Cancer Treatment Reviews 120 (2023) 102604 Contents lists available at ScienceDirect Cancer Treatment Reviews journal homepage: www.elsevier.com/locate/ctrv Systematic or Meta-analysis Studies Liquid biopsy after resection of pancreatic adenocarcinoma and its relation to oncological outcomes. Systematic review and meta-analysis☆ Laura Vidal a, b, Elizabeth Pando a, b, *, Laia Blanco a, Carles Fabregat-Franco c, Florian Castet c, Alexandre Sierra c, Teresa Macarulla c, Joaquim Balsells a, Ramón Charco a, Ana Vivancos d a Department of HPB and Transplant Surgery, Vall d’Hebron University Hospital, Barcelona, Spain b Universitat Autònoma de Barcelona, Barcelona, Spain c Gastrointestinal and Endocrine Tumour Unit, Vall d’Hebron Institute of Oncology (VHIO), Hospital Universitari Vall d’Hebron, Barcelona, Spain d Cancer Genomics Lab, Vall d’Hebron Institute of Oncology (VHIO), Barcelona, Spain A R T I C L E I N F O A B S T R A C T Keywords: Background: It has been hypothesised that manipulation during surgery releases tumoral components into cir­ ctDNA culation. We investigate the effect of surgery on plasma-borne DNA biomarkers and the oncological outcomes in Resectable pancreatic adenocarcinoma resectable pancreatic ductal adenocarcinoma (PDAC). We also compare non-touch isolation techniques (NTIT) Liquid biopsy with standard techniques. Pancreatic surgery Non-touch isolation technique Materials and methods: We performed a systematic review and a meta-analysis of studies analysing liquid biopsy Meta-analysis as circulating tumour DNA (ctDNA), circulating tumour cells (CTCs), and messenger RNA (mRNA) in resectable PDAC patients who underwent surgery and its association with overall survival (OS) and disease-free survival (DFS). Research in EMBASE, Web of Science and PubMed was performed. The ctDNA shift negative-to-positive (ctDNA − /+) or ctDNA shift positive-to-negative (ctDNA +/− ) before and after surgery was evaluated. Results: Twelve studies comprising 413 patients were included. Shorter OS and DFS were identified in patients with positive ctDNA status before (HR = 2.28, p = 0.005 and HR = 2.16, p = 0.006) or after surgery (HR = 3.88, p < 0.0001 and HR = 3.81, p = 0.03), respectively. Surgical resection increased the rate of ctDNA +/− . There were no differences in OS or DFS in the ctDNA +/− group compared with ctDNA +/+ or ctDNA − /+. However, there was a trend to shorter OS in the ctDNA − /+ group (HR = 5.00, p = 0.09). No differences between NTIT and standard techniques on liquid biopsy status were found. Conclusion: Positive ctDNA in the perioperative period is associated with a worse prognosis. Surgical resection has a role in the negativisation of liquid biopsy status. More studies are needed to assess the potential of minimally invasive techniques on ctDNA dynamics. Introduction an incomplete chemotherapy regimen, among others [8,9]. Addition­ ally, inaccurate staging and hidden micrometastasis during the preop­ Pancreatic ductal adenocarcinoma (PDAC) is the seventh cause of erative period are reasons for early recurrence. Metastatic capacity at cancer-related death worldwide and third place in Europe [1,2]. The the early stages of the disease, even in precancerous lesions (PanIN), has actual 5-year survival rate is 12 % [3]. This low survival rate is mainly also been described [10]. due to late diagnosis, aggressive biology and resistance to current Circulating tumour cells (CTCs) and circulating tumour (ct) nucleic chemotherapy regimens [4–7]. Only 20 % are resectable at diagnosis; acids, such as ctDNA and ctRNA, have been recognised in PDAC pa­ despite surgical resection, approximately 80 % recurred. Furthermore, tients’ blood, being released from the primary tumour and or metastatic several factors may contribute to the recurrence of PDAC after resection, site. The circulating capacity of PDAC cells is linked to the transition such as R1 status (<1mm), tumour size, lymph node involvement, and from epithelial to mesenchymal (EMT) phenotype, acquiring survival This publication has been made as part of the Doctoral Program in Surgery and Morphologic Sciences of the Universitat Autònoma de Barcelona, Spain. ☆ * Corresponding author at: Hospital Universitari Vall d’Hebron, Department of Hepato-Pancreato-Biliary and Transplant Surgery, Passeig de la Vall d’Hebron, 119- 129, 08035 Barcelona, Spain. E-mail address: elizabeth.pando@vallhebron.cat (E. Pando). https://doi.org/10.1016/j.ctrv.2023.102604 Received 15 June 2023; Received in revised form 2 August 2023; Accepted 4 August 2023 Available online 6 August 2023 0305-7372/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/). L. Vidal et al. Cancer Treatment Reviews 120 (2023) 102604 and self-renewal properties that confer cells the ability to metastasise. We defined the main question of the study according to the PICO Pancreatic cancer cells have also shown anoikis resistance, a defence cell framework (population, intervention/exposure, comparator and mechanism that induces apoptosis when the cell is detached and in the outcome) [32]. Studies were selected if they met the following eligibility absence of an extracellular matrix. Anoikis resistance confers higher criteria: migratory and invasive properties to pancreatic cancer cells [11,12]. The detection rate of CTCs in PDAC is relatively low; however, the ad­ ● (P) Study population included adults ≥ 18 years with potentially vances achieved in ctDNA have improved the detection rate of circu­ resectable pancreatic cancer undergoing surgery. lating tumoral components and have therefore become the most used ● (I) All patients had preoperative and postoperative liquid biopsy liquid biopsy technique[13]. The incidence of ctDNA detection in samples. pancreatic cancer is 26 %–62 %, being higher in advanced stages. For ● (C) Two analyses were performed: 1) Time-point liquid biopsy status resectable disease, rates up to 50 % have been reported [14–18]. (preoperative and postoperative) as positive versus negative and 2) Detectable ctDNA has been associated with shorter disease-free survival the shift occurrence after surgery in every individual as a shift to (DFS) and overall survival (OS) after pancreatic cancer resection positive versus shift to negative. [14,15,19]. ○ ctDNA − /+ was defined as a positivisation of liquid biopsy in the In that line, it has been reported that tumour manipulation during postoperative period in a patient with preoperatively negative surgery releases cancerous cells into blood circulation, showing an in­ liquid biopsy. crease in ctDNA and CTCs detection after surgery which could hypo­ ○ ctDNA +/− was defined as a negativisation of liquid biopsy in the thetically increase the risk of recurrence [15,20]. This phenomenon postoperative period in a patient with preoperatively positive could equally increase the risk of metastasis. Based on this observation, liquid biopsy. non-touch isolation techniques (NTIT) or vessel-first approaches have ● (O) Compared outcomes were OS, DFS, mortality rate and recurrence emerged as alternative techniques that could decrease the release of rate. cancerous cells into the bloodstream. Some authors reported decreases in CTCs levels using NTIT compared to the standard technique in colo­ A subanalysis also based in the PICO framework was performed in rectal cancer [21–27]. However, strong evidence is still needed to sup­ (P) potentially resectable pancreatic cancer patients (I) treated with a port the oncological superiority of NTIT against standard techniques, non-touch surgery (C) compared to the standard technique evaluating especially in pancreatic cancer surgery. (O) ctDNA − /+ and ctDNA +/− . All studies selected were available in Previous systematic reviews analysing ctDNA status and its impact English. on OS and DFS in PDAC included unresectable and metastatic stages [28–30]. Additionally, there is little evidence addressing the real effect Selection of reviews, extraction, and management of information of surgical resection on releasing tumoral components and its effect on DFS and OS in PDAC patients. All studies in this search were included in an Excel database, and We systematically reviewed circulating tumoral cell components repeated articles were excluded. Two reviewers (L.V.P and E.P.R) (CTCs, ctDNA, ctRNA) in patients who underwent pancreatic cancer separately assessed publication titles, abstracts, and full-text selection resection. We aimed to investigate if surgical resection impacts tumoral based on eligibility criteria. If disagreement was found in a study, it was component release by favouring shift dynamics in liquid biopsy status solved in consensus between reviewers J.B.V. and A.V.P. Some refer­ and if the type of surgical technique plays a role. Besides, we intend to ences were searched and selected manually from the reference list of demonstrate the association of liquid biopsy status with survival and previously included studies. First author’s name, year and country of recurrence in resectable pancreatic cancer patients. publication, the number of included subjects, TNM, type of surgery, margin status, neoadjuvant chemotherapy, origin and time of sample, Methods liquid biopsy detection method, marker, tissue sample, follow-up time and outcomes were extracted by each reviewer independently. Out­ This systematic review has been done according to Preferred comes were collected as Hazard Ratio (HR) with a corresponding 95 % Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) confidence interval or as Risk Ratio (RR) or Odds Ratio (OR). [31]. In some studies that included metastatic stages, locally advanced For this study, different types of liquid biopsies were analysed, such non-resectable and resectable pancreatic cancer patients, only the as ctDNA, CTCs and ctRNA. potentially resectable pancreatic cancer patients that underwent surgery were considered for this meta-analysis. If this stratification was not possible, the article was excluded from the meta-analysis. Study selection and eligibility criteria Information was obtained from articles’ tables, graphics and texts. Supplementary tables and graphics were also used. If data was not A search in EMBASE, Web of Science and PubMed was performed available, the IP of each study was contacted. between November and December 2021. We used the following Medical Subject Heading (MeSH) keywords: Risk of bias ● “Pancreatic cancer” AND “tumour manipulation” OR “Non-Touch” OR The quality and risk of bias of all studies included in the meta- “circulating DNA” OR “circulating tumour DNA” OR “ctDNA” OR analysis were assessed independently by L.V.P and E.P.R. using the “ctRNA” OR “Circulating tumour cells” OR “CTC”. Quality In Prognosis Studies (QUIPS) tool [33], and inconsistencies were ● “Pancreatic adenocarcinoma” AND “tumour manipulation” OR “Non- solved by a third reviewer, J.B.V and A.V.P. For each quantitative Touch” OR “circulating DNA” OR “circulating tumour DNA” OR analysis, publication bias was assessed using a Funnel plot. “ctDNA” OR “ctRNA” OR “Circulating tumour cells” OR “CTC”. ● “Pancreatic tumour” AND “tumour manipulation” OR “Non-Touch” Data analysis OR “circulating DNA” OR “circulating tumour DNA” OR “ctDNA” OR “ctRNA” OR “Circulating tumour cells” OR “CTC”. This meta-analysis data was analysed by one reviewer L.V.P and ● “Pancreatic cancer surgery” AND “tumour manipulation” OR “Non- supervised by E.P.R. The compared groups have been established ac­ Touch” OR “circulating DNA” OR “circulating tumour DNA” OR cording to positive and negative liquid biopsy status in each study. We “ctDNA” OR “ctRNA” OR “Circulating tumour cells” OR “CTC”. set a subject as positive or negative according to the detectable 2 L. Vidal et al. Cancer Treatment Reviews 120 (2023) 102604 concentration threshold of each molecular test defined in each study plot and Funnel Plot were obtained. methodology. Liquid biopsies were performed at the following time points: pre­ Results operative and postoperative. Scenarios were evaluated separately, comparing mortality, OS, recurrence and DFS in positive and negative Study selection liquid biopsy groups. Additionally, we analysed if negativisation (ctDNA +/− ) or positivisation (ctDNA − /+) occurred after surgery and assessed Following a systematic search, 4157 studies were retrieved, of which its association with survival. The secondary sub-analysis compared 818 duplicates were removed. The titles and abstracts of the remaining liquid biopsy status preoperative and postoperative between NTIT and 3339 were assessed, excluding 3261 because they included animal standard technique in pancreatic cancer surgery. subjects, other biliopancreatic tumours, non-original articles, absence of OS, DFS, mortality, and recurrence rates were analysed and exported surgical treatment, or missing blood samples from before or after sur­ directly from tables as HR with 95 % CI or RR/OR with 95 % CI. If RR gery (Fig. 1). and OR were not available, it was determined from individual subject A total of 78 articles were entirely read, and 67 were excluded due to information from supplementary data. If HR was unavailable, we the absence of preoperative or postoperative samples or due to the in­ extracted individual subject information from the supplementary data clusion of metastatic patients. One additional study which was extracted and calculated it by using a univariate Cox regression analysis on IBM® from the references list before the quantitative data analysis was SPSS ® 27. included. Finally, 12 articles were included in the meta-analysis, All quantitative analyses were performed on RevMan 5.4.1 ®. We including 413 patients with potentially resectable PDAC. The study in­ used the Der-Simonian and Laird random effects model to estimate the clusion period was from 1998 to 2021. pooled effect size in our meta-analysis [34]. The chi-squared heteroge­ Patients were included in the quantitative analysis according to the neity test was used to assess the statistical heterogeneity between the intention-to-treat principle. Of note, 17 (4,1 %) patients were finally trials, and the I2 statistic was used to assess the proportion of variability classified as stage IV due to the positivity of distant lymph nodes after attributable to between-study heterogeneity. For each analysis, a Forest the analysis of resected specimens, false negative CT or unexpected Fig. 1. Flow diagram of the systematic review Cochrane [35]. 3 L. Vidal et al. Cancer Treatment Reviews 120 (2023) 102604 findings during surgery. A consensual decision was made to include according to the Funnel Plot. (Supplementary material, Fig. S9). The these stage IV patients in the meta-analysis to resemble a real clinical analysis showed a decrease in OS in patients with ctDNA+ before sur­ scenario and to perform an intention-to-treat analysis. gery compared with those with ctDNA - (HR = 2.28, CI95 %:1.28–4.08, p = 0.005, Fig. 3B). Study characteristics Six studies analysed DFS according to preoperative ctDNA liquid biopsy status. There was no publication bias (Supplementary material, Table 1 summarises the characteristics of the studies included. Most Fig. S10). Patients with ctDNA+ before surgery had decreased DFS studies were prospective, except for one retrospective [39] and one compared to patients with ctDNA - (HR = 2.16, CI95 %: 1.25–3.72, p = randomised controlled trial [26]. The studies were performed in Japan 0.006, Fig. 3C). [27,38,40–44], China [37,45], England [26], the USA [36] and Australia [39]. The AJCC TNM classification system was the most used, Survival analysis according to liquid biopsy status after surgery followed by the Japanese Pancreas Society Classification; two articles OS according to postoperative ctDNA status was analysed in 8 did not mention the classification used [37,39]. Two publications studies. There was no publication bias, according to Funnel Plot (Sup­ compared different techniques, NTIT versus standard technique [26,27]. plementary material, Fig. S11). The analysis showed a decrease in OS in In all studies, blood samples were taken preoperatively, post­ patients with ctDNA+ after surgery compared to patients with ctDNA - operatively, and in some cases, during surgery. Time of liquid biopsy is (HR = 3.88, CI95 %:2.02–7.46, p < 0.0001, Fig. 3D). detailed in Table 1. Preoperative samples are mostly taken during the 24 Six studies analyse DFS according to postoperative ctDNA status. h before surgery or prior tumour manipulation. Postoperative samples There was no publication bias (Supplementary material, Fig. S12). Pa­ are mainly collected between specimen removal and the 10th post­ tients with ctDNA+ after surgeryhad decreased DFS compared to pa­ operative day. The majority of the studies used Ethylenediaminetetra­ tients with ctDNA - (HR = 3.81, CI95 %:2.03–7.14, p = 0.03, Fig. 3E). acetic (EDTA) preservation tubes (Supplementary material, Table S1). The most widely used liquid biopsy was ctDNA based on KRAS mutation Survival analysis according to the shift dynamics after surgery detection in plasma. Some authors matched tumour mutation in pe­ ripheral blood and tissue [37,39,41–45]. CEA mRNA and CTCs in whole Shift to negative analysis (ctDNA +/− ). Three studies compare OS and blood were used as liquid biopsies by Hirota et al. and Gall et al. and were DFS between patients with ctDNA +/− versus those patients ctDNA +/+ the two studies included in the sub-analysis comparing NTIT techniques. and ctDNA − /+. No publication bias was found in the Funnel plot. Overall survival was the commonest endpoint among the studies. (Supplementary material, Fig. S13). The analyses revealed no differ­ ences in OS between groups (HR = 1.39, CI95 %:0.54––3.54, p = 0.50, Risk of bias Fig. 3F). For DFS analysis no publication bias was found (Supplementary The results of the risk of bias assessment are shown in Fig. 2 (Sup­ material, Fig. S14). The analyses revealed no differences in DFS between plementary material, Fig. S1). groups (HR = 1.62, CI95 %:0.50–5.23, p = 0.42, Fig. 3G). Mortality and recurrence risk analysis according to liquid biopsy Shift to positive analysis (ctDNA − /+). Three studies evaluate OS be­ status after surgery tween patients with ctDNA − /+ versus those patients ctDNA − /− and ctDNA +/− . No publication bias was found (Supplementary material, The mortality risk was analysed in eight studies, including 328 pa­ Fig. S15). The analyses revealed no differences in OS between groups. tients. The analysis showed a trend of an increased risk for death among Although statistical significance was not reached, a trend to a decreased patients with ctDNA+ after surgery (RR = 2.67, CI95 %:0.88–8.08, p = OS in patients ctDNA − /+ versus ctDNA − /− and ctDNA +/− was found 0.08), although it did not achieve statistical significance (Supplemen­ (HR of 5.0, CI95 %:0.79––31.75, p = 0.09, Fig. 3H). tary material, Figs. S2–S3). Two studies evaluate DFS in patients with ctDNA − /+ versus those Sensitivity assessment by repeated quantitative analysis identified patients ctDNA − /− and ctDNA +/− . No publication bias was found. No that the Yamaguchi et al. study increased heterogeneity. An extended differences in DFS between groups were found in this analysis (HR = follow-up period (2–132 months) yielded a higher incidence of events in 1.25, CI95 %:0.73–2.12, p = 0.42, Supplementary material, all groups.Publication bias was confirmed in the Funnel Plot. When Figs. S16–S17). excluding this publication, statistically significant results were achieved (RR = 2.67, CI95 %:1.90–3.76, p < 0.00001) (Supplementary material, Comparison between non-touch techniques (NTIT) versus standard Figs. S4–S5). technique The recurrence rate was analysed in 328 patients from 8 studies. There was no publication bias. Recurrence was higher in patients with A subanalysis was performed to investigate if NTIT were associated ctDNA+ after surgery (RR = 1.42, CI95 %:1.08–1.87, p = 0.01, Sup­ with liquid biopsy negativisation after surgery. Liquid biopsies used plementary material, Figs. S6–S7). were CTC and CEA mRNA. The small number of patients and studies included confers a bias. No differences in negativisation rates after Effect of surgery ctDNA on liquid biopsy dynamics surgery between techniques were found (HR = 1.69, CI95 %0.25–11.28, p = 0.59, Supplementary material, Figs. S18–S19). Seven studies, including 282 patients, analysed the shift status after surgery, ctDNA +/− and ctDNA − /+. There was no publication bias in the Funnel Plot. (Supplementary material, Fig. S8).This comparison Sensitivity analysis revealed a higher rate of patients with ctDNA +/− (RR of 0.34, CI95 %:0.17–0.70, p = 0.003, Fig. 3A). A sub-analysis excluding patients with stage IV diagnosed intra­ operatively, we did not obtain any changes in the primary outcomes Survival analysis analysed concerning liquid biopsy status (Supplementary material, Figs. S20–S24). Survival analysis according to liquid biopsy status before surgery In a sub-analysis after excluding studies from the nineties, no dif­ Seven studies were included for OS analysis. CtDNA was the ferences were found in the primary outcomes (Supplementary material, biomarker reported in all cases. There was no publication bias, Figs. S25–S29). 4 L. Vidal et al. Table 1 General characteristics of the included articles. Study Year Country Sample Stage Type of Margin Neoadjuvant Sample Time of Time of sample Detection Technology/Assay Marker Targets Endpoint Follow up size TNM surgery status chemotherapy sample pre post method (months) (R0) Gall et al. 2014 England 12 II (12) sPD (6), 3/6 (50 0/12(0 %) Blood Before tumor After specimen Fluorescence Cell Search System CTCs counts OS, DFS 14.6 [26] nPD (6) %), 3/6 manipulation removal microscopy & (Veridex) (10.5–27.9) (50 %) scanning Groot et al. 2019 USA 59 I-II PD (39), 46/59 24/59 (40.7 %) Plasma Before Prior to Digital PCR RainDrop Digital ctDNA KRASm (G12V, OS, DFS 16 (13–19) [36] (43); III DP (13), (78 %) incision discharge PCR system G12D, G12R, (16) TP (7) (RainDance Q61H) Technologies) Hirota et al. 2005 Japan 18 I (2), II sDP (4), Not 0/18 (0 %) Blood Before tumor After specimen Real-time RT- LightCycler (Roche ctRNA CEA OS, DFS 42 [27] (1), III sPD (6), reported (PV) manipulation removal PCR Biochemicals) (7), IV nDP (3), (8) nPD (5) Jiang et al. 2020 China 27 I (13), Not Not Not reported Plasma Before POD 7 NGS Custom hybrid- ctDNA 1017 cancer DFS 18.6 [37] II (9), defined reported incision capture panel susceptibility (12.4–28.9) IV (5) genes Kitahata 2021 Japan 27 I (3), II- PD (22), 26/27 27/27 (100 %) Plasma Pre NAC - Post 4–8 weeks PO Digital PCR QX200 Droplet ctDNA KRAS multiplex OS, DFS 14.5 et al. [38] III (24) DP (4), (96.3 %) NAC Digital PCR system assays (G12A, (9.5––38.4) TP (1) (BioRad) G12C, G12D, G12R, G12S, G12V, G13D) Lee et al. 2019 Australia, 35 II (3), Not 16/35 0/35 (0 %) Plasma Diagnosis 4–8 weeks PO NGS SafeSeqS (Illumina) ctDNA KRAS (G12, G13, OS, DFS 38.4 [39] New Zeland, III (32) defined (45.7 %) G61) Singapore 5 Nakano et al. 2018 Japan 45 I (2), II PD (25), 36/45 11/45 (24.4 %) Serum Before Discharge Real-time PCR Peptide nucleic acid ctDNA KRAS (codon 12 DFS, OS 43 [40] (8), IIB DP (16), (80 %) incision (PNA)-directed PCR and 13) (35) TP (4) clamping Nomoto 1998 Japan 10 I (1), II Not Not 0/10 (0 %) Blood Before Intraoperative PCR/RFLP & Amplification KRAS ctDNA KRAS (codon 12) OS 19 et al. [41] (4), III defined reported incision and POD 1–7 Sanger mutant and WT and (4), IV sequencing selective enzym (1) digestion WT Watanabe 2019 Japan 39 I (3) II DP (19), Not 7/39 (18 %) Plasma Before 1–12 weeks PO Digital PCR QX200 Droplet ctDNA KRAS (G12V, OS 16.2 et al. [42] (34) III PD (15), reported incision Digital PCR system G12D, G12R, (1), IV TP (5) (BioRad) Q61H) (1) Yamada 1998 Japan 9 I (2), II PD (5), 5/9 (55.6 0/9 (0 %) Plasma Before 4–8 weeks PO Real-time PCR Mutant allele- ctDNA KRAS (1st & 2nd OS 12.2 et al. [43] (3), III DP (4) %) incision specific nucleotide in (1.3–43.5) (2), IV amplification codon 12) (2) (MASA-PCR) Cancer Treatment Reviews 120 (2023) 102604 Yamaguchi 2021 Japan 97 I (4), II PD (97) 71/97 30/97 (30.9 %) Plasma Before POD 3 Digital PCR QX200 Droplet ctDNA KRAS (G12V, OS, DFS 29 (2–132) et al. [44] (93) (73.2 %) incision Digital PCR system G12R, G12D) (BioRad) Yang et al. 2018 China 35 I (3), II PD (35) 23/35 0/35 (0 %) Plasma Day before POD 10 Digital PCR QuantStudio™ 3D ctDNA KRAS (G12V, PFS, OS 12.4 [45] (29), III (65.7 %) surgery Digital PCR System G12D, G12R) (6.1–17.2) (3) (Thermo Fisher Scientific) TOTAL 413 *Abbreviations: DFS, disease-free survival; OS, overall survival; PFS, progression-free survival; PD, pancreatoduodenectomy; DP, distal pancreatectomy; TP, total pancreatectomy; sDP, standard distal pancreatectomy; sPD, standard pancreatoduodenectomy; nDP, non touch distal pancreatectomy; nPD, non touch pancreatoduodenectomy; NTIT, non touch isolation technique; RT, reverse transcriptasa; PV, portal vein; PA, peripherial artery; NGS, next generation sequencing; SVC, superior vena cava; NAC, neoadjuvancy. POD: postoperative day; PO: postoperative. L. Vidal et al. Cancer Treatment Reviews 120 (2023) 102604 Role of surgical resection on liquid biopsy status This systematic review answers the hypothesis that surgical resection has a role in circulating tumoral component status. Despite our initial hypothesis of increased release of tumour components to blood flow after surgical manipulation, our results demonstrate ctDNA negativisa­ tion after surgery (HR = 0.34, p = 0.003). A potential explanation for these results is that most studies in our meta-analysis reported post­ operative ctDNA determination longer than 24 h after surgical specimen removal and before the fourth week after surgery [26,27,36,37,40,41,44,45]. The half-life of ctDNA has a wide variability from minutes to hours. Consequently, the later the ctDNA is determined after surgery, the more probability of finding a negative ctDNA. Addi­ tionally, these results corroborate that a blood negativisation is expected when the primary source of ctDNA is removed (the primary tumour). Cell free DNA (cfDNA) levels increase and persist for up to two to four weeks following surgery, possibly masking persistent ctDNA in relapse patients, this phenomenon could be explained by the concept of trauma-induced cell free DNA (cfDNA) [20,50]. To decrease the likeli­ hood of false negatives, performing the liquid biopsy between a window period of 2–4 weeks after surgery has to be considered [50]. Interestingly, studies focused on NTIT which performed ctDNA de­ terminations during and immediately after specimen removal found positivisation of ctDNA, in agreement with the short half-life of ctDNA. Nomoto et al. reported a progressive positivisation of liquid biopsy during surgical manipulation in up to 100 % of cases after tumour removal [41]. Nonetheless, it is uncertain whether tumoral cells released immediately after tumour manipulation could resist anoikis and lately develop micrometastatic disease and recurrence weeks after surgery. On the other hand, in patients where ctDNA remains positive after surgical resection, a tumour source that continues releasing ctDNA, such as a missing micrometastasis or a lack of local control after surgery (positive lymph nodes, neural infiltration), could be hypothesised. The contribution of surgery to increase ctDNA concentrations in cases with permanently positive ctDNA remains a matter to clarify in future studies. Impact of liquid biopsy status on survival Resectable PDAC patients with ctDNA -/V were associated with longer OS (HR = 2.28) and DFS (HR = 2.16) compared to patients with Fig. 2. Individual assessment of the risk of bias for each article with the QUIPS tool. ctDNA +/V. Likewise, the time point liquid biopsy status revealed that ctDNA V/+negatively impacted OS (HR = 3.88) and DFS (HR = 3.81). It is essential to analyse the dynamics shift during the intervention in Discussion each individual instead of the liquid biopsy status at a given time point in all the cohorts. When we analysed shift after surgical resection, we To our knowledge, this is the first meta-analysis in the literature that found that surgery increased negativisation of liquid biopsy status, focuses on studying the interplay between surgical resection in PDAC eventually influencing OS and DFS. patients and liquid biopsy status, including only resected patients. Our However, this meta-analysis could not demonstrate the association study demonstrates that surgical resection has a role in ctDNA neg­ between ctDNA +/− and improved OS and DFS. One of the explanations ativisation in PDAC resectable patients (HR = 0.34, p = 0.003). Never­ could be the small number of studies providing exact information on theless, regarding OS and DFS in patients with ctDNA +/− after surgery, liquid biopsy change of status in each individual. Additionally, sensi­ no impact was found in our analysis. However, a tendency to a decreased bility of ctDNA in post-operative PDAC may be low, also the outcome of OS was observed in patients with ctDNA − /+. Patients with ctDNA − /+ PDAC is overall so dismal that it may be complicated to find the cured had worse OS (HR = 3.88, p < 0.0001, Fig. 3D) and DFS (HR = 3.81, p = population. Further explanations are that other factors such as fragility, 0.03, Fig. 3E). advanced age, postoperative complications and adjuvant chemotherapy Biomarkers, more precisely ctDNA, are a revolutionary tool used in could influence DFS and other rescue treatments after recurrence could the diagnostic, prognostic, and guiding treatment of malignant diseases impact OS. In that line, it has been demonstrated that certain adjuvant [46–49]. Over the last few years, publications about ctDNA and chemotherapies, such as FOLFIRINOX, yield more favourable OS and pancreatic cancer have increased substantially. Milin-Lazovic et al. DFS in resectable PDAC patients when compared to Gemcitabine [51]. concluded in their meta-analysis that a positive liquid biopsy signifi­ This meta-analysis did not allow for stratification based on adjuvant cantly negatively impacts OS and progression-free survival (PFS) in therapy. PDAC [29]. This association was also found by Bounduc et al. [30]. Both Regarding NTIT, this meta-analysis failed to prove less release of meta-analyses included different tumour stages, including metastatic tumoral components after surgical manipulation, nor better OS and DFS patients, chemotherapy and surgery treatments with curative and in NTIT compared to the standard technique. These results may be palliative intention in the quantitative analysis. explained due to; the low number of studies published comparing NTIT 6 L. Vidal et al. Cancer Treatment Reviews 120 (2023) 102604 Fig. 3. A, Forest plot of articles analysed comparing ctDNA shift after surgery in patients with resectable PDAC. B, Forest plot of OS comparing positive versus negative ctDNA status before surgery in patients with resectable PDAC. C,Forest plot of DFS comparing positive versus negative ctDNA status before surgery in patients with resectable PDAC. D, Forest plot of OS comparing positive versus negative ctDNA status after surgery in patients with resectable PDAC. E, Forest plot of DFS comparing positive versus negative ctDNA status after surgery in patients with resectable PDAC. F, Forest plot of OS in patients with ctDNA +/− versus those patients with ctDNA +/+ and ctDNA − /+. G, Forest plot of DFS in patients with ctDNA +/− versus those patients with ctDNA +/+ and ctDNA − /+. H, Forest plot of OS between patients with ctDNA − /+ versus those patients with ctDNA − /− and ctDNA +/− . versus standard technique in PDAC, the reduced number of patients Limitations and strengths included (total n = 30) and the heterogeneity of results presentation. Hirota et al. described decreased detection of CEA-mRNA and low This meta-analysis has certain limitations. As an intention-to-treat recurrence rate after NTIT compared to the standard technique [27]. analysis, we included a small number of patients (4.1 %) re-staged as NTIT includes vessels and lymphatic first approach by ligating the gas­ metastatic disease intraoperatively. However, a sub-analysis excluded trocolic trunk at its communication with the superior mesenteric vein, those patients without obtaining changes in the primary outcomes followed by ligation of the gastroduodenal artery, right gastric artery, analysed concerning liquid biopsy status (Supplementary material, inferior pancreatic arteries, all pancreaticoduodenal branches from the Figs. S30–S34). portal vein and ligation and section of the bile duct and lymphatic The sensitivity and specificity of liquid biopsy are also a limitation vessels before manipulation of tumour. Gall et al. have demonstrated a since ctDNA is only detected in 50 % of resectable early-stage patients significantly reduced number of CTCs in the portal vein in NTIT [28–32]. Some mutations are not detected in standardised diagnostic compared to the standard technique but failed to achieve survival dif­ panels, that mainly include KRAS (G12, G13, G61), TP53, and SMAD4. ferences [26]. In 7 of the 12 studies (225 patients), a matching tissue strategy between Fortunately, the mere presence of tumoral components in the blood primary tumour and liquid biopsy was performed in order to increase circulation is insufficient to guarantee metastasis development. Meta­ ctDNA detection. In addition, the fact that different detection methods static implantation is a complex phenomenon which implies many were used, such as ddPCR, NGS or qPCR, could influence detection rate. biological pathways. Tumoral cells from the primary tumour or pre­ Reports included in the meta-analysis only determine patients as posi­ malignant lesions can create a pre-metastatic niche (PMN) in the liver tive or negative considering the threshold concentration of ctDNA of before even starting dissemination. They secrete soluble factors and each test without reporting the exact ctDNA concentration (copies/ml3), extracellular vesicles that enhance vascular permeability and docking, being not possible to determine tumour burden. change the extracellular matrix, and gather immunosuppressive in­ Even though precise ctDNA mutant allelic fraction is not specified in flammatory cells into the liver, conferring a supportive niche to the the articles, the limit of detection (LoD) of the tests utilized is between cancerous cells [52,53]. As previously mentioned, these cells need 1/1000–1/10,000 (mutant templates/normal templates). In general, anoikis resistance and the capacity to transition from epithelial to ctDNA reported fractions in the literature are >1/1000, since tests are mesenchymal phenotype to travel through the blood circulation and also limited by the genome equivalents present in plasma samples. attach to the PMN or travel in groups of cells [10–12]. Some of the Hence, the results obtained in the different studies are comparable fundamentals of neoadjuvant therapy are based on those pathways [53]. (Supplementary material, Table S1). Nevertheless, our results showed that patients with positive tumoral Although different preserving tubes were used, no differences are components after surgery have worse DFS and OS, suggesting that the expected when samples are processed within 6 h in EDTA tubes. Cell­ above-mentioned pathways occur. Save has demonstrated to stabilize ctDNA and cfDNA better than EDTA tube after 48 h, but none of the included studies delay sample processing more than 6 h (Supplementary material, Table S1) [55,56]. 7 L. Vidal et al. Cancer Treatment Reviews 120 (2023) 102604 Additionally, this meta-analysis could not assess the role of neo­ meticulous review of this article. adjuvant treatment, and the radicality status effect over liquid biopsy They would also like to express their gratitude to patients and au­ status. In our meta-analysis, 99 patients (24 %) received neoadjuvant thors of all the studies included in the meta-analysis, whose collabora­ treatment, all included in publications from 2018 forward. tion made this study possible. The publication year range of this meta-analysis goes from 1998 to Authors would like to thank Universitat Autònoma de Barcelona for 2021. Significant improvements have been made during these last de­ Open Access funding. cades, including better diagnosis accuracy, TNM classification modifi­ cations, more effective chemotherapies, enhanced surgical techniques Author contribution and increased knowledge of liquid biopsy [54,57,58]. Regardless, the meta-analysis includes eight studies published in the last five years that LVP: project development, data collection and analysis, manuscript use similar KRAS mutation (ctDNA) detection tests. When those studies writing and editing. EPR project development, data collection and are analysed separately no significant differences are found (Supple­ analysis, manuscript writing and editing. JBV and AVP contribute to mentary material, Figs. S30–S34). solving data disagreements. AVP, TMM, JBV and RCH contributed to the This meta-analysis is focused exclusively on potentially resectable final version of the manuscript. All authors reviewed the manuscript. patients, therefore a likely curable population, trying to understand the role of surgical resection on liquid biopsy status. Our results are Funding encouraging; thus, a role of negativisation of the ctDNA is related to surgery. These results may open new research lines to optimise the Open Access funding provided by Universitat Autònoma de reduction of tumoral components after resection using new promising Barcelona. surgical techniques and molecular biology advances. Minimally invasive surgical procedures allow accurate and minimal manipulation of tu­ Appendix A. Supplementary material mours during surgery, aligning with this proposal [59]. This report also corroborates biomarkers, more precisely ctDNA, as a Supplementary data to this article can be found online at https://doi. reliable prognostic factor in resectable PDAC patients. According to this, org/10.1016/j.ctrv.2023.102604. ctDNA may have a role in PDAC follow-up and could be used to track down patients with higher relapse probabilities after surgery. Our re­ References sults encourage that ctDNA may have the potential to determine which patients with resectable PDAC may benefit from neoadjuvant chemo­ [1] Ferlay J, Colombet M, Soerjomataram I, Mathers C, Parkin DM, Piñeros M, et al. therapy. Despite our efforts, we have identified some publication bias, Estimating the global cancer incidence and mortality in 2018: GLOBOCAN sources and methods. Int J Cancer 2019;144(8):1941–53. https://doi.org/10.1002/ suggesting that our findings should be interpreted cautiously. ijc.31937. Future efforts must determine if minimally invasive surgery could [2] Carioli G, Malvezzi M, Bertuccio P, Boffetta P, Levi F, La Vecchia C, et al. European contribute to a greater negativisation of ctDNA after surgery in resect­ cancer mortality predictions for the year 2021 with focus on pancreatic and female lung cancer. Ann Oncol 2021;32(4):478–87. https://doi.org/10.1016/j. able PDAC patients conferring better oncological outcomes. annonc.2021.01.006. [3] American Cancer Society. Cancer facts & figures 2023. Atlanta: American Cancer Conclusion Society; 2023. [4] Gresham GK, Wells GA, Gill S, Cameron C, Jonker DJ. Chemotherapy regimens for advanced pancreatic cancer: a systematic review and network meta-analysis. BMC In resectable PDAC patients, surgical resection has a role in the Cancer 2014;14:471. https://doi.org/10.1186/1471-2407-14-471. Published 2014 negativisation of liquid biopsy status. Detectable ctDNA is a prognostic Jun 27. [5] Hurton SMS, Hayden JA, Molinari M. Adjuvant therapy for resected pancreatic factor for surveillance and relapse in resectable PDAC patients. NTIT did cancer. Cochrane Database Syst Rev 2017;3:CD011514. https://doi.org/10.1002/ not demonstrate increased liquid biopsy negativisation compared to the 14651858.CD011514.pub2. Published 2017 Mar 10. standard technique, although more studies are needed to obtain more [6] Nitecki SS, Sarr MG, Colby TV, van Heerden JA. Long-term survival after resection for ductal adenocarcinoma of the pancreas. Is it really improving? Ann Surg 1995; robust conclusions. 221(1):59–66. https://doi.org/10.1097/00000658-199501000-00007. [7] Richter A, Niedergethmann M, Sturm JW, Lorenz D, Post S, Trede M. Long-term CRediT authorship contribution statement results of partial pancreaticoduodenectomy for ductal adenocarcinoma of the pancreatic head: 25-year experience. World J Surg 2003;27(3):324–9. https://doi. org/10.1007/s00268-002-6659-z. Laura Vidal: Conceptualization, Methodology, Formal analysis, [8] Chandrasegaram MD, Goldstein D, Simes J, et al. Meta-analysis of radical resection Investigation, Resources, Writing – original draft, Visualization, Funding rates and margin assessment in pancreatic cancer. Br J Surg 2015;102(12): acquisition. Elizabeth Pando: Conceptualization, Methodology, Project 1459–72. doi: 10.1002/bjs.9892. [9] Strobel O, Hank T, Hinz U, Bergmann F, Schneider L, Springfeld C, et al. Pancreatic administration, Investigation, Writing – review & editing, Supervision, cancer surgery: the new R-status counts. Ann Surg 2017;265(3):565–73. https:// Funding acquisition. Laia Blanco: Writing – review & editing. Carles doi.org/10.1097/SLA.0000000000001731. Fabregat-Franco: Writing – review & editing. Florian Castet: Writing – [10] Wang S, Huang S, Sun YL. Epithelial-mesenchymal transition in pancreatic cancer: a review. Biomed Res Int 2017;2017:2646148. https://doi.org/10.1155/2017/ review & editing. Alexandre Sierra: Writing – review & editing. Teresa 2646148 [10]. Macarulla: Writing – review & editing. Joaquim Balsells: Supervision, [11] Fofaria NM, Srivastava SK. STAT3 induces anoikis resistance, promotes cell Writing – review & editing. Ramón Charco: Writing – review & editing. invasion and metastatic potential in pancreatic cancer cells. Carcinogenesis 2015; 36(1):142–50. https://doi.org/10.1093/carcin/bgu233. Ana Vivancos: Supervision, Writing – review & editing. [12] Gilmore AP. Anoikis. Cell Death Differ 2005;12(Suppl. 2):1473–7. https://doi.org/ 10.1038/sj.cdd.4401723. Declaration of Competing Interest [13] Cabel L, Proudhon C, Mariani P, Tzanis D, Beinse G, Bieche I, et al. Circulating tumor cells and circulating tumor DNA: what surgical oncologists need to know? Eur J Surg Oncol 2017;43(5):949–62. https://doi.org/10.1016/j.ejso.2017.01.010. The authors declare the following financial interests/personal re­ [14] Sausen M, Phallen J, Adleff V, Jones S, Leary RJ, Barrett MT, et al. Clinical lationships which may be considered as potential competing interests: implications of genomic alterations in the tumour and circulation of pancreatic cancer patients. Nat Commun 2015;6(1). https://doi.org/10.1038/ncomms8686. Laura Vidal reports article publishing charges was provided by Auton­ [15] Hadano N, Murakami Y, Uemura K, Hashimoto Y, Kondo N, Nakagawa N, et al. omous University of Barcelona. Prognostic value of circulating tumour DNA in patients undergoing curative resection for pancreatic cancer. Br J Cancer 2016;115(1):59–65. https://doi.org/ Acknowledgements 10.1038/bjc.2016.175. [16] Takai E, Totoki Y, Nakamura H, Morizane C, Nara S, Hama N, et al. Clinical utility of circulating tumor DNA for molecular assessment in pancreatic cancer. Sci Rep The authors would like to thank Dr. Rodrigo Dienstmann for his 2015;5(1). https://doi.org/10.1038/srep18425. 8 L. Vidal et al. Cancer Treatment Reviews 120 (2023) 102604 [17] Kinugasa H, Nouso K, Miyahara K, Morimoto Y, Dohi C, Tsutsumi K, et al. [41] Nomoto S, Nakao A, Kasai Y, Harada A, Nonami T, Takagi H. Detection of ras gene Detection of K-ras gene mutation by liquid biopsy in patients with pancreatic mutations in perioperative peripheral blood with pancreatic adenocarcinoma. Jpn cancer. Cancer 2015;121(13):2271–80. https://doi.org/10.1002/cncr.29364. J Cancer Res 1996;87(8):793–7. https://doi.org/10.1111/j.1349-7006.1996. [18] Bettegowda C, Sausen M, Leary RJ, Kinde I, Wang Y, Agrawal N, et al. Detection of tb02102.x. circulating tumor DNA in early- and late-stage human malignancies. Sci Transl Med [42] Watanabe F, Suzuki K, Tamaki S, et al. Longitudinal monitoring of KRAS-mutated 2014;6(224):224ra24. https://doi.org/10.1126/scitranslmed.3007094. circulating tumor DNA enables the prediction of prognosis and therapeutic [19] Pietrasz D, Pécuchet N, Garlan F, et al. Plasma circulating tumor DNA in pancreatic responses in patients with pancreatic cancer. PLoS ONE 2019;14(12):e0227366. cancer patients is a prognostic marker. Clin Cancer Res. 2017;23(1):116–23. doi: Published 2019 Dec 31. doi: 10.1371/journal.pone.0227366. 10.1158/1078-0432.CCR-16-0806. [43] Yamada T, Nakamori S, Ohzato H, et al. Detection of K-ras gene mutations in [20] Henriksen TV, Reinert T, Christensen E, Sethi H, Birkenkamp-Demtröder K, plasma DNA of patients with pancreatic adenocarcinoma: correlation with Gögenur M, et al. The effect of surgical trauma on circulating free DNA levels in clinicopathological features. Clin Cancer Res 1998;4(6):1527–32. cancer patients-implications for studies of circulating tumor DNA. Mol Oncol 2020; [44] Yamaguchi T, Uemura K, Murakami Y, Kondo N, Nakagawa N, Okada K, et al. 14(8):1670–9. https://doi.org/10.1002/1878-0261.12729. Clinical implications of pre- and postoperative circulating tumor DNA in patients [21] Turnbull Jr RB, Kyle K, Watson FR, Spratt J. Cancer of the colon: the influence of with resected pancreatic ductal adenocarcinoma. Ann Surg Oncol 2021;28(6): the no-touch isolation technic on survival rates. Ann Surg 1967;166(3):420–7. 3135–44. https://doi.org/10.1245/s10434-020-09278-9. https://doi.org/10.1097/00000658-196709000-00010. [45] Yang X, Xu W, Tian X, Wu J, Lv A, Li C, et al. Diagnostic and prognostic value of [22] Hayashi N, Egami H, Kai M, Kurusu Y, Takano S, Ogawa M. No-touch isolation KRAS mutations in circulating pancreatic ductal adenocarcinoma tumor DNA. technique reduces intraoperative shedding of tumor cells into the portal vein Transl Cancer Res 2018;7(3):622–33. https://doi.org/10.21037/tcr.2018.05.33. during resection of colorectal cancer. Surgery 1999;125(4):369–74. [46] Vivancos A, Tabernero J. Circulating tumor DNA as a novel prognostic indicator. [23] Wind J, Tuynman JB, Tibbe AGJ, Swennenhuis JF, Richel DJ, van Berge Nat Med 2022;28(11):2255–6. https://doi.org/10.1038/s41591-022-02068-8. Henegouwen MI, et al. Circulating tumour cells during laparoscopic and open [47] Kustanovich A, Schwartz R, Peretz T, Grinshpun A. Life and death of circulating surgery for primary colonic cancer in portal and peripheral blood. Eur J Surg Oncol cell-free DNA. Cancer Biol Ther 2019;20(8):1057–67. https://doi.org/10.1080/ 2009;35(9):942–50. https://doi.org/10.1016/j.ejso.2008.12.003. 15384047.2019.1598759. [24] Wiggers T, Jeekel J, Arends JW, et al. No-touch isolation technique in colon cancer: [48] Diehl F, Schmidt K, Choti MA, Romans K, Goodman S, Li M, et al. Circulating a controlled prospective trial. Br J Surg 1988;75(5):409–15. doi: 10.1002/ mutant DNA to assess tumor dynamics. Nat Med 2008;14(9):985–90. https://doi. bjs.1800750505. org/10.1038/nm.1789. [25] Takii Y, Mizusawa J, Kanemitsu Y, Komori K, Shiozawa M, Ohue M, et al. The [49] Thierry AR, El Messaoudi S, Gahan PB, Anker P, Stroun M. Origins, structures, and conventional technique versus the no-touch isolation technique for primary tumor functions of circulating DNA in oncology. Cancer Metastasis Rev 2016;35(3): resection in patients with colon cancer (JCOG1006): a multicenter, open-label, 347–76. https://doi.org/10.1007/s10555-016-9629-x. randomized phase III trial. Ann Surg 2022;275(5):849–55. https://doi.org/ [50] Cohen SA, Kasi PM, Aushev VN, Hanna DL, Botta GP, Sharif S, et al. Kinetics of 10.1097/SLA.0000000000005241. postoperative circulating cell-free DNA and impact on minimal residual disease [26] Gall TMH, Jacob J, Frampton AE, Krell J, Kyriakides C, Castellano L, et al. Reduced detection rates in patients with resected stage I-III colorectal cancer. J. Clin Oncol dissemination of circulating tumor cells with no-touch isolation surgical technique 2023;41(4_suppl):5. in patients with pancreatic cancer. JAMA Surg 2014;149(5):482. [51] Conroy T, Hammel P, Hebbar M, Ben Abdelghani M, Wei AC, Raoul J-L, et al. [27] Hirota M, Shimada S, Yamamoto K, et al. Pancreatectomy using the no-touch FOLFIRINOX or gemcitabine as adjuvant therapy for pancreatic cancer. N Engl J isolation technique followed by extensive intraoperative peritoneal lavage to Med 2018;379(25):2395–406. https://doi.org/10.1056/NEJMoa1809775. prevent cancer cell dissemination: a pilot study. JOP 2005;6(2):143–51. Published [52] Houg DS, Bijlsma MF. The hepatic pre-metastatic niche in pancreatic ductal 2005 Mar 10. adenocarcinoma. Mol Cancer 2018;17(1):95. https://doi.org/10.1186/s12943- [28] Chen L, Zhang Y, Cheng Y, Zhang D, Zhu S, Ma X. Prognostic value of circulating 018-0842-9. Published 2018 Jun 14. cell-free DNA in patients with pancreatic cancer: a systemic review and meta- [53] Wang H, Pan J, Barsky L, Jacob JC, Zheng Y, Gao C, et al. Characteristics of pre- analysis. Gene 2018;679:328–34. https://doi.org/10.1016/j.gene.2018.09.029. metastatic niche: the landscape of molecular and cellular pathways. Mol Biomed [29] Milin-Lazovic J, Madzarevic P, Rajovic N, et al. Meta-analysis of circulating cell- 2021;2(1). https://doi.org/10.1186/s43556-020-00022-z. free DNA’s role in the prognosis of pancreatic cancer. Cancers (Basel) 2021;13(14): [54] Mohammed S, Van Buren 2nd G, Fisher WE. Pancreatic cancer: advances in 3378. doi: 10.3390/cancers13143378 Published 2021 Jul 6. treatment. World J Gastroenterol 2014;20(28):9354–60. https://doi.org/10.3748/ [30] Bunduc S, Gede N, Váncsa S, Lillik V, Kiss S, Dembrovszky F, et al. Prognostic role wjg.v20.i28.9354. of cell-free DNA biomarkers in pancreatic adenocarcinoma: a systematic review [55] Kang Q, Henry NL, Paoletti C, Jiang H, Vats P, Chinnaiyan AM, et al. Comparative and meta-analysis. Crit Rev Oncol Hematol 2022;169:103548. https://doi.org/ analysis of circulating tumor DNA stability In K3EDTA, streck, and cell save blood 10.1016/j.critrevonc.2021.103548. collection tubes. Clin Biochem 2016;49(18):1354–60. https://doi.org/10.1016/j. [31] Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA clinbiochem.2016.03.012. (editors). Cochrane handbook for systematic reviews of interventions version 6.3 [56] van Dessel LF, Beije N, Helmijr JCA, Vitale SR, Kraan J, Look MP, et al. Application (updated February 2022). Cochrane; 2022. Available from www.training.cochrane. of circulating tumor DNA in prospective clinical oncology trials - standardization of org/handbook. preanalytical conditions. Mol Oncol 2017;11(3):295–304. https://doi.org/ [32] Riley RD, Moons KGM, Snell KIE, et al. A guide to systematic review and meta- 10.1002/1878-0261.12037. analysis of prognostic factor studies. BMJ. 2019;364:k4597. Published 2019 Jan [57] Ansari D, Tingstedt B, Andersson B, Holmquist F, Sturesson C, Williamsson C, et al. 30. doi: 10.1136/bmj.k4597. Pancreatic cancer: yesterday, today and tomorrow. Future Oncol 2016;12(16): [33] Hayden JA, van der Windt DA, Cartwright JL, Côté P, Bombardier C. Assessing bias 1929–46. https://doi.org/10.2217/fon-2016-0010. in studies of prognostic factors. Ann Intern Med 2013;158(4):280–6. https://doi. [58] Neoptolemos JP, Kleeff J, Michl P, Costello E, Greenhalf W, Palmer DH. org/10.7326/0003-4819-158-4-201302190-00009. Therapeutic developments in pancreatic cancer: current and future perspectives. [34] Deeks JJ, Higgins JPT, Altman DG (editors). Chapter 10: analysing data and Nat Rev Gastroenterol Hepatol 2018;15(6):333–48. https://doi.org/10.1038/ undertaking meta-analyses. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li s41575-018-0005-x. T, Page MJ, et al., editors. Cochrane handbook for systematic reviews of [59] Topal H, Aerts R, Laenen A, Collignon A, Jaekers J, Geers J, et al. Survival after interventions version 6.3 (updated February 2022). Cochrane; 2022. Available minimally invasive vs open surgery for pancreatic adenocarcinoma. JAMA Netw from www.training.cochrane.org/handbook. Open 2022;5(12):e2248147. https://doi.org/10.1001/ [35] Page MJ, Moher D, Bossuyt PM, et al. PRISMA 2020 explanation and elaboration: jamanetworkopen.2022.48147. updated guidance and exemplars for reporting systematic reviews. BMJ. 2021;372: n160. Published 2021 Mar 29. doi: 10.1136/bmj.n160. [36] Groot VP, Mosier S, Javed AA, et al. Circulating tumor DNA as a clinical test in Glossary resected pancreatic cancer. Clin Cancer Res. 2019;25(16):4973–84. doi: 10.1158/ 1078-0432.CCR-19-0197. ctDNA: circulating tumour DNA. [37] Jiang J, Ye S, Xu Y, et al. Circulating tumor DNA as a potential marker to detect CTCs: circulating tumour cells. minimal residual disease and predict recurrence in pancreatic cancer. Front Oncol. mRNA: messenger RNA. 2020;10:1220. Published 2020 Jul 30. doi: 10.3389/fonc.2020.01220. PDAC: pancreatic ductal adenocarcinoma. [38] Kitahata Y, Kawai M, Hirono S, Okada K-I, Miyazawa M, Motobayashi H, et al. Ct: circulating tumour. Circulating tumor DNA as a potential prognostic marker in patients with EMT: epithelial to mesenchymal transition. borderline-resectable pancreatic cancer undergoing neoadjuvant chemotherapy DFS: disease-free survival. followed by pancreatectomy. Ann Surg Oncol 2022;29(3):1596–605. https://doi. OS: overall survival. org/10.1245/s10434-021-10985-0. NTIT: non-touch isolation techniques. [39] Lee B, Lipton L, Cohen J, Tie J, Javed AA, Li L, et al. Circulating tumor DNA as a ctRNA: circulating tumour RNA. potential marker of adjuvant chemotherapy benefit following surgery for localized MeSH: medical Subject Heading. pancreatic cancer. Ann Oncol 2019;30(9):1472–8. https://doi.org/10.1093/ HR: Hazar Ratio. annonc/mdz200. RR: Risk Ratio. [40] Nakano Y, Kitago M, Matsuda S, Nakamura Y, Fujita Y, Imai S, et al. KRAS OR: Odds Ratio. mutations in cell-free DNA from preoperative and postoperative sera as a QUIPS tool: Quality In Prognosis Studies tool. pancreatic cancer marker: a retrospective study. Br J Cancer 2018;118(5):662–9. AJCC: American Joint Committee of Cancer. https://doi.org/10.1038/bjc.2017.479. PFS: Progression-free survival. PMN: pre-metastatic niche. 9 L. Vidal et al. Cancer Treatment Reviews 120 (2023) 102604 ddPCR: droplet digital PCR. nPD: Non touch isolation technique pancreatoduodenectomy. NGS: next generation sequencing. RT: reverse transcriptase. MASA-PCR: Mutant allele specific amplification PCR. PV: portal vein. PD: pancreatoduodenectomy. PO: postoperative. DP: distal pancreatectomy. POD: postoperative day. TP: total pancreatectomy. NAC: neoadjuvancy. sDP: standard distal pancreatectomy. LoD: Limit of detection. sPD: standard pancreatoduodenectomy. EDTA: ethylenediaminetetraacetic acid. nDP: Non touch isolation technique distal pancreatectomy. cfDNA: cell free DNA. 10