This article is mentioned in:

Introduction

Pancreatic ductal adenocarcinoma (PDAC) is one of the most aggressive epithelial tumors with a 5-year survival rate of <10% (1), which dismal prognosis is greatly related to a diagnosis at late stages and few effective treatment options. The prognosis of patients with PDAC has barely changed over the past two decades, as there are no reliable biomarkers for early detection (2). Although modest advances have been made in treatment options with combination therapies (3-5), recurrence rates remain high (~80%), with patients relapsing within 2 years (6). Hence, implementation of new diagnostic methods, such as liquid biopsy, may help enhance detection accuracy and monitoring tumor progression in real time.

PDAC occurs due to the accumulation of multiple genetic alterations, including activation of oncogenes or loss of tumor-suppressors, as well as aberrant function of signaling pathways (7). Acquisition of mutations in KRAS (KRAS proto-oncogene, GTPase) is regarded as a driver event in PDAC. However, several in vivo studies showed that mutated KRAS alone is insufficient to trigger metastatic transformation (8). Combination with other frequently found inactivating mutations in genes such as CDKN2A (cyclin-dependent kinase inhibitor 2A), SMAD4 (SMAD family member 4), and TP53 (tumor protein p53), or epigenetic changes in key genes, are recognized to further enhance tumorigenesis and metastasis. In fact, 70-90% of PDAC cases harbor co-occurring KRAS and TP53 mutations (9), constituting the most common genetic alterations in PDAC. Although overexpression of wild-type and mutated KRAS is well recognized in colorectal and non-small cell lung cancers (10,11), it remains poorly known in PDAC.

Here, we report the case of a patient with advanced PDAC with multiple liver metastases found to bear marked amplification of the oncogenic KRASG12D allele as detected by liquid biopsy. Mutant KRAS amplification may have important clinical implications, including increased risk for resistance to treatment.

Case report

A 52-years-old Japanese man with no relevant medical history visited our hospital in early July 2019 with chief complaints of persistent upper abdominal pain for 2 months. The patient had a 36-year history of smoking and daily alcohol consumption. No family history of cancer was reported. Physical examination showed high fever (>38.0˚C) and tenderness in the upper abdomen. Laboratory data revealed mild liver dysfunction and normal levels of carbohydrate antigen 19-9 (CA19-9; 31.9 U/ml), with duke pancreatic monoclonal antigen type 2 (DUPAN-2) >1,600 U/ml. Abdominal contrast-enhanced computed tomography (CT) revealed a hypovascular tumor mass of 25 mm in the head of the pancreas. Multiple liver metastases with different masses were detected and no metastases at other sites were evident on CT (Fig. 1A). The patient was immediately admitted, and endoscopic ultrasound-guided fine needle aspiration was performed on the primary tumor and metastases. Histological analysis confirmed that it was an adenocarcinoma, classified as cT3 cN0 cM1(Hep) and cStage IV according to the Union for International Cancer Control criteria and Tumor-Node-Metastasis classification (12). The patient started FOLFIRINOX (mFFX; folinic acid, fluorouracil, irinotecan, and oxaliplatin combo) therapy 2 weeks after the diagnosis in July 2019. On day 3 of treatment, the patient experienced liver dysfunction with increased levels of uric acid and creatinine. On day 6, a strong myelosuppressive effect [white blood cell (WBC): 600/µl, neutrophils: 256/µl, and platelets: 40,000/µl] was observed along with disseminated intravascular coagulation and acute renal failure. On day 8, the patient experienced encephalopathy and a marked increase in the levels of procalcitonin (PCT; 56.9 ng/ml) and C-reactive protein (CRP; 21.62 mg/dl) were observed, most likely as a result of tumor tissue damage. On day 19, a CT assessment revealed a reduction in the primary lesion; thus, the mFFX treatment was initially considered to be effective. However, soon after, a strong fever recurred with increased CRP levels. Therefore, mFFX re-administration was considered severely adverse and intolerable. With the patient informed consent, the regimen was changed and treatment continued with an intravenously administrated second-line therapy, named GnP, comprising gemcitabine (1,000 mg/m2) and nab-paclitaxel (125 mg/m2) (13). GnP did not induce adverse reactions as potent as mFFX, but still resulted in an inflammatory response and elevated levels of procalcitonin (Fig. 2). Despite a slight recovery after chemotherapy, the patient general condition continued to deteriorated and a myriad of new metastatic liver tumors emerged with uncontrollable growth patterns. The patient died of gastrointestinal bleeding associated with disseminated intravascular coagulation 70 days after treatment initiation.

Figure 1 Dynamics of ctDNA during the course of treatment. (A) CT scan depicting the evolution of the primary tumor and liver metastasis from diagnosis (day 14) to after treatment initiation (day 55). The diameter of the primary tumor lesion and the main liver metastatic tumor is shown at the top of each image. (B) ctDNA dynamics of the detected mutations in association with treatment. The black line represents the dynamics of the tumor marker CA19-9. mFFX and GnP treatment cycles are indicated with arrows. CH, clonal hematopoiesis; GnP, gemcitabine plus nab-paclitaxel; mFFX, modified FOLFIRINOX (folinic acid, fluorouracil, irinotecan, and oxaliplatin combo); ctDNA, circulating tumor DNA; Met., metastasis; GNAS, GNAS complex locus.

Figure 2 Levels of PCT and CRP in response to treatment. GnP, gemcitabine plus nab-paclitaxel; mFFX, modified FOLFIRINOX (folinic acid, fluorouracil, irinotecan, and oxaliplatin combo); PCT, procalcitonin; CRP, C-reactive protein.

Genetic analysis was performed in both tumor tissue and liver biopsies by amplicon-based next-generation sequencing (NGS) with the Ion AmpliSeq Comprehensive Cancer Panel (Thermo Fisher Scientific, Inc.) of 509 genes. Activating KRASG12D and TP53R248Q mutations, along with increased copy number variations of the proto-oncogenes MYC (MYC proto-oncogene, bHLH transcription factor) and MAF (MAF bZIP transcription factor) were detected in both primary tumor and metastasis samples (Tables I and II). Plasma samples were collected just before the first-line chemotherapy and in weeks 2 and 4 after treatment initiation, and circulating tumor DNA (ctDNA) was analyzed (Tables III and IV). Ultradeep targeted NGS with the Oncomine pan-cancer cell-free assay (Thermo Fisher Scientific) was used to investigate genetic alterations in 52 genes. ctDNA from before the first line of treatment confirmed the genetic alterations found in the primary tumor and revealed an amplification in the mutant KRASG12D allele [variant allele frequency (VAF) = 87.2%). Mutant allele amplification was detected based on the capped molecular depth (19,999x] obtained in the ctDNA sequence, which surpassed by at least 4-fold the maximum expected molecular depth in unamplified regions (4,980x) based on the sample input (16.6 ng). A 5-fold amplification of KRASG12D in the metastasis samples (VAF=91.5%) was subsequently confirmed by digital polymerase chain reaction and NGS sequencing. A GNASR201H mutation (VAF=0.05%) was also detected in the plasma liquid biopsy. Since this mutation was not present in the primary tumor tissue, genomic DNA from the WBCs was also analyzed. The GNASR201H mutation was confirmed in the WBCs (VAF=0.1%), indicating its association with clonal hematopoiesis rather than with the pancreatic tumor (Table I). Analysis of ctDNA 2 weeks after the first mFFX cycle showed an initial decrease in KRASG12D and TP53R248Q frequency (VAF = 25.3 and 3.9%, respectively). However, ctDNA analysis at week 4 of mFFX indicated an upregulation of the mutant VAF levels and in KRASG12D amplification, close to the levels prior to treatment (Fig. 1B; Table III).

Table I Mutations detected in tissue samples.

Table I

Mutations detected in tissue samples.

	Primary tissue			Liver metastasis			White blood cells
Mutation	Mol depth, x	Counts, n	VAF, %	Mol depth, x	Counts, n	VAF, %	Mol depth, x	Counts, n	VAF, %
KRAS-G12D	3,831	1,766	46.10	18,844a	17,245	91.51	3,529	0	0.00
TP53-R248Q	2,194	1,129	51.46	1,975	1,209	61.22	9,544	0	0.00
GNAS-R201H	4,073	0	0.00	4,949	0	0.00	6,812	6	0.10

[i] ^aMol depth surpassed maximum expected value for unamplified samples. GNAS, GNAS complex locus; Mol, molecular; VAF, variant allele frequency.

Table II CNVs detected in tissue samples.

Table II

CNVs detected in tissue samples.

	Primary tissue		Liver metastasis		White blood cells
Gene <CNV>	Copy no.	CNV ratio	Copy no.	CNV ratio	Copy no.	CNV ratio
KRAS	4	2.0	10	5.0	0	0
MYC	6	3.1	6	3.0	0	0
MAF	7	3.5	8	4.0	0	0

[i] CNV, copy number variation; GNAS, GNAS complex locus; MAF, MAF bZIP transcription factor.

Table III Mutations detected in ctDNA samples.

Table III

Mutations detected in ctDNA samples.

	Before treatment			2 weeks after treatment			4 weeks after treatment
Mutation	Mol depth, x	Counts, n	VAF, %	Mol depth, x	Counts, n	VAF, %	Mol depth, x	Counts, n	VAF, %
KRAS-G12D	19,999a	17,449	87.20	7,111	1,798	25.30	13,736a	9,493	69.10
TP53-R248Q	3,144	1,602	51.00	4,207	165	3.92	3,794	418	11.00
GNAS-R201H	3,787	4	0.05	6,175	5	0.08	6,606	8	0.12

[i] ^aMol depth surpassed maximum expected value for unamplified samples. Mol, molecular; VAF, variant allele frequency; GNAS, GNAS complex locus.

Table IV CNVs detected in ctDNA samples.

Table IV

CNVs detected in ctDNA samples.

	Before treatment		2 weeks after treatment		4 weeks after treatment
Gene <CNV>	Copy no.	CNV ratio	Copy no.	CNV ratio	Copy no.	CNV ratio
MYC	2.8	1.4	2.1	1.0	2.4	1.2

[i] CNV, copy number variation.

Discussion

KRAS activating mutations and TP53, CDKN2A, and SMAD4 loss-of-function alterations are the most common genetic alterations found in PDAC. Nevertheless, a large number of infrequent mutations and copy number variations in multiple genes are also detected, resulting in significant interindividual heterogeneity (7,14). In addition, the oncogenic effect of MYC is well established as a critical effector of activated RAS in several cancer types, including PDAC (15).

Allelic imbalance caused by amplification of mutant KRAS is more frequently reported in high-grade tumors of NSCLC and can affect its response to therapy (10,11). Amplification of KRASmut in PDAC, although less documented, confers an increased metastatic potential by inducing robust epithelial-mesenchymal transition signatures, being associated to worse prognosis (16). However, one of the challenges in accurately detecting gene amplification in PDAC is the presence of high stromal cell content within the tumor tissue (17), with non-neoplastic stroma confounding precise gene dosage and comprehensive interpretation of copy number alterations. Although tissue biopsies are the gold standard for diagnosis and molecular characterization of tumors, the analysis of ctDNA from liquid biopsies can avoid the interference of non-neoplastic stromal cells and capture the intrinsic influence of tumor heterogeneity during the course of the disease. In this case, we detected KRASG12D amplification in the ctDNA but not in the primary tumor. Amplification was later confirmed in the metastatic tissue, reflecting the heterogeneous evolution of the tumor. KRASG12D amplification was associated with rapid tumor growth, suggesting that it may play an important role in promoting the metastatic spread of PDAC cells. In addition, the poor response of the patient to both lines of treatment suggests that the presence of amplified KRASG12D may also impair the tumor sensitivity to chemotherapy. Recognition of this information in advance may help predict treatment-related deterioration of the patient general condition. This is of particular importance for selecting treatment approaches that consider the rate of tumor collapse to help control poor prognosis PDAC cases. Hence, albeit underrated in the clinical setting, amplification of oncogenic KRASG12D, which constitutes a key driving force that adds to an aggressive PDAC phenotype, can be detected in ctDNA through routine liquid biopsy.

This case highlights the importance of accurate detection of gene-dosage gains in oncogenic KRAS mutations in PDAC. KRASG12D amplification in combination with TP53 mutation and deregulated MYC expression may be associated an aggressive PDAC phenotype. Hence, in light of the heterogeneous characteristics of aggressive pancreatic cancers, monitoring tumor evolution through liquid biopsies can help identify such cases at earlier stages. Importantly, the amplification of oncogenic KRASG12D can be successfully detected through liquid biopsy and is feasible for implementation in the clinical setting.

Acknowledgements

The authors would like to thank Rie Hayashi, Dr Hiu Ting Chan and Dr Yoon Ming Chin for technical support.

Funding

The present study was supported by the Council for Science, Technology and Innovation (CSTI), cross-ministerial Strategic Innovation Promotion Program (SIP), ‘Innovative AI Hospital System’ [Funding Agency: National Institute of Biomedical Innovation, Health and Nutrition (NIBIOHN); grant no. SIPAIH18C03].

Availability of data and materials

All data generated or analysed during this study are included in this published article.

Authors' contributions

MM, YN, SKL, YK and FPS were involved in the conception, design and execution of the study and manuscript writing. MM, SKL and FPS collected and analyzed the data. MM and YK provided study material. MM and FPS confirm the authenticity of all the raw data. All authors read and approved the final manuscript.

Ethics approval and consent to participate

The patient provided informed consent in writing for participation in the study approved by the Japanese Foundation for Cancer Research Review Board (IRB 2018-1016).

Patient consent for publication

The patient provided informed consent in writing regarding the publication of the study approved by the Japanese Foundation for Cancer Research Review Board (IRB 2018-1016).

Competing interests

The authors declare that they have no competing interests.

References