Impact of Tumor-Derived DNA Testing in Peritoneal Lavage of Pancreatic Cancer Patients with and Without Occult Intra-Abdominal Metastases

Preoperative predictive factors for positive peritoneal cytology results in patients with pancreatic ductal adenocarcinomas: a retrospective study

BACKGROUND

The clinical importance of positive peritoneal cytology results in patients with pancreatic ductal adenocarcinomas remains controversial. We evaluated the prognosis of these patients and the predictive preoperative risk factors for positive peritoneal cytology results.

METHODS

We retrospectively reviewed patients who underwent curative-intent surgery at our institution between May 2010 and June 2020. Preoperative risk factors for positive peritoneal cytology results were identified using logistic regression analysis. A scoring model was constructed using the total number of significant independent predictors for positive peritoneal cytology results.

RESULTS

Of 233 patients, 18 (7.7%) had positive peritoneal cytology results. The recurrence-free survival and cancer-specific survival were markedly worse in patients with positive peritoneal cytology results than in those with negative peritoneal cytology results (recurrence-free survival: 6.0 months vs. 16.6 months, p = 0.050; cancer-specific survival: 19.4 months vs. 47.5 months, p = 0.034). Tumor location (odds ratio: 3.760, 95% confidence interval: 1.099-11.818, p = 0.023), tumor size > 25 mm (odds ratio: 3.410, 95% confidence interval: 1.031-11.277, p = 0.046), preoperative serosal invasion (odds ratio: 5.193, 95% confidence interval: 1.099-24.531, p = 0.038), and preoperative carcinoembryonic antigen level > 5.6 ng/mL (odds ratio: 3.816, 95% confidence interval: 1.248-10.667, p = 0.019) were identified as significant independent predictive factors. Our predictive model's optimal cutoff and positive predictive values for positive peritoneal cytology results were 3 and 27.9%, respectively.

CONCLUSIONS

The indications for curative-intent surgery should be considered carefully in patients with high-risk factors for positive peritoneal cytology results.

Role of Diagnostic Laparoscopy During Pancreatic Cancer Surgery in the Modern Era

INTRODUCTION

Despite improvements in preoperative image resolution, approximately 10% of curative-intent resection attempts for pancreatic ductal adenocarcinoma are aborted at the time of operation. To avoid nontherapeutic laparotomy, many surgeons perform intraoperative diagnostic laparoscopy (DL) to identify radiographically occult metastatic disease. There are no consensus guidelines regarding DL in pancreatic cancer. The goal of this study is to investigate the efficacy of same-procedure DL at avoiding nontherapeutic laparotomy.

METHODS

A single-institution retrospective review was performed from 2016 to 2022, identifying 196 patients with pancreatic ductal adenocarcinoma who were taken to the operating room for open curative-intent resection. Patient demographic, tumor characteristic, treatment, and outcome data were abstracted. Univariate and multivariate Cox hazard ratio analysis was performed to investigate risk factors for overall survival and recurrence-free survival. Number needed to treat (NNT) was calculated to identify number of DLs necessary to avoid one nontherapeutic laparotomy.

RESULTS

Curative-intent resection was achieved in 161 (82.1%) patients. One hundred twenty six (64.0%) patients received DL prior to resection and DL identified metastatic disease in three (2.4%) patients with an NNT of 42. NNT of DL in a subgroup analysis performed on clinically high-risk patients (defined by preoperative or preneoadjuvant therapy carbohydrate antigen 19-9 > 500 U/mL) is 11. Receipt of DL did not prolong operative times in patients receiving pancreaticoduodenectomy when accounting for completed versus aborted resection.

CONCLUSIONS

Although intraoperative DL is a short procedure with minimal morbidity, these data demonstrate that same-procedure DL has potential efficacy in avoiding nontherapeutic laparotomy only in a subgroup of clinically high-risk patients. Focus should remain on optimizing preoperative patient selection and further investigating novel diagnostic markers predictive of occult metastatic disease.

Clinical impact of high-quality testing for peritoneal lavage cytology in pancreatic cancer

In pancreatic ductal adenocarcinoma (PDAC) patients, the importance of peritoneal lavage cytology, which indicates unresectability, remains controversial. This study sought to determine whether positive peritoneal lavage cytology (CY+) precludes pancreatectomy. Furthermore, we propose a novel liquid biopsy using peritoneal lavage fluid to detect viable peritoneal tumor cells (v-PTCs) with TelomeScan F35, a telomerase-specific replication-selective adenovirus engineered to express green fluorescent protein. Resectable cytologically or histologically proven PDAC patients (n = 53) were enrolled. CY was conducted immediately following laparotomy. The resulting fluid was examined by conventional cytology (conv-CY; Papanicolaou staining and MOC-31 immunostaining) and by the novel technique (Telo-CY; using TelomeScan F35). Of them, 5 and 12 were conv-CY+ and Telo-CY+, respectively. All underwent pancreatectomy. The two double-CY+ (conv-CY+ and Telo-CY+) patients showed early peritoneal recurrence (P-rec) postoperatively, despite adjuvant chemotherapy. None of the three conv-CY+ Telo-CY- patients exhibited P-rec. Six of the 10 Telo-CY+ conv-CY- patients (60%) relapsed with P-rec. Of the remaining 38 double-CY- [conv-CY-, Telo-CY-, conv-CY± (Class III)] patients, 3 (8.3%) exhibited P-rec. Although conv-CY+ status predicted poor prognosis and a higher risk of P-rec, Telo-CY was more sensitive for detecting v-PTC. Staging laparoscopy and performing conv-CY and Telo-CY are needed to confirm the indication for pancreatectomy.

The prognostic impact of peritoneal tumour DNA in gastrointestinal and gynaecological malignancies: a systematic review

Peritoneal metastases from various abdominal cancer types are common and carry poor prognosis. The presence of peritoneal disease upstages cancer diagnosis and alters disease trajectory and treatment pathway in many cancer types. Therefore, accurate and timely detection of peritoneal disease is crucial. The current practice of diagnostic laparoscopy and peritoneal lavage cytology (PLC) in detecting peritoneal disease has variable sensitivity. The significant proportion of peritoneal recurrence seen during follow-up in patients where initial PLC was negative indicates the ongoing need for a better diagnostic tool for detecting clinically occult peritoneal disease, especially peritoneal micro-metastases. Advancement in liquid biopsy has allowed the development and use of peritoneal tumour DNA (ptDNA) as a cancer-specific biomarker within the peritoneum, and the presence of ptDNA may be a surrogate marker for early peritoneal metastases. A growing body of literature on ptDNA in different cancer types portends promising results. Here, we conduct a systematic review to evaluate the prognostic impact of ptDNA in various cancer types and discuss its potential future clinical applications, with a focus on gastrointestinal and gynaecological malignancies.

Peritoneal Cell-Free Tumor DNA is a Biomarker of Locoregional and Peritoneal Recurrence in Resected Pancreatic Ductal Adenocarcinomas

BACKGROUND

Recurrence after curative-intent pancreatectomy for pancreatic ductal adenocarcinomas (PDAC) is quite frequent with locoregional and peritoneal recurrence in about one-third of cases. We hypothesize that peritoneal cell-free tumor DNA (ptDNA) present in the intraoperative peritoneal lavage (PL) fluid may be used as a predictive biomarker of locoregional and peritoneal recurrence.

PATIENTS AND METHODS

Under institutional review board (IRB)-approved protocol, pre- and postresection PL fluids were collected from PDAC patients undergoing curative-intent pancreatectomy. PL fluids from PDAC patients with pathologically proven peritoneal metastasis were also collected as positive controls. Cell-free DNA was extracted from PL fluids. Droplet digital PCR (ddPCR) was performed using ddPCR KRAS G12/G13 screening kit. Recurrence-free survival (RFS) based on KRAS-mutant ptDNA level was determined using Kaplan-Meier methods.

RESULTS

KRAS-mutant ptDNA was detected in PL fluids from all PDAC patients. KRAS-mutant ptDNA was detected in 11/21 (52%) preresection and 15/18 (83%) postresection PL fluid samples. With a median follow-up of 23.6 months, 12 patients developed recurrence (8 locoregional/peritoneal recurrence, 9 pulmonary/hepatic recurrence); 5/8 (63%) and 6/6 (100%) patients with mutant allele frequency (MAF) of > 0.10% in pre- and postresection PL fluids, respectively, developed recurrence. Using a cutoff value of 0.10% MAF, the presence of KRAS-mutant ptDNA in postresection PL fluid predicted a significantly shortened time to locoregional and peritoneal recurrence (median RFS of 8.9 months versus not reached, P = 0.003).

CONCLUSIONS

This study suggests that ptDNA in postresection PL fluids may be a useful biomarker to predict locoregional and peritoneal recurrence in resected PDAC patients.

Simultaneous and sequential combination of genetic and epigenetic biomarkers for the presence of high-grade dysplasia in patients with pancreatic cyst: Discovery in cyst fluid and test in pancreatic juice

BACKGROUND/OBJECTIVES

Screening patients with intraductal papillary mucinous neoplasms (IPMN) has the primary goal of identifying potentially curable noninvasive precursors. We aimed to evaluate the diagnostic impact of genetic and epigenetic biomarkers in the presence of noninvasive precursors.

METHODS

Mutated KRAS/GNAS and methylated SOX17/TBX15/BMP3/TFPI2 DNA were assessed by droplet digital PCR in a discovery cohort of 70 surgically aspirated cyst fluids, and diagnostic performances for differentiating high-grade dysplasia (HGD) from low-grade dysplasia (LGD) was evaluated. We then tested these markers using an independent test cohort consisting of 156 serially collected pancreatic juice samples from 30 patients with IPMN.

RESULTS

Mutated KRAS and GNAS are specific for IPMNs but are not helpful for the prediction of histological grades. Cyst fluids from IPMN with HGD showed higher methylation levels of SOX17 (median, 0.141 vs. 0.021; P = 0.086) and TBX15 (median, 0.030 vs. 0.003; P = 0.028) than those with LGD. The combination of all tested markers yielded a diagnostic performance with sensitivity of 69.6%, and specificity of 90.0%. Among the 30 pancreatic juice samples exhibiting the highest abundance of KRAS/GNAS mutations in each patient in the test cohort, patients with histologically proven HGD due to pancreatic resection had a significantly higher prevalence (100% vs. 31%, P = 0.018) and abundance (P = 0.037) of methylated TBX15 than those without cytohistological diagnosis undergoing surveillance.

CONCLUSIONS

A simultaneous and sequential combination of mutated and methylated DNA markers in pancreatic cyst fluid and juice sample markers can help detect noninvasive pancreatic precursor neoplasms.

Levels of tumor markers CEA / CA 19–9 in serum and peritoneal lavage predict postoperative recurrence in patients with pancreatic cancer

Aim

This study aimed to clarify the usefulness of tumor markers from peritoneal lavage in selecting patients with a high risk of recurrence and predicting site‐specific recurrence in patients with pancreatic cancer.

Methods

The levels of serum carcinoembryonic antigen (CEA) and carbohydrate antigen (CA) 19–9 (sCEA/sCA 19–9) and paired peritoneal lavage CEA and CA 19–9 (pCEA/pCA 19–9) were measured in 90 patients with pancreatic cancer who underwent surgery. Using the cutoff values determined by maximally selected rank statistics for disease‐free survival (DFS), the risk of recurrence and its patterns were evaluated in combination with different markers and different test specimens.

Results

In univariate and multivariate analysis, an elevated pCA 19–9 level (>1.3 U/mL) was an independent prognostic marker for both DFS (hazard ratio [HR], 2.391; P = .018) and overall survival (HR, 3.194; P = .033). Combination analyses contributed to further stratification of a very high risk of recurrence. Of the 58 patients with resectable pancreatic cancer who underwent curative resection, elevated pCA19–9 was also associated with inferior DFS and overall survival (OS). Patients with elevated pCA 19–9 levels were more likely to have an earlier onset of peritoneal recurrence than those with normal pCA 19–9 levels (P = .048, Gehan–Breslow–Wilcoxon test).

Conclusion

pCA 19–9 is a reliable marker for predicting postoperative recurrence in patients with pancreatic cancer after surgery. Further risk stratification can be achieved by using combination assays. The combination of pCA 19–9 and sCA19–9 also serves as a predictor of recurrence site‐specific recurrence.