Wang, Liang

Abstract Background: Circulating cell-free DNA (cfDNA) analysis is emerging as a less invasive approach to assessing tumor genomic alterations in cancer patients. Although high concordance has been reported between tumor tissue NGS and cfDNA in studies investigating specific genetic alterations, the fidelity of cfDNA to tumor tissue DNA in the global genomic scale is largely unknown. In a correlative study of a prospective clinical trial (NCT# 01953640) conducted in mCRPC stage patients treated with abiraterone acetate/prednisone (A/P), we evaluated correlation of genomic CNA in tumor DNA obtained from biopsy of metastatic lesions and matched plasma cfDNA. Methods: mCRPC patients (pts) underwent two image-guided core biopsies of metastases prior to initiation of A/P (visit 1, V1) and 12 weeks after treatment (visit 2, V2). At both time points plasma was obtained at the time of the core biopsies and cfDNA was extracted using DNA Blood Mini Kit (Qiagen, Valencia, CA). High coverage (for tumor tissue) and low coverage (for cfDNA) whole genome sequencing reads were first mapped to the human genome hg19. Read counts (RC) from the mapped sequence files were then binned into 1Mb windows. The RC ratio in each genomic bin was calculated by comparing tumor tissue DNA to lymphocyte gDNA derived from the same patient, and was further log2 transformed, corrected for GC content, and normalized by CGHnormaliter. The fully normalized log2 ratios data was subjected to segmentation using DNAcopy algorithm. Results: Between 05/2013 and 09/2015, 92 patients (pts) were enrolled of which tissue and plasma NGS data both visits was available for 18 pts. The correlation of CNAs between tumor tissue and its cfDNA counterpart ranges from 0.013 to 0.83 for V1 samples, and -0.05 to 0.9 for V2 samples. The decreased correlation in some pairs of samples is largely due to low tumor content and heterogeneity in the cfDNA. Although there is a wide range of correlation, commonly shared CNAs were identified in multiple chromosomal regions, including loss in 8p, gain in 8q and chromosome 5 and X. On the other hand, several genomic regions show inconsistent CNAs between tumor tissue and cfDNA among 15 pairs of samples in which cfDNA have large tumor content. These include: loss in 6q, 1p, 2p, 9q, 11q, 13p, 15p, 20p and gain in 8q and 10p in cfDNA but not in tumor tissue; marked loss in chromosome 22 and gain in p arm of X chromosome in tumor tissue but much less evident in cfDNA. Conclusion: High concordance of CNAs between cfDNA and tumor tissue DNA can be achieved given sufficient tumor content of cfDNA. However, more CNAs can be identified in cfDNA than its tumor tissue counterpart. Our result suggests that cfDNA NGS is a useful tool to investigate clonal evolution associated with cancer progression. Citation Format: Chiang-Ching Huang, Meijun Du, Liguo Wang, Yen-chen Lin, Hua Huang, Liewei Wang, Liang Wang, Manish Kohli. Comparison of copy number aberrations (CNAs) between plasma cell free DNA (cfDNA) and tissue DNA in metastatic castrate resistant prostate cancer (mCRPC) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 2733. doi:10.1158/1538-7445.AM2017-2733

To evaluate the clinical utility of cell‐free DNA (cfDNA), we performed whole‐genome sequencing to systematically examine plasma cfDNA copy number variations (CNVs) in a cohort of patients with colorectal cancer (CRC, n = 80), polyps (n = 20), and healthy controls (n = 35). We initially compared cfDNA yield in 20 paired serum–plasma samples and observed significantly higher cfDNA concentration in serum (median = 81.20 ng, range 7.18–500 ng·mL−1) than in plasma (median = 5.09 ng, range 3.76–62.8 ng·mL−1) (P < 0.0001). However, tumor‐derived cfDNA content was significantly lower in serum than in matched plasma samples tested. With ~10 million reads per sample, the sequencing‐based copy number analysis showed common CNVs in multiple chromosomal regions, including amplifications on 1q, 8q, and 5q and deletions on 1p, 4q, 8p, 17p, 18q, and 22q. Copy number changes were also evident in genes critical to the cell cycle, DNA repair, and WNT signaling pathways. To evaluate whether cumulative copy number changes were associated with tumor stages, we calculated plasma genomic abnormality in colon cancer (PGA‐C) score by summing the most significant CNVs. The PGA‐C score showed predictive performance with an area under the curve from 0.54 to 0.84 for CRC stages I‐IV. Locus‐specific copy number analysis identified nine genomic regions where CNVs were significantly associated with survival in stage III‐IV CRC patients. A multivariate model using six of nine genomic regions demonstrated a significant association of high‐risk score with shorter survival (HR = 5.33, 95% CI = 6.76–94.44, P < 0.0001). Our study demonstrates the importance of using plasma (rather than serum) to test tumor‐related genomic variations. Plasma cfDNA‐based tests can capture tumor‐specific genetic changes and may provide a measurable classifier for assessing clinical outcomes in advanced CRC patients.

152 Background: Plasma cfDNA ARcopy number variations (CNV) have been suggested to have prognostic significance in mCRPC stage. We compared plasma AR amplification ( ARamp) with CTC counts (the only FDA approved prognostic marker in mCRPC stage) in a prospective trial of patients (pts) progressing on androgen deprivation therapy. Methods: Plasma and CTC counts were obtained concurrently as part of a prospective clinical trial in mCRPC stage pts initiating abiraterone acetate (NCT# 01953640). Plasma cfDNA was purified using DNA Blood Mini Kit (Qiagen, Valencia, CA) and AR CNVs were detected using QuantStudio3D digital PCR system (Life Technologies, Carlsbad, CA, USA). CTC enumeration was performed using the FDA cleared CellSearch platform. The distributions of overall survival (OS) time based on ARamp status and CTC counts ( < 5 vs ≥ 5 cells) were estimated using the Kaplan-Meier method. Log-rank tests were used to test association of ARamp status and CTC counts with overall survival (OS). Univariate and multivariate logistic regression models and ROC evaluated association of OS at 24 months by ARamp status and CTCs ≥ 5. Results: Between 05/2013 and 09/2015, 92 pts were enrolled of which complete CTC, plasma and outcome data was available for 70 pts. Median age of the cohort (range) was 72 yrs (39-91); Median PSA–16.2 ng/ml (range: 0.9-2026.0); Median CTC count was 2.0 (range: 0-372); 58/70 pts had bone metastases; 38/70 had high volume and 32/70 had low volume metastatic disease at the time of sample collection. The median (range) follow up time at the time of this analysis was: 26.5 (7.9-37.0) months at which time 28/70 patients had died. ARamp was present in 19/70 pts while 51/70 did not have amplification. 24/70 had CTCs ≥ 5. A significant association of increased ARamp status with OS (p < 0.0001, log-rank test; AUC: 0.66) as well as with CTC counts ≥ 5 (p = 0.001, log-rank test; AUC: 0.67) was observed. Combining CTC status and plasma ARamp status the AUC was 0.75. Conclusions: Plasma cfDNA ARamp status and CTC counts have comparable and clinically meaningful prognostic marker value in mCRPC stage patients.

260 Background: The purpose of this study was to evaluate cell free nucleic acid based biomarkers for prognosis in mCRPC stage. Methods: Uniformly collected blood from mCRPC patients was processed for plasma which was used for exosomal RNA extraction. RNA sequencing was applied to examine microRNA (miRNA) profiles initially in a discovery cohort of 23 mCRPC patients. The sequences from fastq files were mapped to human miRBase (release 19) for known miRNA read counts. A software package (miRDeep2) was used for novel miRNA prediction. Cox regression was used for association of candidate miRNAs with overall survival (OS), defined as time from development of mCRPC to death or last follow up after adjusting for Gleason Score (GS) and time on androgen deprivation (AD) for hormone sensitive disease. miRNAs significantly associated with OS (FDR<0.05) were then validated in an independent cohort of 100 mCRPC subjects using univariate Cox regression. Results: RNA sequencing generated >10 million mappable reads per specimen in the discovery cohort. Among these reads ~50% were known miRNAs which included 1237 unique RNA sequences. Due to low abundance of most RNA transcripts, only miRNAs (n=250) with >/= 32 reads (on average) were analyzed. Four mRNAs (miR-1290, -1246, -375 and a predicted miRNA on chr. 12) were associated with OS(FDR<0.05). These were evaluated as prognotic markers in the independent cohort (N=100) using real time RT-PCR. Median time on AD for this cohort prior to progression to mCRPC was 35.7 (range: 3.5-136) months; median GS was 7; median follow up after mCRPC development was 12.5 (1-41) months at which time 28 patients had died. Elevated expression for miR-1290 and -375 were associated with shorter OS after adjusting for GS and AD time with a Hazard Ratio (HR) for miR-375 of 1.19 (CI: 1.006-1.18); (p=0.04) and a HR for miR1290 of 1.58 (CI:1.01-2.4); (p=0.04). A stronger association with shorter OS was observed for elevated expression of both miRs-HR of 2.02 (CI: 1.1-3.6) (p=0.01). Conclusions: Blood-based nucleic acid assays are easily accessible for biomarker development in mCRPC. Results from this study suggest specific exosomal miRs are of prognostic value in mCRPC and should be validated in larger cohorts.

Abstract Androgen deprivation therapy (ADT) for hormone-sensitive prostate cancer (HSPC) and chemotherapy for castration-resistant prostate cancer (CRPC) are considered standard of care. However, there are no clinical tests or factors that can reliably predict treatment responses for the advanced prostate cancer patients. Examination of tumor component in circulation, also known as liquid biopsy, has shown promise for predicting treatment outcomes and survival. To capture the tumor-derived genomic alterations and decrease the surrogacy of using PSA measurements as is currently performed, we propose a Plasma Genomic Abnormality (PGA) score and Treatment Efficacy (TEff) index based on copy number variations as estimated by cell free DNA (cfDNA) in plasma. We first performed whole genome and targeted sequencing (NimbleGen comprehensive cancer panel) to examine plasma cfDNA for tumor-associated genetic and genomic aberrations in 10 HSPC patients receiving ADT and 10 CRPC patients receiving chemotherapy. For each patient, we tested plasma from two time points: pre-treatment and 4 months post-treatment. We then calculated the PGA score by summing the most significant genomic abnormalities with higher PGA score indicating greater tumor DNA content in cfDNA. TEff index was derived from comparison of PGA score difference between treatments with higher TEff index reflecting a better treatment response. The sequencing-based copy number analysis revealed locus-specific gains or losses including those previously reported in prostate cancers, such as 8q gains, AR amplifications, PTEN losses and TMPRSS2-ERG fusions. We found that overall PGA score was significantly higher in CRPC patients than in HSPC patients (p&lt;0.0005), possibly reflecting increased tumor burden during the disease progression. Patients with low TEff index often demonstrated poor response to treatments, and patients with high initial PGA score or low TEff index tended to have poor overall survival. As examples of the PGA-based TEff associations with outcome we presented three CRPC patients including two with neuro-endocrine origin (small cell carcinoma). Comparison of mutational profiles between pre- and post-treatments showed that treatments caused more diverse gene mutations, in particular, in the genes involving androgen biosynthesis, AR activation, DNA repair and chemotherapy resistance, suggesting subclonal evolution of heterogeneous tumor genomes in response to these treatments. Our results strongly support the feasibility of using non-invasive liquid biopsy as a potential powerful tool to study biological mechanisms underlying therapy-specific resistance and monitor disease progression in the advanced prostate cancer. Further studies are needed to develop and validate the genomics-based PGA score and TEff index for clinical applications. Citation Format: Liang Wang, Shu Xia, Meijun Du, Rachel Ditmar, Tiezheng Yuan, Yongchen Guo, Yuan Wang, Adam Lee, Michael Tschannen, Elizabeth Worthey, Howard Jacobs, Chiang-Ching Huang, Manish Kohli. Plasma genetic and genomic abnormalities predict treatment response and clinical outcome in advanced prostate cancer. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 5231. doi:10.1158/1538-7445.AM2015-5231

Abstract PURPOSE: Cell free circulating microvesicles (in particular, exosomes) contain proteins and nuclei acids that may serve as biomarkers for disease diagnosis and prognosis. The goal of this study was to identify exosomal microRNAs that may predict clinical outcome of patients with castrate resistant prostate cancer (CRPC). EXPERIMENTAL DESIGN: In discovery stage, we performed RNA sequencing using plasma exosomal RNAs derived from 36 patients with CRPC. We mapped the sequence reads to known miRbase (Release 19, 2043 entries) and normalized the miRNA abundance by sequence counts per million mappable reads. We applied Cox regression analysis to identify the miRNAs that were associated with overall survival of CRPC. In validation stage, we first examined an exosomal small RNA sequencing dataset consisting of 192 individuals with various health conditions. We used Normfinder and Bestkeeper to select most stably expressed miRNAs as candidates for normalization references. We then applied real-time qRT-PCR assays to test selected candidate reference miRNAs and survival-related miRNAs in additional 100 CRPC patients. From validated miRNAs, we constructed multivariate models to predict overall survival for this group of patients. RESULTS RNA sequencing generated over 6 million mappable reads per patient in the initial discovery cohort. Among these reads, ∼45% were mapped to known miRNAs for a total of 483 known and 275 novel miRNAs with normalized read counts ≥ 5. The median follow-up time for the 36 CRPC patients was 35.6 months during which 10 patients had died. Cox regression analysis identified four microRNAs (miR-1290, -1246, -375 and a predicted miRNA at chromosome 12) that were associated with overall survival (FDR&lt;0.05). Of six selected candidate miRNAs, miR-30a/e-5p performed the best in their stability as endogenous references for real-time qRT-PCR. When normalized by these internal controls, we confirmed two of the three known miRs (miR-1290 and -375) showing significant association with overall survival in the validation cohort (p&lt;0.001) with a median follow up time of 12.5 (range: 1-41) months at which time 28 patients had died. The median overall survival for patients with high (N=21) and low (N=40) expression in both miR-1290 and -375 was 7.2 (1-26.7) months and 15.3 (1-41) months, respectively (p&lt;0.05). The miR-1290/-375-based prediction model consistently showed better predictive performance with AUC=72% than clinical variable-based model with AUC=64%. CONCLUSIONS: Plasma exosomal miRNAs provide an easily accessible resource for biomarker development in prognosis of advanced prostate cancer. Exosomal miR-1290 and miR-375 are associated with overall survival in CRPC patients. miR-30a/e-5p, especially the geometric mean of miR-30a/e-5p, are qualified endogenous references for real-time qPCR. Further confirmation of these findings is needed for prognostic and predictive biomarker in development of CRPC stage. Citation Format: Xiaoyi Huang, Tiezheng Yuan, Meihua Liang, Meijun Du, Shu Xia, Rachel Louise Dittmar, Zhifu Sun, Yan Lu, Stephen N. Thibodeau, Lisa Boardman, Manish Kohli, Liang Wang. Exosomal miR-1290 and miR-375 as prognostic markers in metastatic castrate resistant prostate cancer. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 574. doi:10.1158/1538-7445.AM2014-574