Multi-Omic Biomarkers in Pancreatic cancer

Multi-Omic Biomarkers in Pancreatic cancer: Comparison

Please note this is a comparison between Version 4 by Peter Tang and Version 3 by Laura Kane.

Pancreatic cancer (PC) is among the most aggressive types of cancer, having caused over 495,000 deaths worldwide in 2020, with increasing annual incidence. Pancreatic cystic lesions (PCLs) are protrusions found within or on the surface of the pancreas, and in many cases have the potential to become malignant. Current methods of characterising PCLs are imperfect and there is a profound need for improved diagnostic algorithms. This review highlights the importance of biological markers in the context of PCLs and PC, with a focus on ‘omics’-related work. Successful integration of different ‘omics’ data could aid in the identification of a novel integrated biomarker profile for the risk stratification of patients with PCLs and PC.

pancreatic cancer
pancreatic cystic lesion
biomarker
risk
omics
multi-omics
pancreatic adenocarcinoma

1. Introduction

Pancreatic cancer (PC) is the 7th leading cause of cancer-related death globally ^[1]. The five-year survival rate for all pancreatic cancer stages combined is just 9%, with incidence rates continuing to rise every year ^[2]. Pancreatic cancer can be divided into two main subtypes: pancreatic adenocarcinoma, which is responsible for 85–90% of all pancreatic neoplasms and has a five-year survival rate of just 8%, and pancreatic neuroendocrine tumour (PanNET), which is far less common and represents less than 5% of PC [1,3,4]^[1][3][4]. As pancreatic adenocarcinoma is by far the most prevalent type of pancreatic cancer, it is used synonymously with PC and will be the type of PC referred to throughout this review.

Pancreatic cystic lesions (PCLs) are typically fluid-filled structures that can be found within or on the surface of the pancreas, though some may have a solid appearance ^[3]. While many PCLs are benign and show no malignant potential, others, such as intraductal papillary mucinous neoplasms (IPMNs) or mucinous cystic neoplasms (MCNs), possess the ability to undergo malignant transformation and can be regarded as precursor lesions of PC [3,4,5]^[3][4][5]. The risk factors known to be associated with PC are extensive. However, investigations into these factors are largely case–control studies and as such have notable selection and recall biases ^[6]. Risk factors for PC can be classified as modifiable and non-modifiable [6,7]^[6][7]. Modifiable risk factors include lifestyle factors such as smoking and alcohol consumption, as well as conditions such as obesity ^[6]. Non-modifiable risk factors include age, gender, ethnicity, genetic risk factors, diabetes and chronic pancreatitis ^[7].

For patients who have a family history of PC or are predisposed to malignancy due to hereditary genetic mutation, PCLs can be identified in up to one-third of such high-risk individuals [8,9]^[8][9]. Germline mutations in BRCA1 and BRCA2 have been shown to confer an increased risk in PC as well as breast and ovarian cancers ^[10]. Von Hippel–Lindau (VHL) disease, caused by a germline mutation to the VHL tumour-suppressor gene, is associated with an increased risk of pancreatic neuroendocrine tumours and non-malignant serous-type PCLs ^[11].

One of the most significant risk factors for PCLs is age, with patients typically being diagnosed at 50 years or older and the incidence rate rising exponentially with age thereafter [6,7,8,12]^{[6][7][8][12]}. PCL size and number have been shown to increase with age ^[13]. Variations in PCL prevalence from country to country can be shown to correlate with population demographics. This geographic variance is further widened by the differences in imaging resolution and the frequency of routine physical check ups within the population [3,14]^[3][14]. Indeed, a 2017 study showed a positive correlation between socioeconomic development (measured through Human Development Index and Gross Domestic Product) and pancreatic cancer incidence and mortality ^[15]. This observed increase in PC incidence with rising socioeconomic development is thought to be result of the Western lifestyle and ageing population, which are known to be large risk factors of PC ^[15]. A general improvement in imaging technologies, and also a growth in the ageing population, has caused the worldwide prevalence of PCLs to rise drastically over the last two decades [8,12]^[8][12]. The age- and sex-adjusted prevalence of PCLs in the general population is approximately 2%, but this figure increases exponentially with age and can range up to 45% in older generations [16,17,18]^[16][17][18]. Notably, while many PCLs are found in symptomatic patients, PCLs are frequently found incidentally in patients undergoing cross-sectional imaging for other reasons—so called ‘incidentalomas’. The incidence rates of such PCLs vary depending on the imaging technique, but can be as high as 19.6% in patients undergoing magnetic resonance imaging (MRI) ^[19]. With the rising prevalence of PCLs globally and poor survival rates associated with PC, there is a great need for improved characterisation of pre-malignant PCLs to allow surgery in those who need it, while avoiding unnecessary surveillance and intervention in those who do not.

2. Current Management of PCLs

When a patient with a PCL is identified, the first thing to be ascertained is its malignant potential. Broadly speaking, PCLs can be divided into either neoplastic or non-neoplastic cysts, as shown in Figure 1 ^[12]. Neoplastic cysts can be either mucinous or non-mucinous, with non-mucinous PCLs rarely undergoing malignant transformation [8,12]^[8][12]. Solid pseudopapillary neoplasms and cystic neuroendocrine neoplasms are notable, rare exceptions, as both are non-mucinous cystic lesions that do have some malignant potential and may require surgical resection. However, mucinous PCLs such as IPMNs and MCNs are generally regarded as precursor lesions for PC [8,12,20,21]^{[8][12][20][21]}. IPMNs are the most common pre-malignant PCL, being much more common than MCNs. The biological behaviour of IPMNs are notoriously unpredictable and there are currently a number of clinical guidelines that aim to help stratify the risk of IPMNs undergoing malignant transformation ^[22]. At present, there are three sets of guidelines in use to guide EUS and surgical referral of patients presenting with asymptomatic PCLs: the 2017 International Association of Pancreatology Fukuoka guidelines ^[23], the 2015 American Gastroenterological Association (AGA) guidelines ^[24], and the 2018 European evidence-based guidelines (EEG) ^[25]. The fact that there are differing consensus guidelines in use is indicative of the imperfect state of knowledge regarding PCLs and PC and the urgent need for improved biological characterisation of these lesions.

Figure 1. Molecular subgroups of pancreatic cystic lesions. Intraductal papillary mucinous neoplasms (IPMNs) and their distinct subclassifications are highlighted. IPMNs are the most common subgroup and are responsible for 38% of PCLs, while mucinous cystic neoplasms, serous cystic neoplasms and cystic neuroendocrine neoplasm represent 23%, 16% and 7% of PCLs, respectively [29]. Branch-duct IPMNs are most common (46%), followed by combined-type IPMNs (40%) and main-duct IPMNs (14%) [30].

Molecular subgroups of pancreatic cystic lesions. Intraductal papillary mucinous neoplasms (IPMNs) and their distinct subclassifications are highlighted. IPMNs are the most common subgroup and are responsible for 38% of PCLs, while mucinous cystic neoplasms, serous cystic neoplasms and cystic neuroendocrine neoplasm represent 23%, 16% and 7% of PCLs, respectively ^[26]. Branch-duct IPMNs are most common (46%), followed by combined-type IPMNs (40%) and main-duct IPMNs (14%) ^[27].

IPMNs are classified based on the involvement of the pancreatic ductal system as either main-duct (MD) IPMN, branch-duct (BD) IPMN or when both main and branch ducts are involved, combined-type IPMN [3,12]^[3][12]. Approximately 70% of MD-IPMNs undergo malignant transformation, whereas the rate is much lower in BD-IPMNs, ranging from 6 to 46% [3,26]^[3][28]. Indeed, the incidence rate of PC concomitant with BD-IPMN has been shown to range between 2% and 11.2% [27]^[29]. The World Health Organisation describes three grades of IPMN: low–intermediate-grade dysplasia; high-grade dysplasia; and IPMNs with associated invasive carcinoma ^[12]. Current methodologies and guidelines are limited in their ability to stratify patients into high- and low-risk of malignant transformation [28]^[30]. Identification of patients with high-grade dysplasia or early invasive cancer and the ability to predict those most likely to undergo malignant transformation is a key aspect of PCL patient management [24,28]^[24][30].

The most frequently used diagnostic tools for PCLs include computed tomography (CT), magnetic resonance imaging (MRI), and endoscopic ultrasound (EUS) +/- fine-needle aspiration (FNA), all of which have low sensitivity and specificity (SN/SP) for identifying high- and low-risk patients ^[24]. In the case of BD-IPMNs, risk stratification is based on cyst size and the presence or absence of a mural enhancing nodule. For MD-IPMNs, the diameter of the main pancreatic duct is accepted as an indicator of malignant risk [21,24,25]^[21][24][25]. A 2017 retrospective study assessed the ability of these two sets of guidelines to identify malignant cysts, and found that even when combined, 11.8% of malignant cysts were not identified [28]^[30]. Although this study had some self-identified bias (it included only higher risk patients), it highlighted the suboptimal performance of current clinical guidelines, even in expert centres [28]^[30]. Surgical resection is associated with significant morbidity and mortality and should be reserved for those at high risk of malignant transformation or established cancer ^[24]. Moreover, there is a 20% recurrence rate following surgical resection for IPMN ^[31], and recent studies have found multiple distinct regions of dysplasia within the pancreas, sometimes with differing mutational status of the same gene, supporting the notion of a multi-focal tumorigenic process of IPMN within the pancreas [4,31,32]^[4][31][32].

EUS-guided FNA is a safe and accurate method of extracting cyst fluid or pancreatic tissue from a patient for further analysis [26,33]^[28][33]. Cyst fluid cytology has high specificity for malignancy or high-grade dysplasia, but low sensitivity due to the typically low cellularity of PCL samples [26,33]^[28][33]. A 12 year multi-institutional study conducted by the French Surgical Association found that 50% of patient cyst fluid samples collected were non-diagnostic and acellular ^[34]. Diagnosis of PCLs by EUS requires attention to cyst morphology, including size, number of cysts present, characteristics of the wall and internal structures, calcification, positioning in relation to the main pancreatic duct and presence of lesions in the background ^[12]. These descriptors are considerably operator dependent and PCL characterisation without cyst fluid analysis is limited [3,26]^[3][28].

While the low cellularity of PCL fluid limits cytological yield, biochemical analysis of PCL fluid has proven an important adjunct in characterising PCLs. Cyst fluid CEA has been shown to have a sensitivity of between 59 and 67% and specificity of 83–91% for detection of mucinous cysts, and is among the best of the biomarkers currently available [20,35]^[20][35]. Mutational profiling of patients has shown utility in the characterisation of different PCL subtypes. However, genetic evaluation of PCL fluid is currently limited to research. KRAS and GNAS mutations in the cyst fluid are particularly important early mutations in IPMNs as they are not found in other common types of cysts ^[32]. The development of novel biomarkers within the cyst fluid has proven difficult to date due to the heterogeneic nature of the fluid. CA19-9 is a tetrasaccharide antigen released by pancreatic cancer cells and is an established marker for PC, but has low SN/SP and is not elevated in pre-malignant PCLs ^[36]. However, there is some evidence that CA19-9 is contributive, and most PCL guidelines advocate for its use in surveillance. Indeed, the importance of such minimally invasive serological biomarkers for use in tandem with non-invasive imaging to identify high-risk patients should not be underestimated. Based on all the aforementioned limitations of current diagnostics, it is clear therefore that there is an urgent need for novel methods and markers to accurately classify and risk stratify PCLs, and we believe that the ‘omics’ revolution is poised to fill this void of information.

3. Identification of Biomarkers in PCLs and PC Using Omics

The omics field has made huge strides in the past two decades, largely due to technological advancements, enabling the cost-effective and high-throughput analysis of biological molecules or ‘biomarkers’ ^[37]. Some omics disciplines are demonstrating great potential in the search for a novel biomarker for PC (Table 1), but the data for PCLs are much more limited.

Table 1. Overview of biomarkers in PC and PCLs that have been validated in an independent cohort.

Biomarker Name	Biomarker Type		Single or Multi-Study Validated		Platform			Sample Type			Sample Size (Total Number Patients)			Sensitivity (95% Confidence Interval)			Specificity (95% Confidence Interval)			P-Value		Purpose
KRAS and/OR GNAS	Genetic mutation panel		Multi		PCR			Cyst fluid			91			65%			100% (83–100)			N/A		MCN vs. non-MCN	^[38]
								(52–76)			98%
								84%			(86–100)			N/A			IPMN vs. non-IPMN
								(70–92)
									95.50%
					PCR using NGS			Cyst fluid			197			68.50%			(N/A)			N/A		IPMN vs. non-IPMN	^[39]
								(N/A)
									100%
					NGS			Cyst fluid			595			89%			(88–100)			N/A		MCN vs. non-MCN	^[40]
								(79–95)			100%
					Sanger sequencing			Cyst fluid			159			65%			(N/A)			N/A		MCN vs. non-MCN
								(N/A)
lncRNA-TFG	Long noncoding RNA		Single		Affymetrix human exon 1.0 ST			Tissue			28			N/A			N/A			6.23 × 10⁻⁸		Positive correlation with tumorigenesis in IPMNs	^[41]
CTD-2033D15.2	Long non-coding RNA		Single		Affymetrix human exon 1.0 ST			Tissue			28			N/A			N/A			1.47 × 10⁻⁴		Negative correlation with tumorigenesis in IPMNs	^[41]
HAND2-AS1	Long non-coding RNA		Single		Affymetrix human exon 1.0 ST			Tissue			28			N/A			N/A			2.66 × 10⁻³		Negative correlation with tumorigenesis in IPMNs	^[41]
Glucose	Metabolite		Multi		Liquid chromatography			Cyst fluid			19			94%			64%			0.004		Glucose ≤ 66 mg/dL in MCNs vs. non-MCNs	^[42]
								(N/A)			(N/A)
					Glucometer								Glucose ≤ 50 mg/dL in MCNs vs. non-MCNs
						Cyst fluid			153			92%			87%			N/A			^[43]
								(N/A)			(N/A)
Kynurenine	Metabolite		Single		Liquid chromatography			Cyst fluid			19			90%			100%			0.002		Lower in MCNs vs. non-MCNs	^[42]
Kynurenine	Metabolite		Single		Liquid chromatography			Cyst fluid			19			(N/A)			(N/A)			0.002		Lower in MCNs vs. non-MCNs	^[42]
AcSperm and	Metabolite panel		Single		Mass spectrometry			Blood plasma			121			66.70%			95%			N/A		PDAC vs. N	^[44]
DAS and														(N/A)			(N/A)
LPC(18: 0) and LPC(20: 3) and indole derivative
ADAMTS1	Methylated gene		Single		Methylation on beads			Blood cfDNA			39			87.20%			95.80%			N/A		PDAC vs. N	^[45]
ADAMTS1	Methylated gene		Single		Methylation on beads			Blood cfDNA			39			(N/A)			(N/A)			N/A		PDAC vs. N	^[45]
BNC1	Methylated gene		Single		Methylation on beads			Blood cfDNA			39			64.10%			93.70%			N/A		PDAC vs. N	^[45]
BNC1	Methylated gene		Single		Methylation on beads			Blood cfDNA			39			(N/A)			(N/A)			N/A		PDAC vs. N	^[45]
SOX17	Methylated gene		Single		Methylation-specific ddPCR			Cyst fluid			154			78.40%			85.60%			N/A		High-risk PCL vs. low-risk PCLs	^[46]
SOX17	Methylated gene		Single		Methylation-specific ddPCR			Cyst fluid			154			(64.7–88.7)			(78.4–91.1)			N/A		High-risk PCL vs. low-risk PCLs	^[46]
TBX15 and BMP3	Methylated gene marker panel		Single		Whole-genome methylome discovery and qPCR			Cyst fluid			134			90%			92%			N/A		HGD/PC vs. LGD/N	^[47]
TBX15 and BMP3	Methylated gene marker panel		Single		Whole-genome methylome discovery and qPCR			Cyst fluid			134			(70–99)			(85–96)			N/A		HGD/PC vs. LGD/N	^[47]
ADAMTS1 and/OR BNC1	Methylated gene panel		Single		Methylation on beads			Blood cfDNA			39			97.40%			91.60%			N/A		PDAC vs. N	^[45]
ADAMTS1 and/OR BNC1	Methylated gene panel		Single		Methylation on beads			Blood cfDNA			39			(N/A)			(N/A)			N/A		PDAC vs. N	^[45]
FOXE1 and SLIT2 and	Methylated gene panel		Single		Methylation-specific ddPCR			Cyst fluid			154			84.30%			89.40%			N/A		High-risk PCL vs. low-risk PCLs	^[46]
EYA4 and SFRP1	Methylated gene panel		Single		Methylation-specific ddPCR			Cyst fluid			154			(N/A)			(N/A)			N/A		High-risk PCL vs. low-risk PCLs	^[46]
miR-1290	MicroRNA		Multi		MicroRNA array analysis			Blood serum			60			88%			84%			N/A		PC vs. N	^[48]
								(N/A)			(N/A)
							76			83%			69%			N/A			PC vs. CP
								(N/A)			(N/A)
							95			83%			78%			N/A			PC vs. CP and N
								(N/A)			(N/A)
					qRT-PCR			Blood plasma
							49			N/A			N/A			0.027			PDAC vs. N		^[49]
						Blood serum
					qRT-PCR
							200			74.20%			91.20%			N/A			PC vs. C		^[50]
								(N/A)			(N/A)
9-miRNA model ^a	MicroRNA panel		Single		TaqMan miRNA Array			Tissue and			33 and 50			89%			100%			N/A		HG IPMNs, PanNETs and SPNs vs. LG IPMNs and SCAs	^[51]
9-miRNA model ^a	MicroRNA panel		Single		TaqMan miRNA Array			cyst fluid			33 and 50			(N/A)			(N/A)			N/A		HG IPMNs, PanNETs and SPNs vs. LG IPMNs and SCAs	^[51]
miR-3679-5p and miR-940	MicroRNA panel		Single		qPCR			Saliva			80			72.50%			70.00%			N/A		PC vs. N	^[52]
												(N/A)			(N/A)
											60			62.50%			80.00%			N/A		PC vs. BPT
												(N/A)			(N/A)
											100			70.00%			70.00%			N/A		PC vs. N and BPT
												(N/A)			(N/A)
CA19-9	Protein-		Multi		Bead-based xMAP immunoassay			Blood serum			267			57.20%			90%			N/A		PDAC vs. N	^[53]
	associated							(N/A)			(N/A)
		ELISA
				Blood plasma
					176			77.50%			83.10%			N/A			CA19-9 >20.3 U/mL		^[54]
		Retrospective clinical data				Blood serum				(N/A)			(N/A)				PDAC vs. C
					41			90%			83.33%			N/A			2.45 times elevated CA19-9 indicated recurrence of PC		^[55]
						(N/A)			(N/A)
CEA	Protein		Multi		Clinical data			Cyst fluid			31			73%			89%			N/A		CEA > 192 ;ng/mL for MCN	^[42]
								(N/A)			(N/A)
					ELISA			Cyst fluid			149			95.50%			81.50%			<0.0001		CEA ≤ 10 ng/mL for SCN	^[56]
								(N/A)			(N/A)
					Enzyme-linked immunosorbent assay			Cyst fluid			153			58%			96%			N/A		CEA > 192 ng/mL for MC	^[43]
								(N/A)			(N/A)
MUC5AC:WGA and MUC5AC:BGH and Endorepellin:WGA	Protein		Multi		Antibody-lectin sandwich microarray			Cyst fluid			147			92% ^b			94% ^b			N/A		Elevation in any two differentiates MCNs vs. non-MCNs	^[57]
	panel							(N/A)			(N/A)
		Antibody-lectin sandwich arrays									Elevation in any two differentiates MCNs vs. non-MCNs
				Cyst fluid			22			87%			100%			N/A			^[58]
						(N/A)			(N/A)
Thymosin- β4	Protein		Single		MALDI imaging and mass spectrometry			Tissue			45			70%			71%			0.011		Overexpressed in IPMN with HGD	^[59]
Thymosin- β4	Protein		Single		MALDI imaging and mass spectrometry			Tissue			45			(N/A)			(N/A)			0.011		Overexpressed in IPMN with HGD	^[59]
Ubiquitin	Protein		Single		MALDI imaging and mass spectrometry			Tissue			45			94%			86%			0.04		Overexpressed in IPMN with HGD	^[59]
Ubiquitin	Protein		Single		MALDI imaging and mass spectrometry			Tissue			45			(N/A)			(N/A)			0.04		Overexpressed in IPMN with HGD	^[59]
VEGF-A	Protein		Single		ELISA			Cyst fluid			149			100%			83.70%			<0.0001		VEGF-A > 5000 pg/mL benign SCN	^[56]
VEGF-A	Protein		Single		ELISA			Cyst fluid			149			(N/A)			(N/A)			<0.0001		VEGF-A > 5000 pg/mL benign SCN	^[56]
VEGF-A and CEA	Protein panel		Single		ELISA			Cyst fluid			149			95.50%			100%			N/A		VEGF-A > 5000 pg/mL and CEA ≤ 10 ng/mL in benign SCN	^[56]

BPT = benign pancreatic tumour, C = non-cancer control, CP = chronic pancreatitis, ELISA = enzyme-linked immunosorbent assay, HG = high grade, HGD = high-grade dysplasia, IPMN = intraductal papillary mucinous neoplasm, LG = low grade, LGD = low-grade dysplasia, MALDI = matrix-assisted laser desorption ionisation, MC = mucinous cyst, MCN = mucinous cystic neoplasm, N = normal healthy, N/A = not available, NGS = next-generation sequencing, PanNET = pancreatic neuroendocrine tumour, PC = pancreatic cancer, PCL = pancreatic cystic lesion, PCR = polymerase chain reaction, PDAC = pancreatic ductal adenocarcinoma, SCN = serous cystic neoplasm, and SPN = solid-pseudopapillary neoplasm. ^a Model is intellectual property of the authors. ^b Average of three cohorts.

4. Multi-Omics as the Key to Biomarker Identification

In the case of PCL and PC characterisation, though these new ‘omics’ techniques have been utilised to analyse the pancreatic cyst fluid, blood serum and even saliva of patients, no single methodology has proven to be a sufficiently sensitive method for delineating these patients into defined categories. Multi-omics involves the integration of multiple layers of omics-type data to augment our understanding of disease and helps researchers to elucidate the flow of information, from the origin of the disease to the biological and functional consequences ^[37]. By investigating multiple aspects of the PCL fluid or blood serum, and treating these data as an interconnected system, rather than distinct and independent pieces, multi-omics could allow researchers to identify key pathways and players in disease stratification. CompCyst is a comprehensive test developed using machine learning techniques to guide the management of patients with PCLs [113]^[60]. This test utilises selected clinical features such as symptoms, cyst size and location, as well as cyst fluid genetic and biochemical markers, including cyst CEA levels and KRAS and GNAS mutation status [113]^[60]. Interrogation of multiple levels of patient data enabled cut offs for each marker to be determined based on the needs of the test, and the level of importance given to the sensitivity or specificity of each individual marker. The results of this study suggest that if CompCyst were applied to general PCL management, 60% of unnecessary surgeries could be avoided [113]^[60]. While these results seem promising, it is important to note that patients evaluated in this study were those most concerning for cancer and do not represent patients seen in routine clinical practice [113]^[60]. While more research is needed to examine the utility of this test in a normal clinical setting, this study shows that layering multiple levels of patient data can potentially improve management strategies for PCLs. Interestingly, a 2018 report describes a multi-analyte blood test called CancerSeek, which assesses levels of circulating proteins and mutations in cell-free DNA to detect one of eight common cancer types (ovarian, liver, stomach, pancreatic, oesophageal, colorectal, lung, or breast) [114]^[61]. When combined with supervised machine learning, this test was able to localise the source of the cancer to two anatomic sites in a median of 83% of patients (n = 626). While not specific to PC, this test shows the stark advantage of combining distinct approaches to create a robust diagnostic tool.

In terms of PCLs and PC, multi-omics opens the door to the possibility of a biomarker panel for characterisation, such that combined thresholds of several markers could prove more sensitive than a single marker alone. A 2020 systematic review examined novel biomarkers for upper GI cancers, identifying 431 biomarkers, of which more than half (n = 231) were for PC [115]^[62]. Only one-fifth of the biomarkers reported in this review were examined in more than one study, and of those that were, there were only two single markers and one panel of markers for PC. Such reviews of the literature show the current state of PC research, where most of those markers that are identified are not examined further and as such never become clinically useful. As mentioned previously, the addition of GNAS to KRAS testing for the diagnosis of IPMNs does not significantly increase diagnostic accuracy. However, the same study found that the combination of GNAS and KRAS mutational status with CEA testing does produce a significantly better accuracy of 86.2% (p < 0.05) ^[39]. A 2015 multi-centre study retrospectively examined the cyst fluid of 130 patients and identified molecular markers and clinical features that classified PCLs with a sensitivity of 90–100% and a specificity of 92–98% [116]^[63]. Using the Multivariate Organisation of Combinatorial Alterations (MOCA) algorithm to identify composite clinical and molecular markers (subtle mutations, loss-of-heterozygosity, aneuploidy) of PCL type and grade, this study identified a panel of both clinical and molecular markers for the distinction of serous cystadenomas (SCA), solid-pseudopapillary neoplasm (SPN), MCNs and IPMNs. Furthermore, it was shown that these features could identify 67 of the 74 patients who did not require surgery, resulting in a reduction in unnecessary procedures by 91%. These results show great promise for the characterisation of PCLs and the stratification of patients for subsequent referral to surgery, and further studies in more robust, experimental validation cohorts will help to further elucidate the potential of this panel in the context of PC.

A key example of the multi-omic nature of driver mutations in the context of PC is KRAS, which is mutated in ~90% of PC (Figure 2) [117]^[64]. Environmental factors, such as smoking or alcohol consumption, can promote biochemical alterations to DNA at the epigenomic level, for example hypermethylation [118]^[65]. The addition of a methyl group to the CpG island of a DNA repair gene can cause silencing and subsequently result in reduced DNA repair proficiency, allowing a mutated KRAS codon to proceed from the genomic level to the transcriptomic level. The transcription of this KRAS mutation results in altered miRNA expression levels, and the mutated mRNA cannot be bound by the regulatory miRNA let-7, thus causing the aberrant translation of K-Ras protein [119,120,121]^[66][67][68]. The marked increase in K-Ras production promotes various signalling pathways, including phosphoinositide 3-kinase (PI3K), mitogen-activated protein kinase (MAPK) and the RAL-GEFs pathway [117]^[64]. GTP-bound K-Ras proteins can interact with, and influence the activity of, effector proteins causing downstream effects in many cellular pathways [117]^[64]. Moreover, KRAS mutant PCLs have been shown to have increased expression of the glucose transporter GLUT1 and subsequently elevated rates of glycolysis, indicating that KRAS mutations play a role in the metabolic switch observed in PC [122]^[69]. Indeed, the presence of KRAS mutations in PC has been shown to correlate with poor patient prognosis, and this can be attributed to the downstream effects seen in multiple omics layers as a result of this point mutation (Figure 2) [123]^[70]. This example illustrates the multi-omic nature of mutational drivers in cancer and the importance of disentangling each aspect in order to clearly observe the pathways affected and the impacts at each omics level.

Figure 2. The multi-omic nature of KRAS mutations in pancreatic cancer. (A) Environmental factors cause biochemical alterations to the DNA such as hypermethylation. This can result in the silencing of repair genes and subsequently failure in DNA repair pathways; (B) Point mutations in a KRAS codon go unchecked as a result of DNA repair failure causing permanent activation of KRAS gene; (C) Mutant KRAS gene is transcribed into mRNA and subsequently results in an upregulation of miR-34a and miR-31 and a downregulation in miR143 and miR-145; (D) mutant KRAS mRNA cannot be bound by regulatory miR let-7 and leaves the cell nucleus to be translated; (E) Mutant KRAS causes an increase in K-Ras protein expression, which causes activation of PI3K, MAPK and RAL-GEF pathways; (F) GTP bound KRAS interacts with various effector proteins and influences the localisation and activities of these effectors; (G) K-Ras proteins convert GTP to GDP which causes gene deactivations and metabolic alterations such as an increase in GLUT1 expression and subsequently an increase in glucose uptake via glycolysis; (H) Changes to cellular protein expression, gene activation and metabolic processes results in increased cell growth and proliferation, driving transformation.

5. Conclusions

PC is an aggressive disease with extremely poor survival rates. The discovery of precursor lesions often occurs too late, and patients are left with few treatment options. PCLs are a highly diverse group of lesions containing both non-malignant and pre-malignant subtypes, and there exists no robust method for distinguishing PCLs and subsequently, which patients have a high-risk of developing PC and should undergo surgical resection, and which patients are at a lower risk and can be spared this procedure. The advent of omics has enabled significant strides in the detection and treatment of cancer. Unfortunately, for patients with PCLs or PC, current individualised omics studies have produced little success. Multi-omics provides a more comprehensive insight into the mechanisms and pathways involved in cancer and has great potential for use in diagnosis and treatment of PC. The presence of many omics databases online, which are publicly available and contain vast quantities of patient data, enables the interrogation of large datasets and the production of highly powered studies. However, there are many facets of the integration of data that must be acknowledged and sufficiently managed in order for these studies to produce accurate and robust results. There is a need for standardisation of multi-omics approaches in this way, such that more in-depth analyses can be carried out. The discipline of multi-omics imparts much expectation for the further understanding of PC as a disease, and the identification of biological markers that may aid in the characterisation of patient PCLs.

References

Rawla, P.; Sunkara, T.; Gaduputi, V. Epidemiology of pancreatic cancer: Global trends, etiology and risk factors. World J. Oncol. 2019, 10, 10.
Siegel, R.L.; Miller, K.D.; Jemal, A. Cancer Statistics, 2019. CA A Cancer J. Clin. 2019, 69, 7–34.
Park, J.; Han, D.; Do, M.; Woo, J.; Wang, J.I.; Han, Y.; Kwon, W.; Kim, S.W.; Jang, J.Y.; Kim, Y. Proteome characterization of human pancreatic cyst fluid from intraductal papillary mucinous neoplasm by liquid chromatography/tandem mass spectrometry. Rapid Commun. Mass Spectrom. 2017, 31, 1761–1772.
Soufi, M.; Yip-Schneider, M.T.; Carr, R.A.; Roch, A.M.; Wu, H.H.; Schmidt, C.M. Multifocal High-Grade Pancreatic Precursor Lesions: A Case Series and Management Recommendations. J. Pancreat. Cancer 2019, 5, 8–11.
Matthaei, H.; Schulick, R.D.; Hruban, R.H.; Maitra, A. Cystic precursors to invasive pancreatic cancer. Nat. Rev. Gastroenterol. Hepatol. 2011, 8, 141.
McGuigan, A.; Kelly, P.; Turkington, R.C.; Jones, C.; Coleman, H.G.; McCain, R.S. Pancreatic cancer: A review of clinical diagnosis, epidemiology, treatment and outcomes. World J. Gastroenterol. 2018, 24, 4846.
Midha, S.; Chawla, S.; Garg, P.K. Modifiable and non-modifiable risk factors for pancreatic cancer: A review. Cancer Lett. 2016, 381, 269–277.
Stark, A.; Donahue, T.R.; Reber, H.A.; Hines, O.J. Pancreatic cyst disease: A review. JAMA 2016, 315, 1882–1893.
Tersmette, A.C.; Petersen, G.M.; Offerhaus, G.J.A.; Falatko, F.C.; Brune, K.A.; Goggins, M.; Rozenblum, E.; Wilentz, R.E.; Yeo, C.J.; Cameron, J.L. Increased risk of incident pancreatic cancer among first-degree relatives of patients with familial pancreatic cancer. Clin. Cancer Res. 2001, 7, 738–744.
Greer, J.B.; Whitcomb, D.C. Role of BRCA1 and BRCA2 mutations in pancreatic cancer. Gut 2007, 56, 601–605.
Tirosh, A.; Sadowski, S.M.; Linehan, W.M.; Libutti, S.K.; Patel, D.; Nilubol, N.; Kebebew, E. Association of VHL genotype with pancreatic neuroendocrine tumor phenotype in patients with von Hippel–Lindau disease. JAMA Oncol. 2018, 4, 124–126.
Kromrey, M.-L.; Bülow, R.; Hübner, J.; Paperlein, C.; Lerch, M.M.; Ittermann, T.; Völzke, H.; Mayerle, J.; Kühn, J.-P. Prospective study on the incidence, prevalence and 5-year pancreatic-related mortality of pancreatic cysts in a population-based study. Gut 2018, 67, 138–145.
Torisu, Y.; Takakura, K.; Kinoshita, Y.; Tomita, Y.; Nakano, M.; Saruta, M. Pancreatic cancer screening in patients with presumed branch-duct intraductal papillary mucinous neoplasms. World J. Clin. Oncol. 2019, 10, 67.
Wong, M.C.; Jiang, J.Y.; Liang, M.; Fang, Y.; Yeung, M.S.; Sung, J.J. Global temporal patterns of pancreatic cancer and association with socioeconomic development. Sci. Rep. 2017, 7, 1–9.
Brugge, W.R. Diagnosis and management of cystic lesions of the pancreas. J. Gastrointest. Oncol. 2015, 6, 375.
Maggi, G.; Guarneri, G.; Gasparini, G.; Fogliati, A.; Partelli, S.; Falconi, M.; Crippa, S. Pancreatic cystic neoplasms: What is the most cost-effective follow-up strategy? Endosc. Ultrasound 2018, 7, 319.
de Jong, K.; Nio, C.Y.; Hermans, J.J.; Dijkgraaf, M.G.; Gouma, D.J.; van Eijck, C.H.; van Heel, E.; Klass, G.; Fockens, P.; Bruno, M.J. High prevalence of pancreatic cysts detected by screening magnetic resonance imaging examinations. Clin. Gastroenterol. Hepatol. 2010, 8, 806–811.
Chang, Y.R.; Park, J.K.; Jang, J.-Y.; Kwon, W.; Yoon, J.H.; Kim, S.-W. Incidental pancreatic cystic neoplasms in an asymptomatic healthy population of 21,745 individuals: Large-scale, single-center cohort study. Medicine 2016, 95, e5535.
Sureka, B.; Varshney, V. Pancreatic Incidentalomas: Review and Current Management Recommendations. Ann. Natl. Acad. Med. Sci. 2019, 55, 006–013.
Sakhdari, A.; Moghaddam, P.A.; Ok, C.Y.; Walter, O.; Tomaszewicz, K.; Caporelli, M.-L.; Meng, X.; LaFemina, J.; Whalen, G.; Belkin, E.; et al. Somatic molecular analysis augments cytologic evaluation of pancreatic cyst fluids as a diagnostic tool. Oncotarget 2018, 10, 4026–4037.
Tanaka, M.; Fernández-del Castillo, C.; Adsay, V.; Chari, S.; Falconi, M.; Jang, J.-Y.; Kimura, W.; Levy, P.; Pitman, M.B.; Schmidt, C.M. International consensus guidelines 2012 for the management of IPMN and MCN of the pancreas. Pancreatology 2012, 12, 183–197.
Hanania, A.N.; Bantis, L.E.; Feng, Z.; Wang, H.; Tamm, E.P.; Katz, M.H.; Maitra, A.; Koay, E.J. Quantitative imaging to evaluate malignant potential of IPMNs. Oncotarget 2016, 7, 85776.
Tanaka, M.; Fernández-del, C.; Kamisawa, T.; Jang, J.Y.; Levy, P.; Ohtsuka, T.; Salvia, R.; Shimizu, Y.; Tada, M.; Wolfgang, C.L. Revisions of international consensus Fukuoka guidelines for the management of IPMN of the pancreas. Pancreatology 2017, 17, 738–753.
Vege, S.S.; Ziring, B.; Jain, R.; Moayyedi, P.; Adams, M.A.; Dorn, S.D.; Dudley-Brown, S.L.; Flamm, S.L.; Gellad, Z.F.; Gruss, C.B. American gastroenterological association institute guideline on the diagnosis and management of asymptomatic neoplastic pancreatic cysts. Gastroenterology 2015, 148, 819–822.
European Study Group on Cystic Tumours of the Pancreas, European evidence-based guidelines on pancreatic cystic neoplasms. Gut 2018, 67, 789–804.
Valsangkar, N.P.; Morales-Oyarvide, V.; Thayer, S.P.; Ferrone, C.R.; Wargo, J.A.; Warshaw, A.L.; Castillo, C.F.-d. 851 resected cystic tumors of the pancreas: A 33-year experience at the Massachusetts General Hospital. Surgery 2012, 152, S4–S12.
Fritz, S.; Klauss, M.; Bergmann, F.; Strobel, O.; Schneider, L.; Werner, J.; Hackert, T.; Büchler, M.W. Pancreatic main-duct involvement in branch-duct IPMNs: An underestimated risk. Ann. Surg. 2014, 260, 848–856.
Maker, A.V.; Carrara, S.; Jamieson, N.B.; Pelaez-Luna, M.; Lennon, A.M.; Molin, M.D.; Scarpa, A.; Frulloni, L.; Brugge, W.R. Cyst fluid biomarkers for intraductal papillary mucinous neoplasms of the pancreas: A critical review from the international expert meeting on pancreatic branch-duct-intraductal papillary mucinous neoplasms. J. Am. Coll. Surg. 2015, 220, 243–253.
Tanaka, M. Intraductal papillary mucinous neoplasm of the pancreas as the main focus for early detection of pancreatic adenocarcinoma. Pancreas 2018, 47, 544–550.
Lee, A.; Kadiyala, V.; Lee, L.S. Evaluation of AGA and Fukuoka Guidelines for EUS and surgical resection of incidental pancreatic cysts. Endosc. Int. Open 2017, 5, E116–E122.
Miller, J.R.; Meyer, J.E.; Waters, J.A.; Al-Haddad, M.; DeWitt, J.; Sherman, S.; Lillemoe, K.D.; Schmidt, C.M. Outcome of the pancreatic remnant following segmental pancreatectomy for non-invasive intraductal papillary mucinous neoplasm. HPB 2011, 13, 759–766.
Tan, M.C.; Basturk, O.; Brannon, A.R.; Bhanot, U.; Scott, S.N.; Bouvier, N.; LaFemina, J.; Jarnagin, W.R.; Berger, M.F.; Klimstra, D. GNAS and KRAS mutations define separate progression pathways in intraductal papillary mucinous neoplasm-associated carcinoma. J. Am. Coll. Surg. 2015, 220, 845–854.e1.
Maker, A.V.; Lee, L.S.; Raut, C.P.; Clancy, T.E.; Swanson, R.S. Cytology from pancreatic cysts has marginal utility in surgical decision-making. Ann. of Surg. Oncol. 2008, 15, 3187–3192.
Le Borgne, J.; de Calan, L.; Partensky, C.; Association, F.S. Cystadenomas and cystadenocarcinomas of the pancreas: A multiinstitutional retrospective study of 398 cases. Ann. Surg. 1999, 230, 152.
Thornton, G.; McPhail, M.; Nayagam, S.; Hewitt, M.; Vlavianos, P.; Monahan, K. Endoscopic ultrasound guided fine needle aspiration for the diagnosis of pancreatic cystic neoplasms: A meta-analysis. Pancreatology 2013, 13, 48–57.
Hasan, S.; Jacob, R.; Manne, U.; Paluri, R. Advances in pancreatic cancer biomarkers. Oncol. Rev. 2019, 13, 410.
Hasin, Y.; Seldin, M.; Lusis, A. Multi-omics approaches to disease. Genome Biol. 2017, 18, 83.
Singhi, A.D.; Nikiforova, M.N.; Fasanella, K.E.; McGrath, K.M.; Pai, R.K.; Ohori, N.P.; Bartholow, T.L.; Brand, R.E.; Chennat, J.S.; Lu, X. Preoperative GNAS and KRAS testing in the diagnosis of pancreatic mucinous cysts. Clin. Cancer Res. 2014, 20, 4381–4389.
Kadayifci, A.; Atar, M.; Wang, J.L.; Forcione, D.G.; Casey, B.W.; Pitman, M.B.; Brugge, W.R. Value of adding GNAS testing to pancreatic cyst fluid KRAS and carcinoembryonic antigen analysis for the diagnosis of intraductal papillary mucinous neoplasms. Dig. Endosc. 2017, 29, 111–117.
Singhi, A.D.; McGrath, K.; Brand, R.E.; Khalid, A.; Zeh, H.J.; Chennat, J.S.; Fasanella, K.E.; Papachristou, G.I.; Slivka, A.; Bartlett, D.L. Preoperative next-generation sequencing of pancreatic cyst fluid is highly accurate in cyst classification and detection of advanced neoplasia. Gut 2018, 67, 2131–2141.
Ding, J.; Li, Y.; Zhang, Y.; Fan, B.; Li, Q.; Zhang, J.; Zhang, J. Identification of key lncRNAs in the tumorigenesis of intraductal pancreatic mucinous neoplasm by coexpression network analysis. Cancer Med. 2020, 9, 3840–3851.
Park, W.G.; Wu, M.; Bowen, R.; Zheng, M.; Fitch, W.L.; Pai, R.K.; Wodziak, D.; Visser, B.C.; Poultsides, G.A.; Norton, J.A. Metabolomic-derived novel cyst fluid biomarkers for pancreatic cysts: Glucose and kynurenine. Gastrointest. Endosc. 2013, 78, 295–302.e2.
Carr, R.A.; Yip-Schneider, M.T.; Simpson, R.E.; Dolejs, S.; Schneider, J.G.; Wu, H.; Ceppa, E.P.; Park, W.; Schmidt, C.M. Pancreatic cyst fluid glucose: Rapid, inexpensive, and accurate diagnosis of mucinous pancreatic cysts. Surgery 2018, 163, 600–605.
Fahrmann, J.F.; Bantis, L.E.; Capello, M.; Scelo, G.; Dennison, J.B.; Patel, N.; Murage, E.; Vykoukal, J.; Kundnani, D.L.; Foretova, L. A plasma-derived protein-metabolite multiplexed panel for early-stage pancreatic cancer. J. Natl. Cancer Inst. 2019, 111, 372–379.
Eissa, M.A.; Lerner, L.; Abdelfatah, E.; Shankar, N.; Canner, J.K.; Hasan, N.M.; Yaghoobi, V.; Huang, B.; Kerner, Z.; Takaesu, F. Promoter methylation of ADAMTS1 and BNC1 as potential biomarkers for early detection of pancreatic cancer in blood. Clin. Epigenetics 2019, 11, 59.
Hata, T.; Molin, M.D.; Hong, S.-M.; Tamura, K.; Suenaga, M.; Yu, J.; Sedogawa, H.; Weiss, M.J.; Wolfgang, C.L.; Lennon, A.M. Predicting the grade of dysplasia of pancreatic cystic neoplasms using cyst fluid DNA methylation markers. Clin. Cancer Res. 2017, 23, 3935–3944.
Majumder, S.; Taylor, W.R.; Yab, T.C.; Berger, C.K.; Dukek, B.A.; Cao, X.; Foote, P.H.; Wu, C.W.; Mahoney, D.W.; Aslanian, H.R. Novel Methylated DNA Markers Discriminate Advanced Neoplasia in Pancreatic Cysts: Marker Discovery, Tissue Validation, and Cyst Fluid Testing. Am. J. Gastroenterol. 2019, 114, 1539–1549.
Li, A.; Yu, J.; Kim, H.; Wolfgang, C.L.; Canto, M.I.; Hruban, R.H.; Goggins, M. MicroRNA array analysis finds elevated serum miR-1290 accurately distinguishes patients with low-stage pancreatic cancer from healthy and disease controls. Clin. Cancer Res. 2013, 19, 3600–3610.
Karasek, P.; Gablo, N.; Hlavsa, J.; Kiss, I.; Vychytilova-Faltejskova, P.; Hermanova, M.; Kala, Z.; Slaby, O.; Prochazka, V. Pre-operative plasma miR-21-5p is a sensitive biomarker and independent prognostic factor in patients with pancreatic ductal adenocarcinoma undergoing surgical resection. Cancer Genom. Proteom. 2018, 15, 321–327.
Wei, J.; Yang, L.; Wu, Y.-n.; Xu, J. Serum miR-1290 and miR-1246 as Potential Diagnostic Biomarkers of Human Pancreatic Cancer. J. Cancer 2020, 11, 1325–1333.
Matthaei, H.; Wylie, D.; Lloyd, M.B.; Molin, M.D.; JKemppainen Mayo, S.C.; Wolfgang, C.L.; Schulick, R.D.; Langfield, L.; Andruss, B.F. miRNA biomarkers in cyst fluid augment the diagnosis and management of pancreatic cysts. Clin. Cancer Res. 2012, 18, 4713–4724.
Xie, Z.; Yin, X.; Gong, B.; Nie, W.; Wu, B.; Zhang, X.; Huang, J.; Zhang, P.; Zhou, Z.; Li, Z. Salivary microRNAs show potential as a noninvasive biomarker for detecting resectable pancreatic cancer. Cancer Prev. Res. 2015, 8, 165–173.
Brand, R.E.; Nolen, B.M.; Zeh, H.J.; Allen, P.J.; Eloubeidi, M.A.; Goldberg, M.; Elton, E.; Arnoletti, J.P.; Christein, J.D.; Vickers, S.M. Serum biomarker panels for the detection of pancreatic cancer. Clin. Cancer Res. 2011, 17, 805–816.
Chan, A.; Prassas, I.; Dimitromanolakis, A.; Brand, R.E.; Serra, S.; Diamandis, E.P.; Blasutig, I.M. Validation of biomarkers that complement CA19. 9 in detecting early pancreatic cancer. Clin. Cancer Res. 2014, 20, 5787–5795.
Azadeh, A.; Felix, R.; Krause, T.; Bernhardt, M.; Jo, P.; König, A.; Mathias, K.; Andreas, L.; Ghadimi, M.; Jochen, G. CA19-9 for detecting recurrence of pancreatic cancer. Sci. Rep. 2020, 10, 1–10.
Carr, R.A.; Yip-Schneider, M.T.; Dolejs, S.; Hancock, B.A.; Wu, H.; Radovich, M.; Schmidt, C.M. Pancreatic cyst fluid vascular endothelial growth factor A and carcinoembryonic antigen: A highly accurate test for the diagnosis of serous cystic neoplasm. J. Am. Coll. Surg. 2017, 225, 93–100.
Sinha, J.; Cao, Z.; Dai, J.; Tang, H.; Partyka, K.; Hostetter, G.; Simeone, D.M.; Feng, Z.; Allen, P.J.; Brand, R.E. A gastric glycoform of MUC5AC is a biomarker of mucinous cysts of the pancreas. PLoS ONE 2016, 11, e0167070.
Cao, Z.; Maupin, K.; Curnutte, B.; Fallon, B.; Feasley, C.L.; Brouhard, E.; Kwon, R.; West, C.M.; Cunningham, J.; Brand, R. Specific glycoforms of MUC5AC and endorepellin accurately distinguish mucinous from nonmucinous pancreatic cysts. Mol. Cell. Proteom. 2013, 12, 2724–2734.
Rebours, V.; le Faouder, J.; Laouirem, S.; Mebarki, M.; Albuquerque, M.; Camadro, J.-M.; Léger, T.; Ruszniewski, P.; Lévy, P.; Paradis, V. In situ proteomic analysis by MALDI imaging identifies ubiquitin and thymosin-β4 as markers of malignant intraductal pancreatic mucinous neoplasms. Pancreatology 2014, 14, 117–124.
Springer, S.; Masica, D.L.; Molin, M.D.; Douville, C.; Thoburn, C.J.; Afsari, B.; Li, L.; Cohen, J.D.; Thompson, E.; Allen, P.J. A multimodality test to guide the management of patients with a pancreatic cyst. Sci. Transl. Med. 2019, 11, eaav4772.
Cohen, J.D.; Li, L.; Wang, Y.; Thoburn, C.; Afsari, B.; Danilova, L.; Douville, C.; Javed, A.A.; Wong, F.; Mattox, A. Detection and localization of surgically resectable cancers with a multi-analyte blood test. Science 2018, 359, 926–930.
Calanzani, N.; Druce, P.E.; Snudden, C.; Milley, K.M.; Boscott, R.; Behiyat, D.; Saji, S.; Martinez-Gutierrez, J.; Oberoi, J.; Funston, G. Identifying Novel Biomarkers Ready for Evaluation in Low-Prevalence Populations for the Early Detection of Upper Gastrointestinal Cancers: A Systematic Review. Adv. Ther. 2020, 1–42.
Springer, S.; Wang, Y.; Molin, M.D.; Masica, D.L.; Jiao, Y.; Kinde, I.; Blackford, A.; Raman, S.P.; Wolfgang, C.L.; Tomita, T. A combination of molecular markers and clinical features improve the classification of pancreatic cysts. Gastroenterology 2015, 149, 1501–1510.
Liu, P.; Wang, Y.; Li, X. Targeting the untargetable KRAS in cancer therapy. Acta Pharm. Sin. B 2019, 9, 871–879.
Natale, F.; Vivo, M.; Falco, G.; Angrisano, T. Deciphering DNA methylation signatures of pancreatic cancer and pancreatitis. Clin. Epigenetics 2019, 11, 132.
Kent, O.A.; Chivukula, R.R.; Mullendore, M.; Wentzel, E.A.; Feldmann, G.; Lee, K.H.; Liu, S.; Leach, S.D.; Maitra, A.; Mendell, J.T. Repression of the miR-143/145 cluster by oncogenic Ras initiates a tumor-promoting feed-forward pathway. Genes Dev. 2010, 24, 2754–2759.
Kent, O.A.; Mendell, J.T.; Rottapel, R. Transcriptional regulation of miR-31 by oncogenic KRAS mediates metastatic phenotypes by repressing RASA1. Mol. Cancer Res. 2016, 14, 267–277.
Rachagani, S.; Macha, M.A.; Heimann, N.; Seshacharyulu, P.; Haridas, D.; Chugh, S.; Batra, S.K. Clinical implications of miRNAs in the pathogenesis, diagnosis and therapy of pancreatic cancer. Adv. Drug Deliv. Rev. 2015, 81, 16–33.
Pupo, E.; Avanzato, D.; Middonti, E.; Bussolino, F.; Lanzetti, L. KRAS-driven metabolic rewiring reveals novel actionable targets in cancer. Front. Oncol. 2019, 9, 848.
Buscail, L.; Bournet, B.; Cordelier, P. Role of oncogenic KRAS in the diagnosis, prognosis and treatment of pancreatic cancer. Nat. Rev. Gastroenterol. Hepatol. 2020, 17, 153–168.