Search

3050 Background: MSK-IMPACT (Integrated Mutation Profiling of Actionable Cancer Targets), is a high throughput, targeted-DNA-sequencing panel for somatic mutations created by the Department of Pathology at Memorial Sloan Kettering Cancer Center (MSK). The MSK-IMPACT is FDA approved for tumor tissue profiling to guide treatment selection. Recognizing access to tumor tissue in many cancers is difficult and may harbor inter & intra lesional heterogeneity, we sought to evaluate concordance of sequencing single CTCs vs. paired biopsy analyzed by MSK-IMPACT, to assess CTC vs. tumor clonality, and their relationship to outcomes. Methods: 148 biopsy samples from 138 mCRPC pts were submitted for IMPACT analysis and a blood draw within 30-day prior to initiation of a new line of treatment. Blood samples were sent to Epic Sciences for CTC detection. The detected CTCs underwent single cell low pass whole genome sequencing for copy number variation (CNV). For each set of matched samples, DNA copy number profiles from IMPACT and CTC sequencing were compared for similarity. Results: Of the 114 successfully sequenced samples 58 were from lymph nodes, 23 from bone, 25 from liver or lung, and 8 from other soft tissue. Of the 111 patients with CTCs analyzed, a total of 1073 CTCs were sequenced (range 1-27, median = 4 per pt). Of patients with both IMPACT & CTC, >80% pts had multiple subclones with distinguishable CNV. Concordance of genomics clones was found in 63%, and discordant in 37% cases. The clonal concordance between tissue biopsy and CTC was higher when the biopsy was obtained from bone marrow or a visceral site (71% concordance) than from a lymph node (53% concordance). Conclusions: Single CTC sequencing is often concordant to metastatic tissue, but unique CTC clones and the presence of multi-clonal disease highlight the potential to be underrepresented through tissue biopsy, especially when taken from the lymph node.[Table: see text]

5053Background: Biomarkers to predict outcomes for individual patients (pts) on standard of care Rx is an unmet medical need in the treatment of mCRPC. Using baseline (BL) blood draws, we previousl...

273 Background: A previous analysis of 161 patients (pts) tested for nuclear-localized AR-V7(+) CTCs showed a therapy interaction between AR-V7 positivity and improved overall survival (OS) on taxane chemotherapy vs. androgen receptor signaling inhibitors (ARSI). To validate the use of the biomarker result for physician decision making, we prospectively analyzed an independent, multicenter, blinded, cross-sectional cohort (n = 225) to confirm a therapy interaction with AR-V7. We corrected for possible pt selection bias by the treating physician by analyzing the association of therapy to OS in low and high risk groups defined by the test cohort. Methods: Two analyses were performed: (1) the validation of a therapy interaction between AR-V7 positivity and superior OS benefit for pts treated with taxanes in the context of use for 2nd+ line pts; and (2) as the choice between ARSI or taxanes was at the discretion of the attending physician, pt risk was incorporated into the predictive biomarker assessment. A pt-specific risk score was developed from line of therapy and other covariates to stratify pts as low and high risk, and the association of AR-V7 status and OS was performed within each risk group to correct for physician decision making and address possible confounding with treatment assignment. Results: (1) A therapy interaction between AR-V7(+) pts and lower risk of death on taxanes vs. ARSI (HR: 0.23, p = 0.003, 95% CI: 0.09 – 0.61) was validated. (2) In the validation cohort, high risk AR-V7(+) pts had an OS benefit when treated with taxanes (p = 0.02) and the AR-V7(-) pts had an improved OS with ARSI therapy (p = 0.02). AR-V7 incidence was low in the low risk pts, precluding the analysis for this sub-cohort. Conclusions: The results validate that the nuclear-localized AR-V7(+) biomarker has an interaction with therapy and improved survival on taxanes. Further, when adjusting for pt risk, the biomarker is predictive of OS in the high risk group. Nuclear-localized AR-V7 protein in CTCs can aid in the decision between ARSI and taxane chemotherapy in the 2nd or greater line of therapy for mCRPC.

5076 Background: Ra-223 prolongs life in mCRPC pts with symptomatic osseous metastasis with inconsistent effects on PSA. Survival times are prolonged further when combined with Abi/Enza. Data from preclinical studies suggest that Ra-223 may sensitize tumors to DDR agents and/or biologic therapies. But predictive biomarkers of benefit to each or both combinationso are lacking. We studied CTC counts and the prevelance of specific CTC subtypes in patients before and following Ra-223 therapy, both as a single agent and in combination, to identify biomarkers of sensitivity and treatement efficacy, and effects of Ra-223 on tumor biology. Methods: Pre and ~4 week post RA-223 therapy blood samples were collected from 35 pts (2 samples each) given as a single agent (n = 20 pts) or in combination with other therapies (n = 15 pts, 9 w/ Enza, 5 w/ Abi, 1 w/ Taxane). Samples were processed and CTCs analyzed using the Epic Sciences platform. Total CTC count and the prevalence of specific CTC phenotypes present pre and post Rx were identified utilizing high content digital pathology and associated with therapy type and post-treatment change. Results: CTC declines were observed in 55% (11/20) and 60% (9/15) of pts treated with single agent and combination respectively. In Ra-223 alone pts, a novel CTC subtype (high N/C ratio, high nuclear area) was identified at baseline 11/20 samples (med = 33% of CTCs). Which was no longer detected in 10 (90%) of the pts treated. This contrasts with a second novel CTC subtype present at baseline in 4 pts (med CTC = 9%) that increased to 9 cases (med CTC = 18%) at follow-up. Conclusions: A subset of pts demonstrate post-therapy CTC declines following Ra-223 alone or in combination. A novel CTC subtype resolved by RA-223 in conjunction with total CTC kinetics may indicate pt benefit from Ra-223. A novel emergent CTC subtype has also been identified in pts already receiving Ra-223. Single CTC sequencing and protein analysys of these CTC subtypes are ongoing, and may help describe tumor evolution and sensitization to novel therapuetics.

5070 Background: Prior response to A or E does not predict sensitivity to E following A or A following E. The detection of AR-V7 predicts insensitivity to either drug, but identifies only a portion of non-responders. We previously identified 15 CTC subtypes based on unique phenotypic features in mCRPC pts, each with unique biology and different degree of likelihood of predicting resistance to either drug. Here we explored the relationship between individual subtypes and sensitivity to A vs. E, but not both. Methods: 107 pre-treatment blood samples from mCRPC pts starting A (n = 47) or E (n = 60) as a 1st or 2nd line of Tx were analyzed for CTCs utilizing the Epic Sciences platform. Samples were assayed for CTC subtypes based upon 15 pre-defined phenotypic CTC classifiers (Type A-O). Treatment outcomes were assessed by serial PSA changes and landmarked percent time of therapy progression on radiographs, and overall survival following either A or E. Cell type prevalence was also analyzed in relation to clinical outcomes, and subsets of the CTC subtypes subject to single cell NGS to ascertain genomic drivers common to each subtype. Results: CTCs were identified in 94% (101/107) of pt samples. One, cell Type K, found in 25% (27/107) of pts, was associated with a statistically significant inferior outcomes on E for all measures. Whereas similar outcomes were seen between K+ & K- pts treated with A. The distinct features of Cell Type K include a large nucleus, high nuclear entropy and high Nuclear/cytoplasmic AR terminal ratio; and a unique genomic profile enriched for cell cycle and DNA repair alterations relative to other CTC subtypes. Conclusions: The presence of specific CTC subtypes in pre-Rx phlebotomy samples associated with outcomes on A or E. A CTC subtype (Cell Type K) helped to identify pts with poor outcomes on E but not A vs. those without the cell type. Further biologic interrogation of K cells and ongoing clincial validation of the CTC subtype is planned. [Table: see text]

139 Background: Genomic profiling of metastatic tumors can identify mCRPC pts w/ “actionable” targets. However, responses from this approach are infrequent and not durable, broadly attributed to intra-tumor and intrer-cellular het. We sought to understand relationships between single CTC phenotypic and genomic patterns present when starting ARSi or taxane therapy (Tx) to develop more effective biomarker-directed treatment approaches. Methods: 9225 CTCs from 319 blood samples from 179 mCRPC pts were detected and digital phenotypically profiled using the Epic Sciences platform. Unsupervised clustering revealed 15 phenotypic CTC subtypes based on digital pathology features (Type ‘A’-‘O’). Phenotypic diversity was determined by a Shannon index of these subtypes in pt samples. Next, 856 CTCs representing each subtype were individually whole genome sequenced for copy number variation and genomic instability, to define phenogenomic single cell profile and taxonomy. Tx outcomes were associated with CTC profiles. Results: Pre-therapy identification of certain CTC phenotypic subtypes was associated with poor overall survival (OS) following ARSi and/or taxanes. 5/15 subtypes existed primarily in pts w/ high CTC het (84-94%, Bonferroni adjusted p < 0.05, permutation test), and linked to amp of AR and cMYC, and loss of TP53 and PTEN. Clusters of genomic events were linked to phenotypic features (i.e. small cell CTC w/AuroraK & nMYC amp) and provide single cell phenogenomic associations with pathways such as AR/PTEN feedback and DNA repair machinery. Conclusions: Presence ofspecific CTC subtypes was associated w/ poor survival on ARSi and/or taxanes and may inform Tx selection. Identifying CTC subtypes present in high het samples can provide insight into individual pt disease evolution and potential approaches to maximize benefit. Longitudinal monitoring of CTC phenogenomics is being studied in novel Tx trials.[Table: see text]

5038Background: Enzalutamide (ENZ) and abiraterone + prednisone (AA) have transformed the care of pts with mCRPC. Qualified predictive biomarkers are needed to guide treatment selections. Studies o...

5046Background: Upregulation of glucocorticoid receptor protein expression in metastatic biopsies from pts with CRPC has previously been shown to correlate with intrinsic and acquired resistance to...

5013Background: An unmet need in the management of mCRPC is how to use approved AR signaling inhibitors (AR tx) and taxanes (T) to maximize survival for the individual patient. Pts with AR-V7+ CTCs...

5044Background: Prediction of pt response to AR Tx or T is an unmet medical need, as pts may respond to both classes of drug, to one but not the other, or have no response to either. Molecular prof...

163 Background: AR directed Tx, including Abiraterone (A) and Enzalutamide (E), prolong survival in patients with mCRPC and are FDA approved. After primary Tx with either A or E, the response to E or A is decreased as is the duration of response. NGS tissue sequencing of mCRPC shows a range of disease phenotypes and genotypes which are associated with increased tumor heterogeneity. Single site biopsy, may not represent the diversity of disease. Our objective is to develop biomarkers of heterogeneity from the analysis of single cells from patients with mCRPC. Methods: 221 blood samples from 179 unique patients (pts) were collected from pts about to begin AR Tx (n = 150) or Taxanes (n = 71). Samples were analyzed with the Epic Sciences platform. Analysis included digital pathology of 20 discrete phenotypic cell features inclusive of AR, CK, size and shape measures. 9225 single CTCs were characterized, data standardized, features clustered and categorized into 15 phenotypically distinct CTC subtypes. Individual pt samples were then analyzed for the frequency and heterogeneity (Shannon Index) of CTC subtypes and monitored for clinical endpoints. A subset of CTCs (n = 350) were individually sequenced and analyzed for clonality and CNV to assess genomic heterogeneity. Results: A diversity of CTC phenotypes were observed at each line of therapy. In AR Tx pts, high heterogeneity predicts 6mo PFS (OR = 2.57, p = 0.01) and shorter OS (HR = 4.5, p < 0.001). Separately, specific CTC phenotypes also predict 6mo PFS (OR = 5, p < 0.001) and shorter OS (HR = 5.8, p < 0.001). Heterogeneity scores were not associated with taxane resistance. NGS analysis identified subclonal drivers of disease progression including multiple private and/or clustered genomic drivers. Conclusions: Single CTC characterization of phenotype and genotype supports the concept of tumor heterogeneity as a driver of disease resistance. The heterogeneity observed supports the potential value of single cell characterization to identify rare somatic sub-clonal alterations and can aid in the development of rational therapeutic trials.

302 Background: CellSearch is an FDA cleared assay that enumerates intact CK+, DAPI+, CD45- cells via EpCAM capture. CTC number is prognostic for survival pre- and post- systemic therapy. These cells are a subset of confirmed genomically altered CTC’s in blood. We report the prognostic significance for OS in mCRPC of CTC subtypes not enumerated by CellSearch including apoptotic CTCs, CK- CTCs, small CTCs, and CTC clusters. Methods: 221 blood samples from 179 unique patients were collected from pts about to begin AR directed (n = 150) or taxane (n = 71) therapy for progressive CRPC. Samples were analyzed utilizing the Epic Sciences platform to enumerate traditional (CK+, abnormal morph), apoptotic (CK+, fragmented nuclei), CK- (CK-/abnormal morph), small (CK+, normal morph), and CTC clusters. Patients were followed for up to 2.3 yrs. Paired CellSearch and Epic traditional CTC counts were collected on 173 patient samples. Results: In paired analysis, Epic traditional CTCs were identified in 86% of patients vs. 67% by CellSearch, and any Epic CTC subtype was found in 96% of the patient samples. Increased frequency of any CTC subtype was a univariate predictor of OS (see table). Conclusions: Epic identified CTC subtypes not identified by CellSearch or other technologies that select or sort based upon size or epithelial expression. Cells were detected in 96% of the patient samples and all 5 subtypes were found to be prognostic for survival. The biologic characterization of CTC subtypes as a biomarker of drug sensitivity is ongoing. The increased clinical sensitivity of subtypes and prevalence of CTCs identified on the Epic platform provides a greatly expanded array of clinically relevant biomarkers to develop predictive liquid biopsy signatures in the metastatic setting. [Table: see text]

The disclosure provides methods correlating intra-patient genomic heterogeneity of single CTCs with phenotypic heterogeneity in each of a population of prostate cancer (PCa) patients.

5029 Background: STEAP1 is overexpressed in mCRPC and is an ADC target. STEAP1 expression in tissue assayed by IHC, protein and mRNA expression levels in circulating tumor cells (CTC) and by 89Zr-l...

147 Background: The optimal sequence to administer approved life prolonging ARS inhibitors: Abiraterone (A) and Enzalutamide (E),and taxanes: Docetaxel (D) and Cabazitaxel (C), to maximize individual pt benefit is an unmet medical need. Sensitivity to an individual agent is dependent on the intrinsic biology of the tumor and the treatment to which it has been exposed. For each agent, response rates are lower in pts who have received prior treatment with any of the other drugs, but for the individual, sensitivity to one does not necessarily predict sensitivity to the other. We sought to develop response and resistance signatures to A & E or D & C from a baseline blood sample. Methods: 91 pt blood samples collected from 79 pts for CTC analysis with the Epic Sciences platform prior to treatment (27 pre-A, 28 pre-E, 28 pre-D, 8 pre-C). Epic analysis identified traditional CTCs (CK+, CD45-, intact nuclei, morph distinct), CK- CTCs (CK-, CD45-, intact nuclei, morph distinct), small CTCs (CK+, CD45-, intact nuclei, small cell size), and CTC clusters. IF staining for AR N, AR C exp were performed, digital pathology algorithms analyzed CTC morphology. A classifier was developed to associate clinical phenotypes with outcome to a specific agent. Results: A and E signatures included: AR N/C exp. and presence of CK+, AR+, Nucleoli+ CTCs. D and C signatures by: presence of CK+, small, AR-, Nucleoli+ CTCs. Multivariate algorithms for A and E and D & C were statistically associated with de novo resistance. Line of therapy was not a univariate predictor of response. Conclusions: Characterization of CTCs identified predictive biomarkers of sensitivity to ARS Tx and taxane chemotherapy in mCRPC pts. The proposed A&E signature differs from C and D, providing the opportunity to better guide treatment selection and improve patient outcomes using a real-time, non-invasive blood biopsy. Prospective trials to validate results are planned. [Table: see text]

132 Background: Numerous resistance mechanisms have been postulated in progressive mCRPC. Determining the presence of putative predictive biomarkers in patients requires a real-time tumor assessment because the biology changes under the influence of the specific therapy(ies) that a patient (Pt) has received. We examined CTC and CTC subpopulation incidence and molecular characterization of tumors from Pts with progressive mCRPC to assess if determinants of resistance can be assessed through a CTC test. Methods: 48 samples from 21 unique progressive mCRPC pts treated with androgen receptor targeted (AR tx) therapies, 9 (43%) on Abiraterone plus Prednisone (AA+P) and 12 (57%) on Enzalutamide (E). Samples were collected and shipped to Epic Sciences, where cells were stained and CTC identified by fluorescent scanners and algorithmic analysis. CTCs, defined as classic (CK+ CD45- w/intact DAPI nuclei and distinct), apopotic (CK+, CD45-, non-intact nuclei) and CK- (CK-, CD45-, intact and distinct) were identified. CTCs reported per mL of blood and were examined for AR, PTEN & ERG. CTC data were analyzed in context of PSA, Veridex CTC (reported per 7.5mL of blood), and clinical history. Results: With Epic, all 21 pts had detectable CTCs (mdn 23 cells/ml, range 2 to 249), whereas with Veridex, 14 (67%) pts had > 5 CTC/7.5 ml (mdn 5 cells/7.5 ml, range 0 to >200). 0/7 (0%) with high AR responded to AA+P or E, 8/14 (53%) with low AR had a PSA decline and stable radiographic disease (median follow-up 12 wks). Variations in AR protein expression and localization heterogeneity was seen in all pts to varying degrees. 6/21 (29%) Pts demonstrated CK- CTC/CTC ratio >50% (mdn 9 cells/ml, range 5-38). ERG rearrangement and PTEN loss were correlated. Conclusions: Epic CTC analysis provides higher detection rates, while enabling individual cell analyses for predictive biomarkers of sensitivity to AR directed therapies. Notable was the marked heterogeneity of detection of AR, ERG, and PTEN of individual CTCs. Studies are ongoing to further explore associations with outcomes. [Table: see text]