Quantitative Detection Of Circulating Tumor DNA By Targeted Exon Deep Sequencing And Its Application In Monitoring Metastatic Breast Cancer

Breast cancer is the most common cancer and the leading cause of cancer related death in women in both developed and developing countries. Among them, metastatic breast cancer remains an incurable disease but is treatable by means of serial administration of endocrine, cytotoxic, or biologic therapies. Thus, the monitoring of treatment response is essential to avoid continuing ineffective therapies, to prevent unnecessary side effects, and to determine the benefit of new therapeutics.Noninvasive access circulating tumor bio-marker is particularly attractive for solid tumors, which cannot be repeatedly sampled without invasive procedures. Thus, "liquid biopsy" has been considered as having the potential to revolutionize detection and monitoring of metastatic breast cancer. Bio-markers such as cancer antigen 15-3 (CA 15-3) and circulating tumor cells have been widely studied. Circulating cell-free DNA carrying tumor-specific alterations (circulating tumor DNA) is considered as another promising bio-marker for noninvasive assessment of cancer burden.Part Ⅰ Data Mining Technology for the Target Identification and Evaluation of the Breast Cancer Related ExonsBackground:For most cancer types (of course including breast cancer), genie regions with cancer related somatic mutations just account for only a small fraction of the whole genome, which indicating that only a limited number of gene regions are involved in the occurrence and development of certain type of cancer. Thus, we just need to focus on these set of regions related to breast cancer to monitor tumor burden changes in the course of treatment. This fact greatly underscores the urgent need to develop an algorithm to maximize the number of breast cancer related genes while minimizing the target region size.Objective:This section intends to select target regions highly associated with breast cancer by using of bioinformatics and data mining technologies. So that we can provide a target for subsequent tumor type targeted sequencing method, reduce blindness, and make the detection technology achieve broad patient coverage at a reasonable cost.Methods:Seven hundred and seventy-six samples were picked out randomly from the whole-exome sequencing (WES) data of a total of 963 patients with breast cancer profiled by The Cancer Genome Atlas (TCGA) as train-set and another 187 samples were selected as test-set. To design a selector for the targeted regions we began by including exons covering recurrent mutations in potential driver genes from the Catalogue of Somatic Mutations in Cancer (COSMIC) and previous published papers. Next, using whole-exome sequencing (WES) data from 776 train-sets, we applied an iterative algorithm to maximize the number of missense mutations per patient while minimizing selector size. Finally, to evaluate and validate the sensitivity and coverage of the targeted parts of the breast cancer genome, we examined the selector region in WES data from the independent cohort of 187 test-sets and independent samples of 40 cases with breast cancer.Results:Collectively, the breast cancer related genie regions selector design targets 961 exons from 834 recurrently mutated genes, in total covering 23982049 single nucleotide variants (SNVs),134 indels, and 52 fusions within a length of 118.24 kb. According to biological process classification, all 834 breast cancer related genes were divided into 1348 GO-BP categories, and the five most common processes were cellular component organization or biogenesis, cellular component organization, collagen catabolic process, single-organism developmental process, and multicellular organismal catabolic process, respectively. According to cellar components classification, all 834 breast cancer related genes were involved in 239 cellar components, and the five most common components included cell projection, cytosol, endoplasmic reticulum lumen, membrane region, and cytoplasm. According to molecular function classification, all 834 breast cancer related genes joined in 269 molecular functions, and the five most common functions were ATP binding, adenyl ribonucleotide binding, adenyl nucleotide binding, anion binding, and extracellular matrix structural constituent, respectively. According to KEGG-signaling pathway annotation, all 834 breast cancer related genes were involved in 73 signaling pathways, and the five most common pathways were focal adhesion, amoebiasis, GnRH-signaling pathway, small-cell lung cancer, and endometrial cancer, respectively. Within this small target (3.7/100000 of the human genome), the selector identifies a median of two SNVs and covers 100% of the training sets with breast cancer. To validate the sensitivity and coverage, we examined the selector region in WES data from the testing patients and an independent cohort of 40 patients with breast cancer. The selector covered 88.7% and 100% of patients with a median of two and five SNVs per patient, respectively.Conclusions:This selector is extremely sensitive and specific and is able to apply to the vast majority of patients with breast cancer, without the requirement for patient-specific optimization.Part Ⅱ Circulating Cell-Free DNA Copy Number as a Prognostic and Predictive Indicator for Monitoring Metastatic Breast Cancer PatientsBackground:For patients with metastatic breast cancer, successful extension of survival depends mainly on monitoring of the tumor burden to determine the response to treatment and timely action to change to the more efficient treatments. Currently, tumor response are mainly be assessed by the use of imaging (X-ray mammography, MR mammography, and CT) and cancer antigen. However, these conventional tumor markers can not readily reflect functional or metabolic changes that may occur with targeted chemotherapy or even with conventional chemotherapy due to lower sensitivity and specificity and longer half-life. Analysis of circulating cell-free DNA has the potential to revolutionize detection and monitoring of tumors. Noninvasive access to circulating cell-free DNA is particularly attractive for solid tumors, which cannot be repeatedly sampled without invasive procedures.Objective:The main specific objectives are as follows:(ⅰ) to explore the positivity, distribution, and change patterns between pre- and post-chemotherapy of the plasma circulating cell-free DNA levels for patients with metastatic breast cancer, and (ⅱ) to evaluate the prognostic potential of circulating cell-free plasma DNA quantification in the patients with metastatic breast cancer.Methods:A prospective cohort study was designed to include 40 breast cancer patients with measurable metastatic lesions, to be followed for 150 days. The patients were planned at different stages:pre-treatment, after the first to fifth course of chemotherapy, and 150 days after treatment. Here we introduce digital PCR, an ultrasensitive method for quantifying the levels of circulating cell-free DNA, using an 82-bp amplicon on Beta-actin gene and a dilution series of intact male human genomic DNA as a standard curve. Area under curve (AUC) of receiver operator characteristic curve (ROC) was calculated to evaluate the predictive efficacy. The relationship between the circulating cell-free plasma DNA copy and the clinico-pathological features and prognosis of breast cancer patients was investigated.Results:The median follow-up was 106 days (range:67-113 days). The 150-day overall survival (OS) and progression-free survival (PFS) rates for all patients were 100% and 50.2%, respectively. The median circulating cell-free DNA genome equivalents (GEs) in the plasma for all patients pre-and post-treatment were 1015 (749.3-1315.3) GE/mL and 916.5(519.3-1124.3) GE/mL, respectively (Paired Wilcoxon test, Z=-1.74, P=0.08). For the variation in circulating cell-free DNA levels from pre-treatment to the fifth course of chemotherapy follow-up assessments, non-parametric Friedman test for the change did not produce any significance at the usual 5% level (χ2=5.26, P=0.39). However, no significant difference of circulating cell-free plasma DNA levels could be found between subgroups by demographic and clinical characteristics (all P value from Mann-Whitney U test>0.05). With cutoff points of 9386 GE/mL and 812 GE/mL, sensitivity and specificity of circulating cell-free plasma DNA copy of before and after the first course of treatment in distinguish progressed patients from other cases (including complete remission (CR), partial remission (PR), and stable disease (SD)) were 70.0% and 66.7%,80.0% and 54.2%, respectively, and AUC of ROC were 0.66 (95%CI:0.48-0.82, z=1.65, P=0.09) and 0.60 (95%CI:0.42-0.76, z=0.91, P=0.37). Then the 40 breast cancer patients were divided into high plasma circulating cell-free DNA copy group (>9386 GE/mL) and low DNA copy group (<9386 GE/mL). The 150-day PFS for patients with<9386 GE/mL of the median circulating cell-free plasma DNA levels in the plasma after the first course of treatment was superior to those with more than 9386 GE/mL (126 days vs.110 days,44.7% vs.40.9%, Log-rank (Mantel-Cox) test χ2=4.05, P=0.04). In multivariate Cox regression analysis, higher plasma DNA copy group (> 9386 GE/mL) was independently associated with decreased PFS (HR=13.27,95%CI:1.14-154.16) compared to that of≤9386 GE/mL group.Conclusions:These results suggest that no significant relationship was found between the plasma circulating cell-free DNA copy and the clinicopathological features. However, high circulating cell-free plasma DNA level was an independent predictor for the short term disease progression of breast cancer.Part Ⅲ Quantitative Detection of Circulating Tumor DNA by Targeted Exon Deep Sequencing and Its Treatment Monitoring Value in Breast CancerBackground:With the concept of "Liquid Biopsy" more and more popular, from conventional circulating biomarker (such as cancer antigen) to circulating tumor cells (CTCs), recently more and more attention has been focused on another circulating biomarker-circulating tumor DNA (ctDNA). ctDNA is a promising biomarker for noninvasive assessment of cancer burden, but existing circulating tumor DNA detection methods have insufficient sensitivity or patient coverage for broad clinical applicability. Identification of circulating tumor DNA through detecting known point mutations in genes in plasma DNA is the most commonly used strategy, but the majority of patients lack mutations in these genes.Objective:To overcome these limitations, we will explore a new strategy for analysis of circulating tumor DNA on the basis of the first part and the second part work of present study. The main specific objectives of this part are as follows:(ⅰ) to explore the positivity, distribution, and change patterns between pre-and post-chemotherapy of the plasma ctDNA for patients with metastatic breast cancer, (ⅱ) to provide a direct comparison of prognostic value between circulating tumor DNA and other circulating biomarkers (CA 15-3 and circulating total cell-free DNA) in predicting short-term prognosis of metastatic breast cancer patients, and (ⅲ) to evaluate the possibility of circulating tumor DNA quantification for the noninvasive monitoring of metastatic breast cancer.Methods:We used targeted sequencing to identify somatic genomic alterations to identify circulating tumor DNA and quantify its fraction in the circulating total cell-free DNA in serially collected plasma specimens. Then ctDNA copy number of metastatic breast cancer patients before and after treatment were calculated together with the peripheral circulating total cell-free DNA measured in the "Part II". Area under curve (AUC) of receiver operator characteristic curve (ROC) was calculated to evaluate the predictive efficacy of all related circulating biomarkers. Besides the circulating total cell-free DNA in the "Part II", the relationship between the circulating tumor DNA and CA 15-3 and the clinicopathological features and prognosis of breast cancer patients was also investigated.Results:The median circulating tumor DNA levels in the plasma for all patients pre-and post-treatment (after the first course of treatment) were 136.1 (60.5-238.5) GE/mL and 136.3 (52.4-239.4) GE/mL, respectively, and the corresponding fractions in the circulating total cell-free DNA were 1.28%(0.56%-2.57%) and 1.56% (0.80%-3.07%) (Paired Wilcoxon test, Z=-0.17, P=0.86). For the variation between different clinicopathological characteristics in circulating tumor DNA levels from pre-treatment to the first course of chemotherapy follow-up assessments, non-parametric Mann-Whitney U test demonstrated that statistical significance between Her2 positive group and Her2 negative group (114.1 GE/mL vs.224.2 GE/mL) in the pre-treatment circulating tumor DNA level, and between group with diabetes and that without diabetes (357.0 GE/mL vs.125.6 GE/mL) at the usual 5% level. Circulating tumor DNA was successfully detected in 40 of the 40 women (100%) in whom somatic genomic alterations were identified; CA 15-3 (>25U/mL) and circulating total cell-free DNA (>9386 GE/mL) were detected in 15 of 32 women (46.9%) and 28 of 40 women (70%), respectively. The levels of plasma circulating tumor DNA in disease progressed group (PD) at before and 1 course after treatment were 223.3 GE/mL and 259.0 GE/mL, respectively. Both were significantly higher than the other group (including complete remission (CR), partial remission (PR), and stable disease (SD)) (before:111.4 GE/mL,1 course after treatment:86.1 GE/mL) (all P value<0.05). However, there was no statistically significant difference of CA 15-3 and circulating cell-free DNA between the PD group and the CR/PR/SD group at both before and 1 course after the treatment (all P value>0.05). With cutoff points of 135.21 GE/mL and 183.34 GE/mL, sensitivity and specificity of circulating cell-free plasma DNA copy of before and after the first course of treatment in distinguish PD patients from CR/PR/SD cases were 90.0% and 58.3%,90.0% and 79.2%, respectively, and AUC of ROC were 0.74 (95%CI:0.56-0.87, z=2.33, P=0.02) and 0.86 (95%CI:0.70-0.95, z=4.51, P=0.0001). Then the 40 breast cancer patients were divided into high plasma circulating tumor DNA copy group (> 135.21 GE/mL) and low DNA copy group (<135.21 GE/mL) according to the circulating tumor DNA levels before the treatment. The 150-day PFS for patients with<135.21 GE/mL was superior to those with more than 135.21 GE/mL (92.3% and 27.1%, Log-rank (Mantel-Cox) test x2=4.95, P=0.03). Correspondingly, patients were also divided into high circulating tumor DNA copy group (> 183.34 GE/mL) and low DNA copy group (≤183.34 GE/mL) according to the circulating tumor DNA concentration after the first course treatment. The corresponding PFS were 88.8% and 18.8%, respectively, Log-rank (Mantel-Cox) test χ2=8.98, P=0.002). In multivariate Cox regression analysis, higher plasma DNA GE(>183.34 GE/mL) was independently associated with decreased PFS (HR=39.02,95%CI:1.19-1284.67) compared to that of<183.34 GE/mL after 1 course of the treatment.Conclusions:This novel method based on targeted exon deep sequencing for detection of circulating tumor DNA in the blood of cancer patients has sufficient sensitivity or patient coverage for broad clinical applicability. This proof-of-concept analysis suggested that circulating tumor DNA is an informative, inherently specific, and highly sensitive biomarker of metastatic breast cancer. Higher circulating tumor DNA level (>183.34 GE/mL) after the first course of treatment and the increased circulating tumor DNA level compared to that at before treatment were independent predictors for the short term disease progression of metastatic breast cancer. After validated in larger studies, this methods for quantification of circulating tumor DNA could serve as a simple method to aid in breast cancer response monitoring.