%) but was more common in EHCCA (11%) and GBCA (16%) [1].

2.4. BRAF V600E

BRAF (A, B, and C subtypes) protein is a serine/threonine kinase associated with the MAPK pathway ^[17][18][19][32,33,34]. The binding of BRAF ligands forms a dimer, which transmits signals downstream of the MAPK network. Mutated BRAF could play a role as oncogenes. BRAF mutants activate downstream growth-promoting signals continuously, causing oncogenic transformation. The most common oncogenic BRAF mutation is BRAF V600E ^[20][35]. BRAF mutation frequency was reported at 5% in BTC [1]. BRAF V600E is one of the tumor-agnostic targets of precision medicine. The ROAR basket trial indicated that the combination therapy with dabrafenib (BRAF inhibitor) plus trametinib (MEK inhibitor) was effective against the BRAF V600E-mutated BTC ^[21][14].

3. Precision Medicine Targets in Pancreatic Ductal Adenocarcinoma (PDAC)

PDAC has a comparatively uniform genetic background comprising four mutated genes: KRAS, CDKN2A [p16/p14], TP53, and SMAD4 [3]. Although genetic alterations were frequently observed in ARID1A, KMT2C, RNF43, FAT3, and KMT2D [3], KRAS, BRCA2, and NRG1 are considered more valuable in cancer diagnostic using precision medicine for PDAC [3].

3.1. BRCA1/2

PDAC represents a phenotype of hereditary breast and/or ovarian cancer syndrome, the well-known hereditary tumor syndrome caused by the BRCA1/2 mutation. The main phenotypes are breast and ovarian cancers, as its name suggests. Instances of prostate and other cancers have also been reported in carriers of the mutation ^[22][23][24][36,37,38]. BRCA mutation causes decompensation of double-strand DNA break repair—homologous recombination repair deficiency. Poly (ADP-ribose) polymerase (PARP) inhibitor, which inhibits single-strand DNA break compensation, can activate apoptosis in cancer cells with BRCA1/2 mutation and increase anti-cancer platinum sensitivity. However, only 6% of metastatic PDACs had BRCA1/2 mutations ^[25][15].

3.2. KRAS G12C

KRAS mutation is a significant oncogenic mutation and is found in 95% of PDACs. RAS is activated by connecting with GTP, receiving the signal from receptor tyrosine kinase. Activated RAS transmits the proliferation-promoting signal to the downstream MAPK pathway effectors. RAS, which has GTPase function, converts GTP into GDP via hydrolysis and proceeds to its inactivated state. However, mutated RAS has weak GTPase activity and induces a prolongation of activated signaling downstream, contributing to a malignant trait. Several efforts have been made to develop drugs that target mutated RAS. The effectiveness and safety of sotorasib, which affect mutated KRAS G12C, have been assessed in a phase 1/2 study named CodeBreaK 100 trial ^[26][18].

3.3. Neuregulin (NRG)1

NRG1 is now gaining attention as a new target for precision medicine among KRAS-negative PDACs ^[27][28][44,45]. NRG1 protein is a cell adhesion molecule that belongs to the NRG family and affects the epidermal growth factor receptor (EGFR). The epidermal growth factor (EGF)-like domain of NRG1 regulates itself or nearby cells using HER2s or HER3/HER4 dimers. About 0.5% of PDAC have NRG fusion, and the specific fusion partners in PDAC are CDH1, ATP181, and VTCN1 ^[29][19]. The efficacy of zenocutuzumab, an NGR inhibitor, is proved in the phase 1/2 global open-label clinical trial (eNRGy) and the early access program against NSCLC and PDAC showing NRG1 fusion.

4. Possible Targets for Precision Medicine of Hepatocellular Carcinoma (HCC)

The accumulation of gene aberrations induced by several risk factors, such as alcohol, hepatitis virus infection, or steatohepatitis, causes hepatocarcinogenesis. Genetic alterations in TERT, TP53, or Wnt/βcatenin signaling pathway could play a role as a cancer driver for HCC. Genetic analysis of 755 HCC patients using NGS revealed genetic aberrations in 44% (TP53), 35% (CTNNB1), 31% (ARID1A), and 12% (MYC) of the assessed genes. However, these alterations have not been targeted by therapies and precision medicine has not been well-established in patients with HCC. Although some tumor-agnostic gene alterations, including changes in NTRK and MET, lead to precision medicine, the incidences of these genes are low (2–4% for NTRK and 2% for MET) ^[4][5][4,5].

4.1. Telomerase Reverse Transcriptase: TERT

TERT is one of the most common driver oncogenes of HCC with respect to cancer cell immortality. Every time cells divide, the recurrent arrangements at the edge of the chromosomes become shorter, and when the telomere length is shortened to a decided extent, the cells die. Cancer cells can be immortalized because of the high telomerase activity, which maintains the telomere length. TERT codes a telomerase subunit and affects the regulation of telomerase activity, causing the oncogenic progression of various cancers. HCCs have a 50% point mutation in the promotor area that regulates the TERT gene expression. Although drugs targeted for TERT treatment are being developed, their efficacy has not yet been established ^[4][5][4,5].

4.2. TP53

TP53 is one of the most effective tumor suppressing genes ^[30][46]. TP53 mutations are found in various human cancers ^[31][32][47,48]. Point mutations of the gene were found in ~40% of HCC. The protein product of TP53, p53, functions as a transcription factor, suppresses transcription genes concerned with cell proliferation, and is called the “Guardian Deity of Genome.” The inactivation of p53 leads to genome instability and accumulates genome abnormalities that lead to oncogenesis. However, precision medicine based on the TP53 mutation has not yet been well established ^[4][5][4,5].

4.3. Wnt/βcatenin Signaling Pathway

The Wnt/βcatenin signaling pathway is important for organogenesis, maintenance of stem cells, adjustment of cell proliferation, movement, and polarity ^[33][49]. Alterations of this pathway were found in 40–50% of HCCs ^[34][21]. Most of these changes are point mutations of CTNNB1 exon3 that are responsible for CTNNB1 transfer into the nucleus. The mutations can activate MYC and CTNNDI, causing the proliferation of HCC cells ^[34][21]. This pathway alteration is also strongly related to the immunological properties of the tumor and might lead to a response to immunological therapy ^[35][50].

5. Tumor-Agnostic Precision Medicine

Recently, the efficacy of the same drug against common genetic alterations in all types of cancer has been established. Here, we discuss microsatellite instability-high (MSI-H), tumor mutation burden-high (TMB-H), and NTRK fusions. Although there are several reports showing the possible tumor-agnostic precision medicines for hepatobiliary and pancreatic cancers, the evidence is not enough because of an insufficiency in the number of examined cases, and thus further examinations are awaited. The tumor-agnostic genetic alterations and treatment results for hepatobiliary and pancreatic cancers are summarized in Table 12. In addition to these treatments, it is noteworthy that the United States Food and Drug Administration (FDA) recently accelerated approval of tumor-agonistic therapy for solid tumors with BRAF mutation ^[36][51].

	MSI-H	TMB-H	NTRK
Alterations	Mutations of MLH1, MSH2, MSH6, PMS2	-	Fusions
Prevalence Whole cancers Hepatobiliary pancreatic	[52] 5% Around 2%	[53] 13% 4%	[54] 0.26% (87/33997) BTC 0.34% (2/787), PDAC 0.25% (5/1492)
Drugs	Pembrolizumab	Pembrolizumab	Larotrectinib	Entrectinib
Study	Phase 2 KEYNOTE-158	Phase 2 KEYNOTE-158	Integrated analysis of three phase 1/2	Integrated analysis of two phase 2 and one phase 1
Efficacy Whole cancers Hepatobiliary	[55] ☆ ORR 34.3% (80/233) BTC 41% (9/22), PDAC 18% (4/22)	[56] ☆ ORR 29.4% (30/102) no data	[57] ☆ ORR 79% (121/153) ORR of BTC 50% (1/2); PDAC 50% (1/2)	[58] ☆ ORR 57 % (31/54) ORR of BTC 100% (1/1); PDAC 67% (2/3)