Pancreatic biomarkers: Could they be the answer?

Table 1 Suggested risk factors for the development of pancreatic ductal adenocarcinoma[3]

Hereditary syndromes	Non-hereditary risk factors
Hereditary breast/ovarian cancer (BRCA2, BRCA1, PALB2)	Nonhereditary chronic pancreatitis
Familial atypical multiple mole melanoma (FAMMM) syndrome (CDKN2A)	Diabetes mellitus, glucose metabolism, and insulin resistance
Peutz-Jeghers syndrome (STK11)	Cigarette smoking
Familial adenomatous polyposis (APC)	Obesity and physical inactivity
Hereditary nonpolyposis colon cancer (Lynch II) (DNA mismatch repair genes)	Diet (high intake of saturated fat and/or meat, particularly smoked or processed meats)
Familial pancreatic cancer (gene not identified)	Coffee and alcohol consumption
Hereditary pancreatitis (PRSS1, SPINK1)	Aspirin and nonsteroidal anti-inflammatory drug use
Ataxia telangiectasia (ATM)	History of partial gastrectomy or cholecystectomy
Li-Fraumeni syndrome (p53)	Helicobacter pylori infection

Table 2 Summary of the most important randomized clinical trials performed in advanced pancreatic ductal adenocarcinoma

Experimental arm treatment (number of patients included)	Median OS (mo) (Experimental arm) (95%CI)	Control arm treatment (number of patients included)	Median OS (mo) (Control arm) (95%CI)	Hazard ratio (95%CI) (P value)	Ref.
Gemcitabine (63 pts)	5.6 (data not shown)	5-FU (63 pts)	4.4 (data not shown)	Data not shown P = 0.0025	Burris et al[7], 1997
Gemcitabine and erlotinib (285 pts)	6.24 (data not shown)	Gemcitabine (284 pts)	5.91 (data not shown)	0.82 (0.69-0.99) P = 0.038	Moore et al[8], 2007
Gemcitabine and capecitabine (267 pts)	7.1 (6.2-7.8)	Gemcitabine (266 pts)	6.2 (5.5-7.2)	0.86 (0.72-1.02) P = 0.08	Cunningham et al[9] 2009
FOLFIRINOX (combination of 5FU, oxaliplatin and irinotecan) (171 pts)	11.1 (9.0-13.1)	Gemcitabine (171 pts)	6.8 (5.5-7.6)	0.57 (0.45-0.73) P < 0.001	Conroy et al[10], 2011
Gemcitabine and nab-paclitaxel (431 pts)	8.5 (7.9-9.5)	Gemcitabine (430 pts)	6.7 (6.0-7.2)	0.72 (0.62-0.83) P < 0.001	Von Hoff et al[11], 2013

OS: Overall survival.

Table 3 Core signalling pathways involved in pancreatic ductal adenocarcinoma

Involved pathways	PDA with pathway aberrations	Representative genes
Apoptosis	100%	CASP10, VCP, CAD, HIP1
DNA repair	83%	ERCC4, ERCC6, EP300, RANBP2, TP53
Regulation of G1/S phase	100%	CDKN2A, FBXW7, CHD1, APC2
Hedgehog pathway	100%	TBX5, SOX3, LRP2, GLI1, GLI3, BOC, BMPR2, CREBBP
Celular adhesion	79%	CDH1, CDH10, CDH2, CDH7, FAT, PCDH15, PCDH17, PCDH18, PCDH9, PCDHB16, PCDHB2, PCDHGA1, PCDHGA11, PCDHGC4
Integrin signaling	67%	ITGA4, ITGA9, ITGA11, LAMA1, LAMA4, LAMA5, FN1, ILK
c-Jun N-terminal kinase signaling	96%	MAP4K3, TNF, ATF2, NFATC3
KRAS signaling	100%	KRAS, MAP2K4, RASGRP3
Regulation of invasion	92%	ADAM11, ADAM12, ADAM19, ADAM5220, ADAMTS15, DPP6, MEP1A, PCSK6, APG4A, PRSS23
GTP-ase dependent signaling (not κ-ras)	79%	AGHGEF7, ARHGEF9, CDC42BPA, DEPDC2, PLCB3, PLCB4, RP1, PLXNB1, PRKCG
TGF-β pathway	100%	TGFBR2, BMPR2, SMAD4, SMAD3
Wnt/Notch pathway	100%	MYC, PPP2R3A, WNT9A, MAP2, TSC2, GATA6, TCF4

Addapted from Jones et al[19], 2008. PDA: Pancreatic ductal adenocarcinoma.

Table 4 Biomarkers in pancreatic ductal adenocarcinoma

Biomarker	Prognostic biomarker	Predictive biomarker	Comments and references
MUC1	Yes		Predictive of early cancer-related death[37]
MSLN	Yes		Predictive of early cancer-related death[37]
6-gene signature	Yes		Expression of FOSB, KLF6, NFKBIZ, ATP4A, GSG1 and SIGLEC11 is related with metastatic spread[38]
VEGF	Yes		Worse survival in resected PDA[39]
p16	Yes		Higher expression was related to poorer prognosis[40]
TP53	Yes		Relation with tumour dedifferentiation and higher locorregional recurrence[40]
SMAD4	Yes		Higher Smad4/Dpc4 was related to bigger tumours, lymph node metastases and shorter survival[40]. Higher relapse rate (distant spread)[41]. Loss of expression correlated with resectability and better survival after surgery[42]
EGFR			No predictive/prognostic power[43,44]
K-ras	Yes		Better prognosis in Kras wild-type tumours[43,44]
RRM1	Yes	Yes	High expression of RRM1 showed significantly better overall survival[45-47] and worse response to treatment[47-49]
ERCC1	Yes		High ERCC1 expression showed significantly better overall survival[47,50,51]. No predictive power[49]
CTCs	Yes		More studies are awaited[52]
hENT1	Yes	Yes	High expression of hNENt1: worse prognosis, higher response to gemcitabine in the adjuvant setting; unclear impact in metastatic patients[50,53-57]
HuR	Yes	Yes	Low expression of HuR: worse prognosis[58] and better response to gemcitabine[59,60]
SPARC	Yes		Expression of SPARC in the peritumoural stroma is related with worse prognosis[61,62]. No predictive effect
CTGF			Preclinical data seem to suggest prognostic impact and potential predictive power for FB-3019[63-66]

VEGF: Vascular endothelial growth factor; EGFR: Epidermal growth factor receptor; CTCs: Circulating tumour cells; SPARC: Secreted protein acidic and rich in cysteine; CTGF: Connective tissue growth factor.

Table 5 “Classic” predictive biomarkers for “classic” chemotherapies with potential interest in pancreatic cancer

Predictive biomarker	Drug	Theoretical impact1	Studies performed in pancreatic cancer (predictive outcome)?	Impact confirmed in pancreatic cancer?	Notes	Ref.
Thymidylate synthase	5FU	When negative, better response to 5FU	Yes	No	Predictive value in PDA not validated	[55,72-74]
DPD	5FU	When mutation DPD, more 5FU related toxicity	Yes	No	Survival benefit with S1 and DPD mutation	[73]
Topoisomerase I	Irinotecan	When positive, better response to Irinotecan	No	No	No data in pancreatic cancer	-
RRM1	Gemcitabine	When positive, better response to gemcitabine	Yes	Yes	Low expression correlates with better response	[47-49]
ERCC1	Oxaliplatin	When negative, better response to Oxaliplatin	Yes	No	No predictive effect	[49,51]
XRCC1	Oxaliplatin	When negative, better response to Oxaliplatin	No	No	No data in pancreatic cancer	-
EGFR/kras	Erlotinib	Erlotinib effective when EGFR mutation/kras wild type present	Yes	No	No predictive effect	[43,44]
PALB2	Mitomycin C	Mitomicin C effective when PALB2 mutation present	No	Yes	Case report	[75]
BRCA2	PARP inhibitors	PARP inhibitors effective when BRCA2 mutation present	Yes	Yes	Phase I trial	[76]

¹This impact is suggested in other malignancies.