Table 1.
Summarizing table of the most relevant features of the studies included in this meta-analysis.
| First Author of the Study, Year [21,22,23,24,25,26,27,28,29,30,31,32,33,34] | N. of Patients | Stage | Type of Tissue Specimen | Molecular Test for Tissue Specimen and Genes | Time Point of Tissue and Liquid Biopsy Test and Genes | Type of Liquid Biopsy | Molecular Test for Liquid Biopsy | TP | FP | TN | FN |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Ako, 2017 [21] | 40 | I–II: 60%, III–IV: 40% | 16 SR and 24 EUS-FNA | PCR, KRAS | The same time | Plasma and serum, cfDNA | Droplet PCR, KRAS | 19 | 0 | 3 | 18 |
| Bernard, 2019a [22] | 34 | I–II: 68%, III–IV: 32% | 22 SR and 12 EUS-FNA | PCR, KRAS | The same time | Blood for cfDNA | Droplet digital PCR, KRAS | 20 | 1 | 11 | 2 |
| Brychta, 2016 [23] | 50 | I–II: 82%, III–IV: 18% | SR | Chip-based digital PCR, KRAS | The same time | Plasma (cfDNA) | Chip-based digital PCR, KRAS | 13 | 0 | 14 | 23 |
| Earl, 2015 [24] | 12 | NA | SR | PCR, KRAS | The same time | Plasma (cfDNA) | Droplet digital PCR, KRAS | 3 | 3 | 2 | 4 |
| Kinugasa, 2015 [25] | 75 | I–II: 3%, III–IV: 97% | EUS-FNA | PCR, KRAS | The same time | Serum (cfDNA) | Droplet digital PCR, KRAS | 43 | 4 | 15 | 13 |
| Kulemann, 2016 [26] | 11 | I–II: 91%, III–IV: 9% | NS | PCR, KRAS | Retrospective | Blood with isolation and analysis of CTCs | PCR, KRAS | 5 | 0 | 0 | 6 |
| Marchese, 2006 [27] | 30 | I–II: 83%,III–IV: 17% | 25 SR, 5 EUS-FNA | rflp-PCR KRAS | The same time | Serum (cfDNA) | rflp-PCR KRAS | 0 | 0 | 9 | 21 |
| Park, 2018a [28] | 17 | I–II: 18%,III–IV: 82% | EUS-FNA | PCR, KRAS | The same time | Plasma (cfDNA) | PCR, KRAS | 10 | 0 | 4 | 3 |
| Pishvaian, 2017a,* [29] | 16 | I–II: 0%,III–IV: 100% | EUS-FNA of pancreas or metastasis | 321 genes panel NGS | During treatment | cfDNA | 68 genes panel NGS | 6 | 1 | 0 | 9 |
| Sefrioui, 2017 [30] | 27 | NS | EUS-FNA/biopsy/SR | Digital PCR, KRAS | The same time | Plasma (cfDNA) | Digital PCR, KRAS | 14 | 3 | 5 | 5 |
| Shibata, 1998 [31] | 3 | I–II: 66.6%, III–IV: 33.3% | NS | nPCR, KRAS | The same time | Peripheral blood (CTCs separation) | nPCR, KRAS | 3 | 0 | 0 | 0 |
| Vietsch, 2018a,* [32] | 5 | I–II: 100%, III–IV: 0% | SR | 56 genes panel NGS | LB before surgery | cfDNA | 56 genes panel NGS | 0 | 0 | 0 | 5 |
| Wu, 2014 [33] | 36 | NS | NS | COLD-PCR, KRAS | The same time | Plasma (cfDNA) | COLD-PCR, KRAS | 26 | 0 | 10 | 0 |
| Zill, 2015 [34] | 13 | NS | EUS-FNA | NGS | The same time | Plasma (cfDNA) | 54 genes panel NGS | 12 | 0 | 0 | 1 |
| Total | 369 | I–II: 57%, III–IV: 43% | 11 studies: same time, 3 studies: other times | 12 studies: cfDNA from blood, 2 studies: CTCs separation | 174 | 12 | 73 | 110 |
Abbreviation: TP: true positive, FP: false positive, TN: true negative, FN: false negative; SR: surgical resected specimen; EUS-FNA: endoscopic ultrasound-guided fine-needle aspiration; cfDNA: circulating-free DNA; CTCs: circulating tumor cells; NS: not specified; rflp-PCR: restriction fragment length polymorphism-polymerase chain reaction; NGS: targeted next-generation sequencing for specific pancreatic cancer genes; nPCR: nested PCR; LB: liquid biopsy; COLD-PCR: co-amplification-at-lower-denaturation-temperature polymerase chain reaction; Note: Bernard, 2019a refers to the analysis of exoDNA in liquid biopsy; Park, 2018a refers to the analysis using PCR; Pishvaian, 2017a refers to a cohort of patients in which cfDNA has been analyzed; Vietsch, 2018a refers to patients whose liquid biopsy has been analyzed before surgical resection of pancreatic tumor; * in these studies, the rate of concordance of mutations between tumor tissue and liquid biopsy has been shown taking into account the four most important genes in pancreatic cancer (KRAS, TP53, SMAD4, and CDKN2A).