Table 1 Studies on cell-free DNA methylation in plasma/serum as diagnostic and/or prognostic biomarkers for pancreatic adenocarcinoma

83 PDAC:

0

ppENK

p16 (9 patients): sensitivity 70%; specificity 100%

16 I-II

37 III

p16

30 IV

34 PDAC:

19 I-II

2 III

13 IV

30 HC

CCND2

Based on the unmethylated status of the composited biomarker: sensitivity 76%; specificity 59%

SOCS1

THBS

PLAU

VHL

30 PDAC

30HC

30CP

BRCA1²

CCND2^1,2

CDKN1C^1,2

CDKN2B¹

DAPK1¹

ESR1¹

MGMT¹

MLH1^1,2

MUC2¹

MYOD1¹

PGK1¹

PGR proximal^1,2

PGR distal²

RARB¹

RB1¹

SYK^1,2

¹Methylation of 14 gene promoters distinguishes between CP and PDAC: sensitivity 91.2% (95% CI 76.5-97.1); specificity 90.8% (95% CI 76.1-96.8)

²Methylation of 8 gene promoters distinguishes between HC and CP: sensitivity 81.7% (95% CI 67.3-90.6); specificity 78% (95% CI 63.8-87.7)

30 PDAC:

18 I-II

5 III

7 IV

30 HC

VHL

Combined 5 markers to differentiate PDAC from HC: sensitivity 80%; specificity 66%

GPC3: sensitivity 63.3%; specificity 83.3%

MYF3

TMS

GPC3

SRBC

16 PDAC:

1 I

8 III

7IV

29 HC

UCHL1

Higher methylation detection in PDAC compared to HC (p<0.05)

NPTX2

SARP2

13 CP

ppENK

Methylated p16 significantly higher in PDAC than in CP (p = 0.016)

p16

RASSF1A

104 PDAC:

60 CP

NPTX2

NPTX2: sensitivity 80%; specificity 76%

24 I-II

5 benign biliary tract stone disease

43 III

37 IV

47 PDAC

197: colon, lung, gastric, breast cancers and hepatocarcinoma

APC

Methylation frequencies:

DCC

p16

RASSF1A 34% APC 23.4%

p14

p16 17% p14 14.9%

RASSF1A

DCC 6.4%

42 PDAC:

10 I

32 II-IV

26 HC

BNC1

BNC1: sensitivity 79% (95% CI 66-91); specificity 89% (95% CI 76-100)

ADAMTS1: sensitivity 48% (95% CI 33-63); specificity 92% (95% CI 82-100)

ADAMTS1

BNC1+ADAMTS1: sensitivity 81% (95% CI 69-93); specificity 85% (95% CI 71-99)

90% sensitivity in stage I for both genes

95 PDAC:

40 I-II

13 III

42 IV

97 CP

APC

Diagnostic prediction model with 8 genes methylation panel that differentiate malign from benign conditions: AUC=0.86 (95% CI 0.81-0.91), sensitivity 76%; specificity 83%

Performance of prediction model in early-stages (I-II): AUC=0.86 (95% CI 0.79–0.92), 73% sensitivity; 83% specificity

BMP3

59 acute pancreatitis

BNC1

MESTv2

RASSF1A

27 benign conditions

SFRP1

SFRP2

TFPI2

95 PDAC:

40 I-II

13 III

42 IV

0

ALX4 ^1,2

Two methylation-based prognostic prediction models:

BNC1 ^1,2

CDKN2B ^1,2

HIC1 ¹

¹Methylation of 8 genes that differentiate stage IV from stage I-III disease: AUC=0.87, sensitivity 74%; specificity 87%

MLH1 ^1,2

NEUROG1 ^1,2

SEPT9v2 ^1,2

²Methylation of 8 genes that differentiate stage I-II from stage III-IV disease: AUC=0.82, sensitivity 73%; specificity 80%

SST ¹

TFPI2 ²

WNT5A ²

95 PDAC:

40 I-II

13 III

42 IV

0

BNC1

Gene hypermethylation based survival prediction model (Hazard ratios (95% CI)):

BNC1 2.00 (1.26-3.18); GSTP1 9.55 (2.70-33.82); SFRP1 1.94 (1.24-3.02); SFRP2 0.45 (0.27-0.73) and TFPI2 2.52 (1.42-4.47)

GSTP1

SFRP1

SFRP2

TFPI2

39 PDAC:

37 I-II

2 III-IV

95 HC

8 CP

BNC1

ADAMTS1

BCN1: AUC=0.79 (95% CI 0.70-0.85), sensitivity 64.1%; specificity 93.7%

ADAMTS1: AUC=0.91 (95% CI 0.85-0.95) sensitivity 87.2%; specificity 95.8%

BNC1 + ADAMTS1: AUC=0.95 (95% CI 0.90-0.98) sensitivity 97.4%; specificity 91.6%

57 PDAC

53 HC

BNC1

SEPT9

BCN1: sensitivity 50.9% (95% CI 37.3-64.4); specificity 88.7% (95% CI 77.0-95.7)

SEPT9: sensitivity 36.8% (95% CI 24.5-50.7); specificity 96.2% (95% CI 87.0-99.5)

BNC1 + SEPT9: sensitivity 64.9% (95% CI 55.0-78.8); specificity 86.8% (95% CI 74.7-94.5)

Combined genes + CA19-9: sensitivity 86% (95% CI 74.2-93.7); specificity 81.1% (95% CI 68.0-90.6)

14 PanIN

44 benign conditions

61 PDAC:

20 I-II

38 III-IV

2 unspecified

22 HC

UCHL1

PENK

NPTX2

SPARC

Methylation index (MI) of 4 genes higher in PDAC than in HC (p < 0.05)

Lower survival in patients with high MI for SPARC and NPTX2 genes (p < 0.05)

21 CP

47 PDAC:

2 II

41 III-IV

4 unknown

14 HC

ADAMTS2

HOXA1

PCDH10

SEMA5A

SPSB4

Methylation levels in the 5 genes not significantly different between PDAC and HC

49% of PDAC patients with at least one methylated gene, 49% sensitivity; 86% specificity

DNA methylation in ≥ 1 gene and/or KRAS mutation: sensitivity 68%; specificity 86%

4 PDAC:

2 II

2 III

2 HC

TRIM73

Combined 8 gene panel: sensitivity 97.1%; specificity 98%

FAM150A

EPB41L3

SIX3

MIR663

MAPT

LOC100128977

LOC100130148

30 PDAC:

15 nonmetastatic

15 metastatic

18 HC

SST

SST: sensitivity 93%; specificity 89%

67 PDAC:

8 I

26 II

17 III

16 IV

97 HC

5mC

5hmC

A 24-feature 5mC model that can discriminate PDAC from HC, sensitivity 82.4%; specificity 100%

A 27-feature 5hmC model that can discriminate PDAC from HC, sensitivity 85.7%; specificity 100%

The 51-feature model combining 5mC and 5hmC markers: sensitivity 93.8%; specificity 95.5%

22 PDAC:

10 HC

ADAMTS1

BNC1

LRFN5

PXDN

Pancreatic cancer detection with 4-gene panel: AUC=0.94, sensitivity 100%; specificity 90%

3 I

15 II

1 III-IV

3 not available

346 PDAC:

11 I

165 II

33 III

137 IV

25 CP

APC

Validation study of the diagnostic prediction model proposed in Henriksen et al., 2016: AUC=0.77 (95% CI 0.69-0.84)

Diagnostic prediction model + CA19-9 in:

Resectable disease (I-II): AUC=0.89 (95% CI 0.83-0.95)

Unresectable disease (IV): AUC=0.95 (95% CI 0.92-0.98)

Entire cohort: AUC=0.85 (95% CI 0.79-0.91)

BMP3

BNC1

MESTv2

RASSF1A

SFRP1

SFRP2

TFPI2

170 PDAC:

5 I

45 II

60 III

60 IV

170 HC

GRIN2D

Methylated DNA marker (MDM) panel: AUC=0.90 (95% CI 0.86-0.94)

MDM panel + CA 19-9: AUC=0.97 (95% CI 0.94-0.99), sensitivity 92% (95% CI 83-98); specificity 92% (95% CI 81-100)

MDM panel for early-stage detection: AUC=0.84 (95% CI 0.76-0.92)

MDM + CA19-9 for early-stage detection: AUC=0.90 (95% CI 0.84-0.97)

CD1D

ZNF781

FER1L4

RYR2

CLEC11A

AK055957

LRRC4

GH05J042948

HOXA1

PRKCB

SHISA9

NTRK3

25 PDAC:

1 I

7 II

4 III

13 IV

20 HC

ZNF154

ZNF154 for early stage (I-II): AUC=0.87, sensitivity 100% and specificity 80%

ZNF154 for late stage (III-IV): AUC=0.85, sensitivity 94.1% and specificity 80%

19 PDAC

19 IV

44 benign conditions

MIR129-2

LINC01158

CCDC181

PRKCB

TBR1

ZNF781

MARCH11

VWC2

SLC9A3

HOXA7

Biomarker set of 10 genes capable of distinguishing malignant from benign cases: AUC=0.999 (95% CI 0.995-1.0) sensitivity 100% and specificity 95%

Biomarker set useful for monitoring: methylation decrease after treatment (p=3.9x10^-3)

44 PDAC

44 IV

2 HC

BMP3

NPTX2

SFRP1

SPARC

TFPI2

NPTX2 methylation distinguished between low- and high-risk poor prognosis patients (p-=6.7x10^-3)

NPTX2 methylation dynamics during patients monitoring predict evolution disease and survival: AUC=0.80 (95% CI 0.66-0.94), sensitivity 85%; specificity 65%