Original Article
Detection of K-ras Gene Mutation by Liquid Biopsy in Patients With Pancreatic Cancer Hideaki Kinugasa, MD, PhD1; Kazuhiro Nouso, MD, PhD1; Koji Miyahara, MD, PhD1; Yuki Morimoto, MD1; Chihiro Dohi, MD1; Koichiro Tsutsumi, MD, PhD1; Hironari Kato, MD, PhD1; Takehiro Matsubara, PhD2; Hiroyuki Okada, MD, PhD1; and Kazuhide Yamamoto, MD, PhD1
BACKGROUND: Cell-free circulating tumor DNA (ctDNA) in serum has been considered to be a useful candidate for noninvasive cancer diagnosis. The current study was designed to estimate the clinical usefulness of genetic analysis for ctDNA by digital polymerase chain reaction in patients with pancreatic cancer. METHODS: The authors compared K-ras mutations detected in endoscopic ultrasound-guided fine-needle aspiration biopsy tissue DNA and in ctDNA from 75 patients with pancreatic cancer. K-ras mutations in the serum of 66 independent, consecutive patients with pancreatic cancer were also analyzed and the authors compared the results with survival rates. RESULTS: The frequencies of the mutations in tissue samples at G12V, G12D, and G12R in codon 12 were 28 of 75 samples (37.3%), 22 of 75 samples (29.3%), and 6 of 75 samples (8.0%), respectively. Conversely, the rates of the mutations in ctDNA were 26 of 75 samples (34.6%), 29 of 75 samples (38.6%), and 4 of 75 samples (5.3%), respectively. Overall, the K-ras mutation rates in tissue and ctDNA were 74.7% and 62.6%, respectively, and the concordance rate between them was 58 of 75 samples (77.3%). Survival did not appear to differ by the presence of K-ras mutations in tissue DNA, but the survival of patients with K-ras mutations in ctDNA was significantly shorter than that of patients without mutations in both a development set (P 5.006) and an independent validation set (P 5.002). The difference was especially evident in cases with a G12V mutation. CONCLUSIONS: Analysis of ctDNA is a new useful procedure for detecting mutations in patients with pancreatic cancer. This noninvasive method may have great potential C 2015 as a new strategy for the diagnosis of pancreatic cancer as well as for predicting survival. Cancer 2015;000:000-000. V American Cancer Society. KEYWORDS: digital polymerase chain reaction, liquid biopsy, circulating tumor DNA (ctDNA), pancreatic cancer, K-ras, circulating DNA.
INTRODUCTION Mutations in the ras genes (K-ras, H-ras, and N-ras) are common alterations that are identified in many types of human tumors and have been implicated in the development of human cancer. These genes are converted to active oncogenes by point mutations occurring in either codon 12, codon 13, or codon 61.1 K-ras is known to be mutated in colon, lung, and pancreatic cancer, whereas H-ras and N-ras have been reported to be associated with bladder cancer, myeloid leukemia, and breast tumors.2 Of these genes, K-ras is the most frequently mutated gene in pancreatic cancer, with a reported mutation rate ranging from 75%3 to 95%.4 Furthermore, activation of the K-ras oncogene appears to be an early event in pancreatic cancer, and some types of K-ras mutations affect patient survival.5,6 Histological examination of tumor tissues obtained by endoscopic ultrasound-guided fine-needle aspiration (EUSFNA) is the gold standard for the diagnosis of pancreatic cancer, in addition to imaging modalities such as computed tomography and magnetic resonance imaging. However, the differential diagnosis of pancreatic solid masses with EUSFNA is still challenging in approximately 15% of cases.7 Therefore, rapid assays with probes specific for K-ras point mutations that have been developed for increasing the detection rate are often used clinically.8 Some studies have indicated that K-ras mutation analysis from tissue obtained by EUS-FNA can be useful in the diagnostic workup of pancreatic masses, in particular when the tissue obtained by EUS-FNA is insufficient and the diagnosis is inconclusive.7 However, the diagnostic accuracy of EUS-FNA is influenced by several factors, such as the quality and quantity of the material obtained, the size and location of the mass, and the technical skill of the endoscopist as well as the presence of a cytopathologist on site.9,10
Corresponding author: Hideaki Kinugasa, MD, PhD, Department of Gastroenterology and Hepatology, Okayama University Graduate School of Medicine, Dentistry, and Pharmaceutical Sciences, 2-5-1 Shikata-cho, Kita-ku, Okayama 700-8558, Japan; Fax: (011) 81-86-225-5991; [email protected] 1
Department of Gastroenterology and Hepatology, Okayama University Graduate School of Medicine, Dentistry, and Pharmaceutical Sciences, Okayama, Japan; BioRepository/BioMarker Analysis Center, Innovative Clinical Medicine, Okayama University Hospital, Okayama, Japan
2
DOI: 10.1002/cncr.29364, Received: December 6, 2014; Revised: January 28, 2015; Accepted: February 19, 2015, Published online Month 00, 2015 in Wiley Online Library (wileyonlinelibrary.com)
Cancer
Month 00, 2015
1
Original Article
Cell-free circulating tumor DNA (ctDNA) in the plasma or serum has been considered to be a candidate for noninvasive cancer diagnosis by liquid biopsy, which has been explored to detect somatic mutations from malignant tumors.11 However, to the best of our knowledge, previous mutation detection assays have not been sufficiently sensitive, specific, or quantitative for the assessment of the clinical usefulness of ctDNA because the number of ctDNA molecules with somatic mutations is very low compared with wild-type molecules. Newly developed technologies such as digital polymerase chain reaction (PCR) and next-generation sequencing have provided new insight in this area.12-14 These technologies afford a very high level of sensitivity and specificity, and enable quantification of the ratio of mutant to normal DNA molecules. Such technologies have dramatically improved the detection rate of rare mutations.12-14 The objective of the current study was to elucidate the clinical usefulness of detecting K-ras mutations of ctDNA by digital PCR in patients with pancreatic cancer.
MATERIALS AND METHODS Patient Characteristics
EUS-FNA biopsy tissue and serum samples were obtained from a total of 75 consecutive patients with pancreatic cancer between January 2008 and December 2010, and were analyzed as a development cohort study. All patients were Japanese and all the samples were obtained before therapy. The median age of the patients was 66 years (range, 47-85 years) and the male:female ratio was 2.57 (54/21). A total of 0, 2, 5, and 68 patients, respectively, were diagnosed with stage I, stage II, stage III, and stage IV disease by the Japanese Pancreatic Society (JPS) classification. Tumors were located in the head, body, and tail of the pancreas in 44.0%, 37.3%, and 18.7% of the patients, respectively. The median carbohydrate antigen 19-9 (CA 19-9), DUPAN2, and SPAN-1 levels were 335 U/mL (range, 0.6-99,999 U/mL), 684 U/mL (range, 25-16,000 U/ mL), and 176 U/mL (range, 10-210000 U/mL), respectively. Of these 75 patients, diabetes, a body mass index >30, smoking, and drinking were found to be associated with 32.4%, 0%, 59.7%, and 63.6% cases of pancreatic cancer, respectively. The median body mass index was 20.8 (range, 12.5-29.5). Diabetes, smoking, drinking, CA 19-9 level (>335 U/mL), and tumor location were not found to be correlated with K-ras mutation (P values of .92, .53, .49, .14, and .57, respectively). Additional serum samples were obtained 2
from 66 consecutive patients with pancreatic cancer between January 2011 and June 2013, and were analyzed as an independent validation cohort set. In the independent cohort, the median age of the patients was 66 years (range, 38-85 years) and the male:female ratio was 1.87 (43/23). A total of 1, 0, 8, and 57 patients were diagnosed with stage I, stage II, stage III, and stage IV disease, respectively. The median CA 19-9, DUPAN2, and SPAN-1 levels were 544 U/mL (range, 0.6-68,190 U/mL), 379 U/mL (range, 25-16,000 U/ mL), and 158 U/mL (range, 10-20,000 U/mL), respectively (Table 1). All EUS-FNA samples were assessed for the presence of the 6 most frequent mutations in K-ras codon 12 (AGT [G12S], CGT [G12R], TGT [G12C], GAT [G12D], GCT [G12A], and GTT [G12V]) with a PCR-preferential homoduplex formation assay (PCR-PHFA)15 before entrance into the current study. All patients provided written informed consent to examine their serum and use their clinical data. The study protocol conformed to the ethical guidelines of the World Medical Association Declaration of Helsinki and was approved by the Okayama University Ethics Committee. Serum from 20 healthy patients and 20 patients with chronic pancreatitis were collected as controls. The median age of the patients was 44 years (range, 22-76 years) and 61 years (range, 40-84 years), respectively. Chronic pancreatitis was diagnosed according to the diagnostic criteria of the Japan Pancreas Society.16 EUS-FNA Technique
We imaged the pancreas at a 7.5-megahertz frequency using a convex linear-array echoendoscope (GF-UCT240 or GF-UCT260; Olympus, Tokyo, Japan) connected to an ultrasound device (ProSound Alpha 10; Aloka, Tokyo, Japan) and 22-gauge needle (Expect Endoscopic Ultrasound Aspiration Needle; Boston Scientific, Marlborough, Mass). The aspiration materials were used for cytopathological evaluation, histological examination, and K-ras mutation analysis by PCR-PHFA. Sections were stained with hematoxylin and eosin as well as immunohistochemical staining for p53 when necessary. Extraction of Cell-Free ctDNA in Serum
Blood samples were collected in tubes (BD Vacutainer blood collection tubes; Becton, Dickinson and Company, Franklin Lakes, NJ) and processed within 1 hour after collection. The samples were centrifuged at 3000 3 g at 4 C to separate the serum from the peripheral blood cells, and stored at 280 C. ctDNA in serum was extracted from Cancer
Month 00, 2015
K-ras in Blood With Pancreatic Cancer/Kinugasa et al
TABLE 1. Characteristics of Patients From the Development Set and the Validation Set Characteristic Total Age, y Sex
Male Female
Median CA 19-9 (range), U/mL Median DUPAN2 (range),U/mL Median SPAN-1 (range), U/mL Stage of disease
Initial therapy
I II III IV GEM S-1 None Surgery Others
Development Set
Validation Set
75 66 (range, 47-85) 54 21 335 (0.6-99,999) 684 (25-16,000) 176 (10-210,000) 0 2 5 68 47 1 3 11 13
66 71 (range, 38-85) 43 23 544 (0.6-68,190) 379 (25-16,000) 158 (10-20,000) 1 0 8 57 40 5 3 13 5
P
.019 .381 .568 .042 .909 .250
.171
Abbreviations: CA 19-9, carbohydrate antigen 19-9; GEM, gemcitabine; S-1, oral fluoropyrimidine.
aliquots (1 mL) of serum obtained from 5 mL of blood with the QIAamp Circulating Nucleic Acid Kit (Qiagen, Valencia, Calif) according to the manufacturer’s instructions. Digital PCR
We used droplet digital PCR (ddPCR) (QX200; BioRad, Hercules, Calif), which is one of several available digital PCR17 methods. The sample was partitioned into 20,000 droplets, with target and background (reference) DNA randomly distributed among these droplets and uniformly. We used 3 types of PrimePCR products (BioRad) (G12V, G12D, and G12R) for ddPCR (cat 1863115, 1863113, and 1863112) because G12V, G12D, and G12R encompass nearly all K-ras mutations in pancreatic cancer18 (our EUS-FNA samples revealed that only these 3 types of K-ras mutations were detected in cases with K-ras mutations). Reactions were performed in 20 mL of reaction volume, which consisted of up to 130 ng of extracted DNA (5 mL), 2 3 droplet PCR supermix (10 mL), 20 3 target primer/probe mix (FAM) (1 mL), 20 3 reference primer/probe mix (HEX) (1 mL), and deionized distilled water (3 mL). Emulsified PCR reactions were run on a 96-well plate on a C1000 Touch thermal cycler incubating plate (Bio-Rad) at 95 C for 10 minutes followed by 40 cycles of incubation at 95 C for 15 seconds and 60 C for 60 seconds, followed by a 10-minute incubation at 98 C. Plates were read on a Bio-Rad QX200 droplet reader using QuantaSoft version 1.4.0 software provided by Bio-Rad to assess the numbers of droplets that were positive for wild-type K-ras and/or K-ras mutations. Digital PCR assay was performed on ctDNA in serum from 75 patients with pancreatic cancer in a development cohort and 66 patients with pancreatic canCancer
Month 00, 2015
cer in an independent validation cohort. All samples were analyzed in duplicate and variations were 0.001%.19 We confirmed that the control without serum ctDNA demonstrated no positive signal. Statistical Analysis
Overall survival was measured from the day of diagnosis by EUS-FNA to the date of death or the last follow-up examination. Survival curves were calculated using the Kaplan-Meier method. The log-rank test was used to compare survival curves. Univariate and multivariate analyses were used to examine potential prognostic factors and a Cox proportional hazards model was used to determine the most significant variables contributing to survival. Differences with a P value of
Detection of K-ras Gene Mutation by Liquid Biopsy in Patients With Pancreatic Cancer Hideaki Kinugasa, MD, PhD1; Kazuhiro Nouso, MD, PhD1; Koji Miyahara, MD, PhD1; Yuki Morimoto, MD1; Chihiro Dohi, MD1; Koichiro Tsutsumi, MD, PhD1; Hironari Kato, MD, PhD1; Takehiro Matsubara, PhD2; Hiroyuki Okada, MD, PhD1; and Kazuhide Yamamoto, MD, PhD1
BACKGROUND: Cell-free circulating tumor DNA (ctDNA) in serum has been considered to be a useful candidate for noninvasive cancer diagnosis. The current study was designed to estimate the clinical usefulness of genetic analysis for ctDNA by digital polymerase chain reaction in patients with pancreatic cancer. METHODS: The authors compared K-ras mutations detected in endoscopic ultrasound-guided fine-needle aspiration biopsy tissue DNA and in ctDNA from 75 patients with pancreatic cancer. K-ras mutations in the serum of 66 independent, consecutive patients with pancreatic cancer were also analyzed and the authors compared the results with survival rates. RESULTS: The frequencies of the mutations in tissue samples at G12V, G12D, and G12R in codon 12 were 28 of 75 samples (37.3%), 22 of 75 samples (29.3%), and 6 of 75 samples (8.0%), respectively. Conversely, the rates of the mutations in ctDNA were 26 of 75 samples (34.6%), 29 of 75 samples (38.6%), and 4 of 75 samples (5.3%), respectively. Overall, the K-ras mutation rates in tissue and ctDNA were 74.7% and 62.6%, respectively, and the concordance rate between them was 58 of 75 samples (77.3%). Survival did not appear to differ by the presence of K-ras mutations in tissue DNA, but the survival of patients with K-ras mutations in ctDNA was significantly shorter than that of patients without mutations in both a development set (P 5.006) and an independent validation set (P 5.002). The difference was especially evident in cases with a G12V mutation. CONCLUSIONS: Analysis of ctDNA is a new useful procedure for detecting mutations in patients with pancreatic cancer. This noninvasive method may have great potential C 2015 as a new strategy for the diagnosis of pancreatic cancer as well as for predicting survival. Cancer 2015;000:000-000. V American Cancer Society. KEYWORDS: digital polymerase chain reaction, liquid biopsy, circulating tumor DNA (ctDNA), pancreatic cancer, K-ras, circulating DNA.
INTRODUCTION Mutations in the ras genes (K-ras, H-ras, and N-ras) are common alterations that are identified in many types of human tumors and have been implicated in the development of human cancer. These genes are converted to active oncogenes by point mutations occurring in either codon 12, codon 13, or codon 61.1 K-ras is known to be mutated in colon, lung, and pancreatic cancer, whereas H-ras and N-ras have been reported to be associated with bladder cancer, myeloid leukemia, and breast tumors.2 Of these genes, K-ras is the most frequently mutated gene in pancreatic cancer, with a reported mutation rate ranging from 75%3 to 95%.4 Furthermore, activation of the K-ras oncogene appears to be an early event in pancreatic cancer, and some types of K-ras mutations affect patient survival.5,6 Histological examination of tumor tissues obtained by endoscopic ultrasound-guided fine-needle aspiration (EUSFNA) is the gold standard for the diagnosis of pancreatic cancer, in addition to imaging modalities such as computed tomography and magnetic resonance imaging. However, the differential diagnosis of pancreatic solid masses with EUSFNA is still challenging in approximately 15% of cases.7 Therefore, rapid assays with probes specific for K-ras point mutations that have been developed for increasing the detection rate are often used clinically.8 Some studies have indicated that K-ras mutation analysis from tissue obtained by EUS-FNA can be useful in the diagnostic workup of pancreatic masses, in particular when the tissue obtained by EUS-FNA is insufficient and the diagnosis is inconclusive.7 However, the diagnostic accuracy of EUS-FNA is influenced by several factors, such as the quality and quantity of the material obtained, the size and location of the mass, and the technical skill of the endoscopist as well as the presence of a cytopathologist on site.9,10
Corresponding author: Hideaki Kinugasa, MD, PhD, Department of Gastroenterology and Hepatology, Okayama University Graduate School of Medicine, Dentistry, and Pharmaceutical Sciences, 2-5-1 Shikata-cho, Kita-ku, Okayama 700-8558, Japan; Fax: (011) 81-86-225-5991; [email protected] 1
Department of Gastroenterology and Hepatology, Okayama University Graduate School of Medicine, Dentistry, and Pharmaceutical Sciences, Okayama, Japan; BioRepository/BioMarker Analysis Center, Innovative Clinical Medicine, Okayama University Hospital, Okayama, Japan
2
DOI: 10.1002/cncr.29364, Received: December 6, 2014; Revised: January 28, 2015; Accepted: February 19, 2015, Published online Month 00, 2015 in Wiley Online Library (wileyonlinelibrary.com)
Cancer
Month 00, 2015
1
Original Article
Cell-free circulating tumor DNA (ctDNA) in the plasma or serum has been considered to be a candidate for noninvasive cancer diagnosis by liquid biopsy, which has been explored to detect somatic mutations from malignant tumors.11 However, to the best of our knowledge, previous mutation detection assays have not been sufficiently sensitive, specific, or quantitative for the assessment of the clinical usefulness of ctDNA because the number of ctDNA molecules with somatic mutations is very low compared with wild-type molecules. Newly developed technologies such as digital polymerase chain reaction (PCR) and next-generation sequencing have provided new insight in this area.12-14 These technologies afford a very high level of sensitivity and specificity, and enable quantification of the ratio of mutant to normal DNA molecules. Such technologies have dramatically improved the detection rate of rare mutations.12-14 The objective of the current study was to elucidate the clinical usefulness of detecting K-ras mutations of ctDNA by digital PCR in patients with pancreatic cancer.
MATERIALS AND METHODS Patient Characteristics
EUS-FNA biopsy tissue and serum samples were obtained from a total of 75 consecutive patients with pancreatic cancer between January 2008 and December 2010, and were analyzed as a development cohort study. All patients were Japanese and all the samples were obtained before therapy. The median age of the patients was 66 years (range, 47-85 years) and the male:female ratio was 2.57 (54/21). A total of 0, 2, 5, and 68 patients, respectively, were diagnosed with stage I, stage II, stage III, and stage IV disease by the Japanese Pancreatic Society (JPS) classification. Tumors were located in the head, body, and tail of the pancreas in 44.0%, 37.3%, and 18.7% of the patients, respectively. The median carbohydrate antigen 19-9 (CA 19-9), DUPAN2, and SPAN-1 levels were 335 U/mL (range, 0.6-99,999 U/mL), 684 U/mL (range, 25-16,000 U/ mL), and 176 U/mL (range, 10-210000 U/mL), respectively. Of these 75 patients, diabetes, a body mass index >30, smoking, and drinking were found to be associated with 32.4%, 0%, 59.7%, and 63.6% cases of pancreatic cancer, respectively. The median body mass index was 20.8 (range, 12.5-29.5). Diabetes, smoking, drinking, CA 19-9 level (>335 U/mL), and tumor location were not found to be correlated with K-ras mutation (P values of .92, .53, .49, .14, and .57, respectively). Additional serum samples were obtained 2
from 66 consecutive patients with pancreatic cancer between January 2011 and June 2013, and were analyzed as an independent validation cohort set. In the independent cohort, the median age of the patients was 66 years (range, 38-85 years) and the male:female ratio was 1.87 (43/23). A total of 1, 0, 8, and 57 patients were diagnosed with stage I, stage II, stage III, and stage IV disease, respectively. The median CA 19-9, DUPAN2, and SPAN-1 levels were 544 U/mL (range, 0.6-68,190 U/mL), 379 U/mL (range, 25-16,000 U/ mL), and 158 U/mL (range, 10-20,000 U/mL), respectively (Table 1). All EUS-FNA samples were assessed for the presence of the 6 most frequent mutations in K-ras codon 12 (AGT [G12S], CGT [G12R], TGT [G12C], GAT [G12D], GCT [G12A], and GTT [G12V]) with a PCR-preferential homoduplex formation assay (PCR-PHFA)15 before entrance into the current study. All patients provided written informed consent to examine their serum and use their clinical data. The study protocol conformed to the ethical guidelines of the World Medical Association Declaration of Helsinki and was approved by the Okayama University Ethics Committee. Serum from 20 healthy patients and 20 patients with chronic pancreatitis were collected as controls. The median age of the patients was 44 years (range, 22-76 years) and 61 years (range, 40-84 years), respectively. Chronic pancreatitis was diagnosed according to the diagnostic criteria of the Japan Pancreas Society.16 EUS-FNA Technique
We imaged the pancreas at a 7.5-megahertz frequency using a convex linear-array echoendoscope (GF-UCT240 or GF-UCT260; Olympus, Tokyo, Japan) connected to an ultrasound device (ProSound Alpha 10; Aloka, Tokyo, Japan) and 22-gauge needle (Expect Endoscopic Ultrasound Aspiration Needle; Boston Scientific, Marlborough, Mass). The aspiration materials were used for cytopathological evaluation, histological examination, and K-ras mutation analysis by PCR-PHFA. Sections were stained with hematoxylin and eosin as well as immunohistochemical staining for p53 when necessary. Extraction of Cell-Free ctDNA in Serum
Blood samples were collected in tubes (BD Vacutainer blood collection tubes; Becton, Dickinson and Company, Franklin Lakes, NJ) and processed within 1 hour after collection. The samples were centrifuged at 3000 3 g at 4 C to separate the serum from the peripheral blood cells, and stored at 280 C. ctDNA in serum was extracted from Cancer
Month 00, 2015
K-ras in Blood With Pancreatic Cancer/Kinugasa et al
TABLE 1. Characteristics of Patients From the Development Set and the Validation Set Characteristic Total Age, y Sex
Male Female
Median CA 19-9 (range), U/mL Median DUPAN2 (range),U/mL Median SPAN-1 (range), U/mL Stage of disease
Initial therapy
I II III IV GEM S-1 None Surgery Others
Development Set
Validation Set
75 66 (range, 47-85) 54 21 335 (0.6-99,999) 684 (25-16,000) 176 (10-210,000) 0 2 5 68 47 1 3 11 13
66 71 (range, 38-85) 43 23 544 (0.6-68,190) 379 (25-16,000) 158 (10-20,000) 1 0 8 57 40 5 3 13 5
P
.019 .381 .568 .042 .909 .250
.171
Abbreviations: CA 19-9, carbohydrate antigen 19-9; GEM, gemcitabine; S-1, oral fluoropyrimidine.
aliquots (1 mL) of serum obtained from 5 mL of blood with the QIAamp Circulating Nucleic Acid Kit (Qiagen, Valencia, Calif) according to the manufacturer’s instructions. Digital PCR
We used droplet digital PCR (ddPCR) (QX200; BioRad, Hercules, Calif), which is one of several available digital PCR17 methods. The sample was partitioned into 20,000 droplets, with target and background (reference) DNA randomly distributed among these droplets and uniformly. We used 3 types of PrimePCR products (BioRad) (G12V, G12D, and G12R) for ddPCR (cat 1863115, 1863113, and 1863112) because G12V, G12D, and G12R encompass nearly all K-ras mutations in pancreatic cancer18 (our EUS-FNA samples revealed that only these 3 types of K-ras mutations were detected in cases with K-ras mutations). Reactions were performed in 20 mL of reaction volume, which consisted of up to 130 ng of extracted DNA (5 mL), 2 3 droplet PCR supermix (10 mL), 20 3 target primer/probe mix (FAM) (1 mL), 20 3 reference primer/probe mix (HEX) (1 mL), and deionized distilled water (3 mL). Emulsified PCR reactions were run on a 96-well plate on a C1000 Touch thermal cycler incubating plate (Bio-Rad) at 95 C for 10 minutes followed by 40 cycles of incubation at 95 C for 15 seconds and 60 C for 60 seconds, followed by a 10-minute incubation at 98 C. Plates were read on a Bio-Rad QX200 droplet reader using QuantaSoft version 1.4.0 software provided by Bio-Rad to assess the numbers of droplets that were positive for wild-type K-ras and/or K-ras mutations. Digital PCR assay was performed on ctDNA in serum from 75 patients with pancreatic cancer in a development cohort and 66 patients with pancreatic canCancer
Month 00, 2015
cer in an independent validation cohort. All samples were analyzed in duplicate and variations were 0.001%.19 We confirmed that the control without serum ctDNA demonstrated no positive signal. Statistical Analysis
Overall survival was measured from the day of diagnosis by EUS-FNA to the date of death or the last follow-up examination. Survival curves were calculated using the Kaplan-Meier method. The log-rank test was used to compare survival curves. Univariate and multivariate analyses were used to examine potential prognostic factors and a Cox proportional hazards model was used to determine the most significant variables contributing to survival. Differences with a P value of
