Mahoney, Douglas W.

Early detection improves hepatocellular carcinoma (HCC) outcomes, but better noninvasive surveillance tools are needed. We aimed to identify and validate methylated DNA markers (MDMs) for HCC detection. Reduced representation bisulfite sequencing was performed on DNA extracted from 18 HCC and 35 control tissues. Candidate MDMs were confirmed by quantitative methylation‐specific PCR in DNA from independent tissues (74 HCC, 29 controls). A phase I plasma pilot incorporated quantitative allele‐specific real‐time target and signal amplification assays on independent plasma‐extracted DNA from 21 HCC cases and 30 controls with cirrhosis. A phase II plasma study was then performed in 95 HCC cases, 51 controls with cirrhosis, and 98 healthy controls using target enrichment long‐probe quantitative amplified signal (TELQAS) assays. Recursive partitioning identified best MDM combinations. The entire MDM panel was statistically cross‐validated by randomly splitting the data 2:1 for training and testing. Random forest (rForest) regression models performed on the training set predicted disease status in the testing set; median areas under the receiver operating characteristics curve (AUCs; and 95% confidence interval [CI]) were reported after 500 iterations. In phase II, a six‐marker MDM panel (homeobox A1 [HOXA1], empty spiracles homeobox 1 [EMX1], AK055957 , endothelin‐converting enzyme 1 [ECE1], phosphofructokinase [PFKP], and C‐type lectin domain containing 11A [CLEC11A]) normalized by beta‐1,3‐galactosyltransferase 6 (B3GALT6) level yielded a best‐fit AUC of 0.96 (95% CI, 0.93‐0.99) with HCC sensitivity of 95% (88%‐98%) at specificity of 92% (86%‐96%). The panel detected 3 of 4 (75%) stage 0, 39 of 42 (93%) stage A, 13 of 14 (93%) stage B, 28 of 28 (100%) stage C, and 7 of 7 (100%) stage D HCCs. The AUC value for alpha‐fetoprotein (AFP) was 0.80 (0.74‐0.87) compared to 0.94 (0.9‐0.97) for the cross‐validated MDM panel (P < 0.0001). Conclusion : MDMs identified in this study proved to accurately detect HCC by plasma testing. Further optimization and clinical testing of this promising approach are indicated.

Abstract Purpose: Gastric adenocarcinoma is the third most common cause of cancer mortality worldwide. Accurate and affordable noninvasive detection methods have potential value for screening and surveillance. Herein, we identify novel methylated DNA markers (MDM) for gastric adenocarcinoma, validate their discrimination for gastric adenocarcinoma in tissues from geographically separate cohorts, explore marker acquisition through the oncogenic cascade, and describe distributions of candidate MDMs in plasma from gastric adenocarcinoma cases and normal controls. Experimental Design: Following discovery by unbiased whole-methylome sequencing, candidate MDMs were validated by blinded methylation-specific PCR in archival case–control tissues from U.S. and South Korean patients. Top MDMs were then assayed by an analytically sensitive method (quantitative real-time allele-specific target and signal amplification) in a blinded pilot study on archival plasma from gastric adenocarcinoma cases and normal controls. Results: Whole-methylome discovery yielded novel and highly discriminant candidate MDMs. In tissue, a panel of candidate MDMs detected gastric adenocarcinoma in 92% to 100% of U.S. and South Korean cohorts at 100% specificity. Levels of most MDMs increased progressively from normal mucosa through metaplasia, adenoma, and gastric adenocarcinoma with variation in points of greatest marker acquisition. In plasma, a 3-marker panel (ELMO1, ZNF569, C13orf18) detected 86% (95% CI, 71–95) of gastric adenocarcinomas at 95% specificity. Conclusions: Novel MDMs appear to accurately discriminate gastric adenocarcinoma from normal controls in both tissue and plasma. The point of aberrant methylation during oncogenesis varies by MDM, which may have relevance to marker selection in clinical applications. Further exploration of these MDMs for gastric adenocarcinoma screening and surveillance is warranted. Clin Cancer Res; 24(22); 5724–34. ©2018 AACR.

Abstract Purpose: Lung cancer is the leading cause of cancer deaths worldwide. Most present symptomatically at late stage with high lethality. Early detection reduces mortality but accurate and readily accessible tools for population screening are limited. By whole methylome sequencing, we have identified novel methylated DNA markers (MDMs) for lung cancer in tissue (Giakoumopoulos et al. ASCO 2016). Using top candidate MDMs in the present study, we now explore their clinical accuracy for lung cancer detection when assayed from plasma. Experimental Procedures: Archival plasmas from two independent study groups were tested in blinded fashion. Lung cancer cases and controls (apparently healthy smokers) for each group were balanced on age and sex (Group 1: 64 cases, 231 controls; Group 2: 23 cases, 80 controls). Using multiplex PCR followed by QuARTS (Quantitative Allele-Specific Real-time Target and Signal amplification) assay, a post-bisulfite quantification of MDMs on DNA extracted from plasma was performed. We selected 31 MDM candidates for initial evaluation in Group 1 (1 ml plasma/patient); top individual MDMs were subsequently tested in Group 2 to identify optimal MDM panels for lung cancer detection (2 ml/patient). Results: From Group 1 analyses, 13 high performance MDMs were selected for further testing (CYP26C1, SOBP, SUCLG2, SHOX2, ZDHHC1, NFIX, FLJ45983, HOXA9, B3GALT6, ZNF781, SP9, BARX1, EMX1) with individual areas under the receiver operator curve (AUCs) ranging from 0.593 to 0.939. Discrimination by individual MDMs was corroborated in Group 2 in which data was analyzed using two methods: a logistic regression fit and a regression partition tree approach. The logistic fit model identified a 4-marker panel (ZNF781, BARX1, EMX1, and SOBP) with an AUC of 0.96 and an overall sensitivity of 91% and 90% specificity. Analysis of the data using a regression partition tree approach identified 4 markers (ZNF781, BARX1, EMX1, and HOXA9) with AUC of 0.93 and an overall sensitivity of 96% and specificity of 94%. For both approaches, B3GALT6 was used as a standardizing marker of total DNA input. Conclusion: A panel of MDMs assayed in plasma achieved high sensitivity and specificity for all types of lung cancer. Further clinical evaluation and validation of this promising panel in larger patient groups are clearly indicated. Citation Format: Hatim T. Allawi, Maria Giakoumopoulos, Evan Flietner, Austin Oliphant, Carla Volkmann, Brian Aizenstein, Tamara Sander, Drew Eckmayer, Ashley Poenitzsch Strong, Melissa Gray, Barry Berger, Tracy Yab, William Taylor, Douglas Mahoney, John B. Kisiel, David E. Midthun, David A. Ahlquist, Graham P. Lidgard. Detection of lung cancer by assay of novel methylated DNA markers in plasma [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 712. doi:10.1158/1538-7445.AM2017-712

11 Background: Currently there are no serum markers of response to AA. We performed a proteome based analysis of serum from prostate cancer patients, followed by mapping of candidate markers to networks based on known molecular interactions in Ingenuity Knowledge Base (IKB). Methods: Isobaric mass tags for relative and absolute quanititation (iTRAQ) analyzed by reverse-phase liquid chromatography coupled to tandem mass spectrometry (LC/MS/MS) was performed in serum of 3 non-localized prostate cancer cohorts. The first cohort included 15 paired untreated hormone-sensitive “pre AA” and 3-month “post AA” specimens; the second included 10 “early AA failure” (median failure time: 11 months) and the third included 10 “late AA failure” specimens (median failure time: 95 months). Differentially expressed candidate proteins were identified by comparisons of (i) paired pre/post AA proteomes and (ii) post AA proteome with the combined AA failure cohorts at a False Discovery Rate of 0.2. To facilitate biological interpretation of multiple candidates identified in the comparisons, IPA was used. Candidate markers implicated in IPA networks with statistical significance were pursued in a second, independent patient cohort for association with time to AA failure using univariate Cox regression analysis. Results: Median PSA for pre/post AA first cohort were 3.15 and 0.29 ng/ml. Median PSA for the second and third cohorts were 27.3 and 4.3 ng/ml. Between post AA and AA failure cohorts, 149 proteins were differentially expressed. Between early and late AA failure 98 proteins were differentially expressed; 47 proteins were common in both comparisons. Network enrichment analysis of the 47 proteins by IPA identified four interaction networks (p<0.01), one of which highlighted a role for 17-β-estradiol (E1). Gas chromatography used for measuring 3-month post AA initiation serum E1, estrone (E2) and testosterone levels (N=38) detected high E1, E2 levels associated with longer time to AA failure (P=0.07 for E1; P=0.08 for E2) in the independent cohort. Conclusions: A global proteomic analysis identified post AA initiation serum E1, E2 levels as potential response markers, which needs validation.

341 Background: Precursors to 1/3 of colorectal cancer (CRC), serrated polyps have historically been underdetected by screening due to their inconspicuous, nonhemorrhagic, and proximal nature. New generation sDNA-MT show high sensitivity for both CRC and advanced adenomas (Gastroenterology 2012;142:248), but screen detection of serrated polyps by this approach requires further validation. Aim: To assess and compare noninvasive detection of sessile serrated polyps >1cm (SSP) by sDNA-MT and a commercial quantitative FIT. Methods: In a blinded prospective study, a single stool sample used for both tests was collected from 456 asymptomatic adults prior to screening or surveillance colonoscopy (criterion standard). All 29 patients with SSP were included as cases and all 232 with no neoplastic findings as controls. Buffered stool samples were processed and frozen on receipt; Exact Sciences performed sDNA-MT in batches using optimized analytical methods (Gastroenterology 2012;142:S770). The sDNA-MT panel targets methylated BMP3 (mBMP3) and NDRG4, mutant KRAS, β-actin, and hemoglobin. FIT (Polymedco OC-FIT Check) was performed in separate lab <2 days post defecation and evaluated at cutoffs of 50 (FIT-50) and 100 ng/ml (FIT-100). Results: Age/sex distributions were similar in cases and controls: 40% vs 45% > 65 years and 51% vs 59% women, p=NS respectively. SSP median size was 1.2cm (range 1-3cm), 93% were proximal, and 64% had synchronous diminutive polyps. Among sDNA-MT markers, mBMP3 proved highly discriminant for SSP detection (AUC=0.87, p<0.00001); other DNA markers were less informative and provided no incremental sensitivity. Hemoglobin alone showed no discrimination (AUC=0.50, p=NS). At matched specificities, SSP detection by stool mBMP3 was significantly greater than by either FIT-50 or FIT-100 (Table). Conclusions: From a screening and surveillance setting, SSP can be detected noninvasively by stool assay of exfoliated DNA markers, especially mBMP3. FIT appears to have no value in SSP detection. [Table: see text]

255 Background: In a prior study, we found that 8 single nucleotide polymorphisms (SNPs) from 4 candidate hormone-related genes (JAK2, TRMT11, NKX3-1 and HSD17B12) were associated with overall survival (OS) in CRPC stage in a North American cohort of 519 subjects. We attempt to replicate these findings in an independent cohort of Portuguese patients. Methods: A hundred and forty CRPC stage patients in the new cohort were genotyped fort he candidate SNPs. The primary endpoint was overall survival (OS), defined as time from development of CRPC to death. For SNP level results we estimated hazard ratios (HR) and 95% confidence intervals (CI) under the dominant allele model using Cox regression and adjusted for age and Gleason score (GS). Results: The median age of the Portuguese cohort was 69 years (range 53-84). The GS distribution was 48% subjects with GS≥8; 35% with GS=7 and 17% with GS<7. Median time from castration resistance to death for the cohort was 2.2 years (IQ range: 0.8-3.4, 93 deaths). One of the 3 SNPs (rs2295005; C>G; MAF = 0.16) in the TRMT11gene was associated with OS after adjustment for age and GS: (dominant model p=0.05, HR=0.56; 95 CI=0.31-1.00). No other SNP was observed to have statistical significance with OS (Table 1). Conclusions: Variation in the TRMT11 gene is significantly associated with overall survival in patients with CRPC and warrants further validation as a potential prognostic biomarker. [Table: see text]

209 Background: Predictive markers of response to AA are not known. In a prior study, we found that 8 single nucleotide polymorphisms (SNPs) from 4 candidate hormone-related genes (JAK2, TRMT11, NKX3-1 and HSD17B12) were associated with response to AA in HSPC in a North American cohort of 519 subjects. We attempt to replicate these findings in a cohort of Portuguese patients. Methods: The Portuguese cohort consisted of 308 HSPC patients, and the primary endpoint was time to AA failure, defined as time from initiating continuous AA to two serial PSA increases over the nadir, initiation of chemotherapy for progressive disease, or clinical or imaging based progression, whichever came first. For each SNP, we calculated hazard ratios (HRs) and 95% confidence intervals (CI) for time to AA failure using a recessive model with adjustment for Gleason score (GS). Results: The median age of the Portuguese cohort was 69 years (range 43, 85). The GS distribution was 35% subjects with GS≥8; 35% with GS=7 and 30% with GS<7. Median time to AA failure for the cohort was 1.8 years (IQ range: 0.3, 3.6). Association of each SNP with time to AA failure is shown in the Table. None of the SNPs replicated. However, there was a suggestive inverse association for rs10503733 (NKX3.1) with time to AA failure (P=0.07). Conclusions: While we were not able to replicate our initial findings, there was an unexpected association with rs10503733 (NKX3.1 gene) that may warrant additional clinical validation. [Table: see text]

313 Background: We evaluated variation in candidate genes implicated in either initiation or progression of prostate cancer with survival in CRPC stage. Methods: Germline DNA was extracted from peripheral blood mononuclear cells of CRPC patients enrolled in a clinically annotated registry. Fourteen candidate genes were tagged using single nucleotide polymorphisms (SNPs) from HapMap with minor allele frequency of > 5%. The primary endpoint was overall survival (OS), defined as time from development of CRPC to death. Principal component analysis was used for gene levels tests of significance. For SNP level results the per allele hazard ratios (HR) and 95% confidence intervals (CI) under the additive allele model using Cox regression adjusted for age at CRCP and Gleason score (GS) were used. Results: Two hundred and forty five CRPC patients who met the criteria of having adequate DNA were genotyped (14 genes, 84 SNPs). The median age of the cohort was 69 years (range 43-93). The GS distribution was 55% with GS≥8, 32% with GS = 7 and 13% with GS < 7 or unknown. Median time from castration resistance to death for the cohort was 2.67 years (IQ range: 1.6-4.07, 144 deaths). At the gene level JAK2 was detected to be associated with OS. Six of 18 JAK2 SNPs were associated with OS after adjustment for age and GS as well as pruning of SNPs in high linkage disequilibrium with each other (see table). In our multivariate model including age, GS, rs2149556, and rs4372063, the adjusted HRs for rs2149556 and rs4372063 were 0.67 (95% CI 0.38-1.18) and 2.17 (95% CI 1.25-3.76), respectively. The protective effect of rs2149556 appears after adjusting for the presence of minor alleles for rs4372063 in this dataset. Conclusions: Germline variation in the JAK2gene is associated with survival in CRPC stage and warrants further validation as a potential prognostic biomarker. [Table: see text]