OBJECTIVES:Anetumab ravtansine is an antibody-drug conjugate consisting of a fully human anti-mesothelin monoclonal antibody conjugated to cytotoxic maytansinoid tubulin inhibitor DM4. Mesothelin is highly expressed in ovarian cancer. This phase Ib study determines the safety, pharmacokinetics, and anti-tumor activity of anetumab ravtansine and pegylated liposomal doxorubicin in mesothelin-expressing platinum-resistant ovarian cancer.METHODS:Anetumab ravtansine (5.5 or 6.5 mg/kg) and pegylated liposomal doxorubicin (30 mg/m2) were administered intravenously every 3 weeks to 65 patients with platinum-resistant epithelial ovarian cancer. Mesothelin expression was assessed by central immunohistochemistry. Adverse events, tumor response (RECIST 1.1), and progression-free survival were determined. Biomarker samples were assessed by ELISA and next-generation sequencing.RESULTS:In dose escalation, nine patients received anetumab ravtansine across two doses (5.5 or 6.5 mg/kg). The maximum tolerated dose of anetumab ravtansine was 6.5 mg/kg every 3 weeks and no dose-limiting toxicities were observed. In dose expansion, 56 patients were treated at the maximum tolerated dose. The most common treatment-emergent adverse events of any grade were nausea (47.7%), decreased appetite (43.1%), fatigue (38.5%), diarrhea (32.3%), and corneal disorder (29.2%). In all treated patients the objective response rate was 27.7% (95% CI 17.3% to 40.2%), including one complete (1.5%) and 17 partial responses (26.2%), with median duration of response of 7.6 (95% CI 3.3 to 10.2) months and median progression-free survival of 5.0 (95% CI 3.2 to 6.0) months. In an exploratory analysis of a sub-set of patients (n=19) with high mesothelin expression who received ≤3 prior lines of systemic therapy, the objective response rate was 42.1% (95% CI 20.3% to 66.5%) with a median duration of response of 8.3 (95% CI 4.1 to 12.0) months and median progression-free survival of 8.5 (95% CI 4.0 to 11.4) months.CONCLUSIONS:Anetumab ravtansine and pegylated liposomal doxorubicin showed tolerability and promising clinical activity. These results established the dose schedule and the mesothelin-positive target population of this combination for a phase III study in platinum-resistant ovarian cancer.TRIAL REGISTRATION NUMBER:NCT02751918.

Introduction: The combination of the PI3K inhibitor copanlisib (C) plus the anti-CD20 antibody rituximab (R) has been shown to be superior to R plus placebo (P) in patients with relapsed iNHL (Matasar et al. Lancet Oncol 2021). We report here the biomarker analysis to assess correlation with response of baseline samples for PTEN expression and GEP signatures. Methods: Adult patients with relapsed B-cell iNHL were randomly assigned 2:1 to either C + R or P + R; standard dosing on a 28-day cycle applied. Either fresh or archival tumor tissues were collected for central pathology review and biomarker analysis. Extraction of the formalin-fixed, paraffin-embedded (FFPE) tissue slides was performed using the Qiagen AllPrep DNA/RNA FFPE Kit, where quality was assessed using a combination of Agilent TapeStation for integrity and ThermoFisher NanoDrop for quantity and purity. Immunohistochemistry (IHC) for PTEN expression was performed by Mosaic Laboratories (Lake Forest, CA, USA) and RNA sequencing analysis was performed by Almac Diagnostics (Durham, NC, USA). GEP was also conducted using claraT pathway signatures provided by Almac. Results: Patients with complete, very good partial, and partial objective responses were combined in a “responders” group and compared against all other patients (“non-responders” group). Data from PTEN IHC were available for 222 patients (149 for C + R; 73 for P + R). There was no statistically significant difference in response with PTEN presence or absence for the 3 subgroups - total iNHL, follicular lymphoma (FL), and non-FL. Also, baseline PTEN expression did not show significant correlation with best overall response. C + R treatment improved median progression-free survival (PFS) for all patients, with a statistically significant benefit to those with PTEN presence in all 3 subgroups. For patients treated with P + R, Kaplan-Meier plots and log-rank tests showed statistically improved median PFS in patients with PTEN absence vs PTEN presence in overall iNHL (p=0.031) and FL (p=0.0021). Data from GEP were available for 202 patients (133 for C + R; 69 for P + R). The association with best overall response and PFS was analyzed using specific biomarker gene sets and/or 23 preselected clinical parameters. Association of GEP analysis based on hypothesis-based modeling, single-gene hypothesis-free modeling, and claraT pathway signatures showed no predicted responsiveness or PFS that was statistically significant. Conclusions: C + R improved median PFS in the overall iNHL, FL, and non-FL groups, with a statistically significant improvement in the PTEN-positive population. In the P + R group, PTEN-positive patients had worse PFS in all 3 subgroups. Data from GEP and pathway analysis did not show correlation with overall response or PFS. Citation Format: Shalini Chaturvedi, Anke Schulz, Lidia Mongay Soler, Barrett H. Childs, Matthew Matasar, Pier Luigi Zinzani. Biomarker assessment of PTEN protein expression and gene expression profiling (GEP) in a phase III study of copanlisib in combination with rituximab in patients with indolent non-Hodgkin lymphoma (iNHL) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr LB523.

89 Background: Combination treatment with regorafenib (80–120 mg/day, PO, 3 wks on/1 wk off) plus nivolumab (480 mg IV Q4W) showed manageable safety but modest efficacy in a phase 2 study of 70 pts from North America with pMMR/MSS chemotherapy-resistant metastatic CRC (mCRC). Five pts had a partial response (PR; objective response rate [ORR]: 7%); all did not have liver metastases at baseline (n = 5/23; ORR: 22%). One pt had a confirmed complete response (CR) after the primary completion analysis of the study (ASCO 2021). This retrospective exploratory analysis investigated the potential association between specific biomarkers and anti-tumor activity, and how those biomarkers are modulated by treatment with regorafenib plus nivolumab. Methods: In formalin-fixed paraffin-embedded tumor samples obtained at baseline and Cycle (C) 2 Day (D) 8, immune-related biomarkers were assessed via immunohistochemistry (IHC), and RNA sequencing was used for gene expression profiling/gene signatures. Pre-/on-treatment blood samples were collected to measure circulating tumor DNA (ctDNA) and soluble biomarkers. Results: A total of 40 and 27 baseline tumor samples and 14 and 5 paired tumor samples at baseline/C2D8 were available for IHC and RNA sequencing, respectively. Higher baseline protein and mRNA expression of biomarkers for pre-existing immune sensitivity (eg, effector T cells) trended with anti-tumor activity. These biomarkers were expressed at lower levels in pts with liver metastases vs those without liver metastases at baseline. Cytotoxic T cell density was elevated on C2D8 but did not correlate with anti-tumor activity. Increased mean variant allelic frequency in ctDNA at C2D1 predominated in pts with progressive disease (PD), while clearance of ctDNA at C2D1 was only noted for the one pt with a CR. High clonal tumor mutational burden in ctDNA showed a numerical trend with anti-tumor activity (PD vs. SD/PR; P=0.058) and PFS (P = 0.072). Baseline serum levels of select markers related to angiogenesis (eg, vascular endothelial growth factor [VEGF] D) were associated with inferior anti-tumor activity (P = 0.002). Serum levels of immune-related soluble biomarkers (eg, tumor necrosis factor alpha) increased on treatment (P < 0.005), while levels of soluble VEGF receptor 2 decreased (P < 0.001). Conclusions: This study of pts with MSS mCRC treated with regorafenib plus nivolumab suggests that baseline tumor biomarkers for pre-existing immune sensitivity trended with anti-tumor activity, whereas select baseline peripheral biomarkers related to angiogenesis trended with inferior anti-tumor activity. Pharmacodynamics effects were observed, yet no significant correlation with anti-tumor activity was found. Due to the small sample size and retrospective nature, these analyses are hypothesis-generating. Clinical trial information: NCT04126733.

Background: The pan-PI3 K inhibitor copanlisib is approved as monotherapy in adult patients (pts) with relapsed follicular lymphoma. While PI3 K activation is often observed in high-risk pediatric cancers, copanlisib had not been studied in this setting. We report the results of a trial sponsored by Bayer AG and conducted in collaboration with the Children’s Oncology Group on the safety, efficacy, and pharmacokinetics (PK) of copanlisib in pediatric pts with relapsed/refractory solid tumors (NCT03458728). Methods: Pts were aged 6 months to 21 years, with a confirmed histologic diagnosis of malignancy with no standard curative or effective therapy after ≥1 line of therapy. Copanlisib was given i.v. over 1 hour on days 1, 8, and 15 of a 28-day cycle. Primary endpoints were safety and the determination of the maximum tolerated dose (MTD) and/or recommended Phase II dose (RP2D); secondary endpoints included PK, response, and analysis of biomarkers including % pAKT. Dose-limiting toxicity (DLT) criteria were amended to exclude non-clinically significant laboratory abnormalities after 5 pts were treated at 28 mg/m2. Results: 23 pts received copanlisib: 16 at 28 mg/m2 and 7 at 35 mg/m2. Median age was 14 years (range 7–19) and median number of prior anticancer therapies was 4 (range 2–18). The most common tumor types were neuroblastoma (26%) and osteosarcoma (22%). DLTs were observed in 1/8 pts in the 28 mg/m2 cohort (hyperuricemia) and in 2/7 pts in the 35 mg/m2 cohort (decreased platelet count [n = 1]; transient grade 3 hypertension and fever [n = 1]). The MTD/RP2D was 28 mg/m2; 3 additional pts were treated at the MTD with no DLTs. The most common treatment-emergent adverse events (TEAEs) were hyperglycemia (70%), nausea (61%), and decreased white blood cell count (52%). Grade 3/4 TEAEs were observed in 56%/25% and 86%/0% of pts in the 28 and 35 mg/m2 cohorts, respectively, with no grade 5 TEAEs. Copanlisib exposure at 28 mg/m2 was ∼80% of the adult exposure at the 60 mg dose. No objective responses occurred (22 evaluable pts); best response was stable disease (3 pts, 14%) and progressive disease (19 pts, 86%); 2 pts received >2 treatment cycles. A significant decrease in % pAKT from baseline to 3 hours after dosing occurred in surrogate tissue of platelet-rich plasma on day 15 of cycle 1 (p < 0.001), consistent across samples and copanlisib dose, irrespective of response (mean inhibition 75%; median inhibition 77%). Conclusions: The RP2D of single-agent copanlisib in pediatric pts was 28 mg/m2 i.v. on days 1, 8, and 15 of a 28-day cycle; the PK expansion cohort is ongoing. No unexpected toxicities were observed. The PK profile and pharmacodynamic effects were similar to those seen in adults, but no evidence of efficacy as monotherapy was observed. Studies of copanlisib plus other active agents may be warranted. Conflict of interest: Ownership: Margaret Macy: stock in Johnson & Johnson. Funan Huang: stock in Bayer. Charles Phelps: stock in Bayer, Johnson & Johnson, Pfizer, Amgen, and Eli Lilly and Company. Advisory Board: Thomas Cash: Y-mAbs Therapeutics. Elizabeth Fox: de minimis compensation from Moffit Cancer Center Sunshine Project. Board of Directors: Elizabeth Fox: leadership or fiduciary role in other board, society, committee or advocacy group, paid or unpaid, with Children’s Oncology Group/Pediatric Early Phase Clinical Trial Network, ACCELERATE, and CureSearch. Corporate-sponsored Research: Margaret Macy: salary support and institutional support for this abstract from Bayer HealthCare Pharmaceuticals, Inc. Thomas Cash: institutional grants or contracts from Celgene/BMS and F. Hoffman-La Roche/Genentech. Joseph G. Pressey: grants or contracts from the Jeff Gordon Foundation. Gregory K. Friedman: grants or contracts from Eli Lilly and Company, Pfizer, and Eisai. Joachim Grevel: institutional support for this abstract from Bayer Germany. Stephen G. DuBois: support for this abstract from Bayer. Other Substantive Relationships: Margaret Macy: consulting fees from Y-mAbs Therapeutics. Thomas Cash: payment or honoraria for lectures, presentations, speaker bureau, manuscript writing or educational events, and support for attending meetings and/or travel, from EUSA Pharma. Joseph G. Pressey: payment or honoraria for lectures, presentations, speaker bureau, manuscript writing or educational events from General Dynamics. Gregory K. Friedman: patent pending related to oncolytic virotherapy for brain tumors, and receipt of equipment, materials, drugs, medical writing, gifts or other services from Treovir LLC, who provided G207 for an investigator-initiated trial related to brain cancer treatment. Joachim Grevel: consulting fees from BAST Inc Ltd. Funan Huang: employee of Bayer HealthCare Pharmaceuticals, Inc. Charles Phelps: employee of Bayer HealthCare Pharmaceuticals, Inc. Barrett H. Childs: employee of Bayer HealthCare Pharmaceuticals, Inc. John Chung: employee of, and support for attending meetings and/or travel from, Bayer HealthCare Pharmaceuticals, Inc. Shalini Chaturvedi: employee of Bayer HealthCare Pharmaceuticals, Inc. Anke Schultz: employee of Bayer AG. Steven G. DuBois: consulting fees from Amgen, Bayer, Jazz, and Loxo Oncology, and travel expenses from Loxo Oncology, Roche, and Salarius. Navin Pinto, Luca Szalontay, Wayne L. Furman, Jennifer H. Foster, Josephine HaDuong, Andrew Bukowinski, and Brenda J. Weigel have no conflicts of interest to declare.

We previously described a prognostic role for the glycoprotein STC1 in patients with metastatic CRC (mCRC) receiving first-line chemotherapy and in a small number of patients with refractory mCRC previously treated with regorafenib (Jary, et al. Int J Cancer 2020). Regorafenib is approved for the treatment of patients with mCRC based an improvement in overall survival (OS) in the placebo-controlled, randomized, phase 3 CORRECT trial (Grothey, et al. Lancet 2013; NCT01103323). Here, we investigate the relationship between STC1 and clinical outcome in patients from CORRECT.

The CORRECT trial showed an overall survival (OS) benefit for regorafenib in mCRC pts who progressed after standard therapies. No predictive biomarker exists for regorafenib and treatment is often burdened by severe toxicity. We investigated the prognostic and predictive value of genetic variants selected based on previous associations with antiangiogenic treatment. Genomic DNA from 507 pts (337 regorafenib, 170 placebo) was genotyped using a customized OncoArray (Illumina). 149 candidate SNPs representing 78 genes and 15 pathways were identified. Stratified multivariable Cox proportional hazards regression was used to jointly assess the predictive and prognostic effects, using a 2-df likelihood ratio test of the SNP effect and SNP*treatment interaction. SNPs were coded additively. False discovery rate (FDR) was controlled at Q = 0.1. A functional SNP, rs1056560, in the CRY1 gene in the clock pathway was significantly associated with OS (P < .001, Q = .06). Both prognostic and predictive effects (marginal SNP effect and SNP*treatment interaction) achieved nominal significance (P = .04 and P < .001, respectively). Significant OS benefit was found for pts with rs1056560 C allele variant treated with regorafenib: 6.5 vs 3.3 months with placebo for CC genotype carriers (N = 93, HR 0.48, 95%CI 0.30-0.78, P = .003) and 6.9 vs 5.4 months for CA genotype (N = 221, HR 0.62, 95%CI 0.45-0.85, P = .003). No treatment benefit was observed for the AA genotype (N = 174). No SNP achieved FDR significance for progression free survival, however, CRY1 rs1056560 was nominally significant for the 2-df test (P = .01) and the predictive effect (P = .007). For the first time we identify a potential predictive biomarker for regorafenib in the CORRECT population. CRY1 is a negative regulator of the clock machinery, which modulates the circadian expression of key target genes, including VEGF. Growing interest is focusing on alterations of this pathway in cancer. Our finding helps better understand regorafenib response and may support alternative treatment choices based on CRY1 rs1056560 genotypes to optimize regorafenib benefit.

Introduction  This X-VeRT (eXplore the efficacy and safety of once-daily oral riVaroxaban for the prevention of caRdiovascular events in patients with nonvalvular aTrial fibrillation scheduled for cardioversion) substudy evaluated the effects of treatment with rivaroxaban or a vitamin-K antagonist (VKA) on levels of biomarkers of coagulation (D-dimer, thrombin-antithrombin III complex [TAT] and prothrombin fragment [F1.2]) and inflammation (high sensitivity C-reactive protein [hs-CRP] and high-sensitivity interleukin-6 [hs-IL-6]) in patients with atrial fibrillation (AF) who were scheduled for cardioversion and had not received adequate anticoagulation at baseline (defined as, in the 21 days before randomization: no oral anticoagulant; international normalized ratio <2.0 with VKA treatment; or <80% compliance with non-VKA oral anticoagulant treatment). Methods  Samples for biomarker analysis were taken at baseline ( n  = 958) and treatment completion (42 days after cardioversion; n  = 918). The influence of clinical characteristics on baseline biomarker levels and the effect of treatment on changes in biomarker levels were evaluated using linear and logistic models. Results  Baseline levels of some biomarkers were significantly associated with type of AF (D-dimer and hs-IL-6) and with history of congestive heart failure (hs-CRP, D-dimer, and hs-IL-6). Rivaroxaban and VKA treatments were associated with reductions from baseline in levels of D-dimer (-32.3 and -37.6%, respectively), TAT (-28.0 and -23.1%, respectively), hs-CRP (-12.5 and -17.9%, respectively), and hs-IL-6 (-9.2 and -9.8%, respectively). F1.2 levels were reduced from baseline in patients receiving a VKA (-53.0%) but not in those receiving rivaroxaban (2.7%). Conclusion  Anticoagulation with rivaroxaban reduced levels of key inflammation and coagulation biomarkers to a similar extent as VKAs, with the exception of F1.2. Further investigation to confirm the value of these biomarkers in patients with AF is merited.

BACKGROUNDThe analysis of subgroups in clinical trials is essential to assess differences in treatment effects for distinct patient clusters, that is, to detect patients with greater treatment benefit or patients where the treatment seems to be ineffective.METHODSThe software application subscreen (R package) has been developed to analyze the population of clinical trials in minute detail. The aim was to efficiently calculate point estimates (eg, hazard ratios) for multiple subgroups to identify groups that potentially differ from the overall trial result. The approach intentionally avoids inferential statistics such as P values or confidence intervals but intends to encourage discussions enriched with external evidence (eg, from other studies) about the exploratory results, which can be accompanied by further statistical methods in subsequent analyses. The subscreen application was applied to 2 clinical study data sets and used in a simulation study to demonstrate its usefulness.RESULTSThe visualization of numerous combined subgroups illustrates the homogeneity or heterogeneity of potentially all subgroup estimates with the overall result. With this, the application leads to more targeted planning of future trials.CONCLUSIONThis described approach supports the current trend and requirements for the investigation of subgroup effects as discussed in the EMA draft guidance for subgroup analyses in confirmatory clinical trials (EMA 2014). The lack of a convenient tool to answer spontaneous questions from different perspectives can hinder an efficient discussion, especially in joint interdisciplinary study teams. With the new application, an easily executed but powerful tool is provided to fill this gap.

Investigation of the mRNA expression profile in endometrial tissue of healthy women treated with different doses of the Selective Progesterone Receptor Modulator Vilaprisan (VPR). Endometrial biopsy samples were obtained from 163 healthy tubal-ligated women enrolled in a double blind, parallel-group study, randomized to VPR doses of 0.1, 0.5, 1, 2 and 5 mg/d or placebo and treated over 84 days. Biopsy specimen were taken before treatment (d9±2), end of treatment (d84-2) and after start of next menstrual bleeding (d9±2). RNA isolated from all biopsy samples were labeled and hybridized to Illumina HumanHT 12 v4. For Quality control (QC) the distribution and the effect of QC parameters like for example background, noise, RNA yield and batch was visualized. Illumina samples were pre-processed by simple scaling normalization on the Background. Genes identified as up-or down-regulated in comparison of pre-treatment versus treatment in the high dose groups (1, 2 and 5mg) were used for GO analysis. The group or time differences in gene expression were investigated using moderated t-tests (paired and unpaired). Treatment with VPR exerts strong effects on the endometrial gene expression which are almost all reversible or reduced after first bleeding post-treatment. Gene ontology analysis revealed that the majority of affected genes were highly down-regulated and involved in processes that are directly related to the cell cycle regulation/progression or control of cell growth, like AURKA, CDK2, CCNB1 DLG7, MELK and Ki67. This indicates VPR's strong inhibitory effect on endometrial proliferation. These genes are also known to be overexpressed in endometrial carcinoma and other malignant tumors. In addition, genes playing a specific role in endometrial decidualisation and receptivity (e.g. FOXM1, HOXA10) are also found to be repressed under VPR treatment. Endometrial gene expression is strongly affected by VPR treatment. The majority of genes were down-regulated and involved in cell growth suggesting an antiproliferative effect of VPR on the endometrium. A dedicated mechanism of action study of VPR has been recently initiated in fibroid patients to further investigate the effect of VPR on menstrual bleeding and fibroid shrinkage.

562 Biomarkers -Heart Failure 7 day ECG monitoring to exclude silent AF.Blood samples were analysed for 40 cardiovascular biomarkers.Results: Patients with known AF (n=298) had higher blood levels of adrenomedullin (ADM), interleukin-27 (IL-27), fibroblast-growth factor 23 (FGF-23), TRAIL 2, growth hormone (GH), receptor for advanced glycosylation end products (RAGE), VEGF-D, and brain natriuretic peptide (BNP).Patients in sinus rhythm had a higher pappalysin-1 and melusin levels compared to AF patients.Patients in sinus rhythm (n=369) were significantly younger (66.3±11.5 years in sinus group, 71.3±12.1 years in AF group), with higher proportion of diabetes, hypertension, and coronary artery disease than AF patients.Figure 1 shows the differences in mean BNP and FGF-23 levels between AF and sinus rhythm groups.A logistic regression model including clinical factors, echocardiographic parameters, and biomarkers with univariate differences between groups identified FGF-23 (OR 1.355 per log2 unit increase, p=0.000186),BNP (OR 1.316 per log2 unit increase, p=0.000002) as independent markers for AF.Pappalysin-1 and ADM were also significantly different (p=0.02 for both) Conclusion: BNP and FGF-23 are associated with AF in an unselected cohort of patients presenting to a large hospital.In addition to natriuretic peptides, activation of fibroblasts seems a common driver of AF.

Activating KRAS mutations are reported in up to 90% of pancreatic cancers. Refametinib potently inhibits MEK1/2, part of the MAPK signaling pathway. This phase I/II study evaluated the safety and efficacy of refametinib plus gemcitabine in patients with advanced pancreatic cancer.Phase I comprised dose escalation, followed by phase II expansion. Refametinib and gemcitabine plasma levels were analyzed for pharmacokinetics. KRAS mutational status was determined from circulating tumor DNA.Ninety patients overall received treatment. The maximum tolerated dose was refametinib 50 mg twice daily plus standard gemcitabine (1000 mg/m(2) weekly). The combination was well tolerated, with no pharmacokinetic interaction. Treatment-emergent toxicities included thrombocytopenia, fatigue, anemia, and edema. The objective response rate was 23% and the disease control rate was 73%. Overall response rate, disease control rate, progression-free survival, and overall survival were higher in patients without detectable KRAS mutations (48% vs. 28%, 81% vs. 69%, 8.8 vs. 5.3 months, and 18.2 vs. 6.6 months, respectively).Refametinib plus gemcitabine was well tolerated, with a promising objective response rate, and had an acceptable safety profile and no pharmacokinetic interaction. There was a trend towards improved outcomes in patients without detectable KRAS mutations that deserves future investigation.

Background: The single hotspot mutation AKT1 [G49A:E17K] has been described in several cancers, with the highest incidence observed in breast cancer. However, its precise role in disease etiology remains unknown.Methods: We analyzed more than 600 breast cancer tumor samples and circulating tumor DNA for AKT1(E17K) and alterations in other cancer-associated genes using Beads, Emulsions, Amplification, and Magnetics digital polymerase chain reaction technology and targeted exome sequencing.Results: Overall AKT1(E17K) mutation prevalence was 6.3 % and not correlated with age or menopausal stage. AKT1(E17K) mutation frequency tended to be lower in patients with grade 3 disease (1.9 %) compared with those with grade 1 (11.1 %) or grade 2 (6 %) disease. In two cohorts of patients with advanced metastatic disease, 98.0 % (n = 50) and 97.1 % (n = 35) concordance was obtained between tissue and blood samples for the AKT1(E17K) mutation, and mutation capture rates of 66.7 % (2/3) and 85.7 % (6/7) in blood versus tissue samples were observed. Although AKT1-mutant tumor specimens were often found to harbor concurrent alterations in other driver genes, a subset of specimens harboring AKT1(E17K) as the only known driver alteration was also identified. Initial follow-up survival data suggest that AKT1(E17K) could be associated with increased mortality. These findings warrant additional long-term follow-up.Conclusions: The data suggest that AKT1(E17K) is the most likely disease driver in certain breast cancer patients. Blood-based mutation detection is achievable in advanced-stage disease. These findings underpin the need for a further enhanced-precision medicine paradigm in the treatment of breast cancer.

Introduction: In the randomized phase 3 CORRECT trial (NCT01103323), regorafenib improved overall survival (OS) and progression-free survival (PFS) vs placebo in patients with treatment-refractory mCRC (n = 505 regorafenib; n = 255 placebo). Biomarker analysis showed that regorafenib was consistently associated with a clinical benefit in subgroups defined by tumor mutational status and protein biomarker concentrations (Tabernero, Lancet Oncol 2015;16:937). SNPs may be potential biomarkers for predicting treatment benefit and toxicity. We report a retrospective and exploratory analysis of SNP biomarkers from CORRECT. Methods: Genotyping was performed on whole-blood DNA from 528/760 (69%) patients using SEQUENOM's MassARRAY. A potential association of 258 SNPs from the VEGF-A signaling pathway with clinical benefit was assessed using Cox proportional hazard model (alpha = 0.10 after multiplicity adjustment; models were computed under additive, dominant, and recessive assumptions). A possible association of 162 SNPs from regorafenib metabolism genes with selected Grade 3–4 adverse events (AEs) was evaluated using logistic regression (alpha = 0.10). Results: Of the 211 VEGF-A SNPs tested, none showed a significant interaction with treatment and PFS. Under additive assumptions, SNP rs7024233 in the TIE2 gene showed a statistically significant interaction with treatment and OS (adjusted p = 0.0317). The heterozygote of the TIE2 SNP (CT) was associated with improved OS with regorafenib vs placebo (Table), while the direction of the effect is reversed for the homozygous genotype (CC or TT), which questions the robustness of these findings. This is supported by the fact that SNP rs7024233 did not show a significant interaction with treatment in either the dominant or the recessive model. Of the 22 metabolism SNPs that were assessable, none were associated with incidence of any of the Grade 3–4 AEs in patients treated with regorafenib. However, gene-level association testing for AE incidence identified several associations, including association of the UGT1A1/UGT1A9 gene with hypertension and hand–foot skin reaction. Conclusion: Although these preliminary exploratory analyses did not find a clear association between the VEGF-A SNP biomarkers tested and regorafenib clinical efficacy, the results suggest that polymorphisms in metabolism genes have the potential to predict safety in patients with mCRC. Tabled 1 Open table in a new tab

Abstract Background: Mutationally-activated KRAS is present in 90% of pancreatic ductal adenocarcinoma (PDAC) and may represent an early genetic driver, being commonly found in low-grade pancreatic lesions (Eser et al. 2014 BJC 111:817). Refametinib is a potent oral allosteric MEK 1/2 inhibitor with both single-agent activity and synergistic activity in combination with gemcitabine in preclinical models of pancreatic cancer (PC). A Phase 1B/2 study in patients with locally advanced, unresectable or metastatic PC and no prior systemic therapy was conducted and recently reported (Van Laethem et al., ASCO 2014; NCT01251640). We report here on the exploratory biomarker findings from this study. Methods: KRAS mutational analysis was conducted via liquid biopsy at baseline on circulating tumor DNA (ctDNA) by BEAMing (Sysmex-Inostics) as well as circulating micro RNA (miRNA) from plasma collected at baseline and post-dose and analyzed by qPCR using an Exiqon panel of 752 miRNAs and an innovative data preprocessing method for normalization and imputation of undetermined values. Tumor molecular characterization was performed on archival tumor tissue and included targeted tumor gene next-generation sequencing with FOUNDATION ONE and the analysis of Ki67 proliferation index. Results: Samples for biomarker analysis were obtained from 69 treated patients. Forty-six (67%) had detectable KRAS mutations by liquid biopsy. KRAS G12D, G12V and G12R were the most frequent mutations. Interestingly, KRAS wild-type patients had better efficacy outcomes compared to mutant KRAS patients (mut/WT, respectively): overall response rate 15%/30% (OR 2.4, p = 0.147), median progression-free survival (mPFS) 3.7/8.8 mo (HR 0.32, p = 0.001), and overall survival (OS) 7.1/18.2 mo (HR 0.28, p = 0.001). There was a trend correlating KRAS mutant allele frequency with response. The CA19.9 levels correlated with KRAS mutational status. Tumor exome sequencing was performed from 16 patients, 15 of which had a KRAS mutation (G12D or G12V). The discordancy rate compared to BEAMing KRAS data was 26% (4/15). Conclusions: The high prevalence of KRAS mutations in patients with PC has been confirmed using BEAMing technology. In this study, there was an association between improved mPFS and OS in KRAS WT patients. Together with lower baseline levels of CA19.9 in the KRAS WT cohort, we conclude that liquid biopsy may be an approach to identify prognostic or predictive markers in PDAC treated with refametinib and gemcitabine. This hypothesis is sustained by the finding that poor clinical response showed increasing allele frequency of mutant KRAS. These results require confirmation in a larger trial. Citation Format: Michael Teufel, Jean-Luc Van Laethem, Hanno Riess, Marius Giurescu, Vittorio L. Garosi, Anke Schulz, Richardus Vonk, Henrik Seidel, Joachim Reischl, Barrett H. Childs. KRAS wild-type status as detected by circulating tumor DNA analysis may be a prognostic or predictive factor for clinical benefit in patients with unresectable, locally advanced or metastatic pancreatic cancer (PC) treated with the MEK inhibitor refametin [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 5239. doi:10.1158/1538-7445.AM2015-5239

Sharing pharma compounds with academia: experiences with providing vitamin D receptor ligands

4129 Background: Refametinib is a potent oral allosteric MEK 1/2 inhibitor with both single-agent and synergistic activity in combination with gemcitabine in preclinical models of pancreas cancer (PC). A synergistic effect between refametinib and erlotinib was also reported in KRAS wild-type human PC in vitro (Diep et al, Clin Cance Res 2011). We report the biomarker results of a single-arm, open label, phase 2a study combining refametinib and gemcitabine in advanced PC. Methods: Refametinib was administered 50 mg bid in combination with gemcitabine. The primary objective was overall response rate (ORR). Secondary objectives included progression-free survival (PFS), overall survival (OS) and biomarker assessment. KRAS mutational analysis was performed from plasma collected at baseline on circulating tumor DNA (ctDNA) by BEAMing. Circulating miRNA were obtained from plasma collected at baseline and post-dose and analyzed by qPCR. Targeted tumor gene next-generation sequencing, gene expression analysis by RNA sequencing and the proliferation index were conducted. Results: Of the 60 patients treated, 39 (65%) had KRAS mutations. In the KRAS subgroups (mut/WT, respectively), the ORR, mPFS and OS were 28%/48% (p=0.136), 4.6/9.0 mo (HR 0.26) and 6.6/18.2 mo (HR 0.27). There was a trend correlating allele frequency with response. KRAS G12D, G12V and G12R were the most frequent mutations. Mutations of codon 38 or 436 were not observed. Tumor exome sequencing was performed from 16 patients, 15 of which had a KRAS mutation (G12D or G12V). The most frequent co-occurring somatic mutations or amplifications were TP53, CDKN2A, and cMYC. Tumor proliferation index and miRNA signatures were not correlated with response. Gene expression data were tested for correlation with response or the presence of published KRAS pathway signatures. Additional subgroup analyses were performed. Conclusions: The high prevalence of KRAS mutations in patients with PC has been confirmed using BEAMing. There was a trend towards improved response, mPFS and OS in the wild-type KRAS subset and for KRAS allele frequency to correlate with response. Clinical trial information: NCT01251640.

Endometriosis compromises the quality of life of countless women worldwide and is a leading cause of disability. Clinical symptoms of endometriosis can be very heterogeneous leading to a long interval between onset of symptoms and surgical diagnosis. A noninvasive, rapid diagnostic test is urgently needed. In this prospective study, we evaluated the usefulness of Cytokeratin-19 (CK19) as a biomarker for the diagnosis of endometriosis through urine and serum ELISA. 76 reproductive-aged women undergoing laparoscopy for benign conditions were included to this study and divided into two groups by the presence (n = 44) or absence (n = 32) of endometriosis. There was no statistically significant correlation between the concentration of CK19 in urine (p = 0.51) or in serum (p = 0.77) and the diagnosis of endometriosis. Assigning the samples to the proliferative or secretory cycle stage did not sufficiently lower the p values. In this study, the promising data reported in the recent literature about CK19 serving as a sufficient biomarker for endometriosis could not be verified when tested in a larger sample size. Further studies are warranted to explore the usefulness of CK19 in the diagnosis of endometriosis.

Abstract The AKT1 (E17K) mutation is rare and occurs in colon, ovarian, lung, and especially breast cancer where its frequency ranges between 1.4% and 8.2%. It's precise role in cancer development and progression in clinical context is still unknown. To increase our understanding of the AKT1 (E17K) mutation in breast cancer we analyzed more than 600 tumor samples from breast cancer patients (UICC I - IV, including untreated and neoadjuvantly treated patients) which were provided by the non-profit organization PATH (Patients’ Tumor Bank of Hope, Germany). Extensive clinical data with a median follow-up time of 4.8 years to record disease progression were available for 95% of the patients included in this study. The AKT1 (E17K) mutation was detected in ∼6% of samples in the analyzed cohort using the BEAMing technology. Correlation with clinical parameters showed that the prevalence of the AKT1 (E17K) mutation was statistically independent of age or post-/pre-menopausal stage and was comparable between HER-2 positive and negative patients. In addition, FOUNDATION ONE® targeted exome Next Generation Sequencing (NGS) analysis of some of the tumor samples was done to demonstrate the fingerprint of individual tumors in correlation with the AKT1 (E17K) mutation. NGS and BEAMing technology had a ∼98% concordance for AKT1 (E17K) mutated and non-mutated samples. In 12 out of 36 AKT1 (E17K) mutated samples no additional somatic mutations (SNVs, indels) described to drive cancer development were detected. Moreover, neither amplification nor deletion of tested genes known to be recurrently amplified or deleted in cancer were found in 10 out of these 12 samples. This supports the hypothesis that AKT1 (E17K) can be a driver mutation. However, in all of these samples mutations with yet unannotated function in additional oncogenes were detected. It remains open whether these aberrations impact the role of AKT1 (E17K) as a driver mutation in tumor growth. Analyses of patient cohort data from large databases, as demonstrated here, holds promise for discovering the role of rare somatic mutations in known oncogenes (such as AKT1 (E17K)) in the development of breast cancer. Citation Format: Marion Rudolph, Tobias Anzeneder, Matthias Ocker, Eleni Lagkadinou, Oliver Politz, Martin Michels, Anke Schulz, Georg Beckmann, Michael Teufel, Henrik Seidel, Richie Soong, Heinz Bodenmüller, Ulla Ohlms, Khusru Asadullah, Joachim Reischl. AKT1 (E17K) mutation: coexistence with oncogenic alterations, prevalence, and correlation to clinical parameter in a large series of breast cancer patients. [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 569. doi:10.1158/1538-7445.AM2014-569

4025 Background: Refametinib is a potent oral allosteric MEK 1/2 inhibitor with single-agent and synergistic activities in combination with gemcitabine in preclinical pancreatic cancer (PC) models. We report the results of a single-arm, open-label, phase 2a study combining refametinib and gemcitabine in advanced PC. Methods: Eligibility criteria included: ECOG PS ≤ 2; locally advanced, unresectable or metastatic pancreatic adenocarcinoma; and no prior systemic therapy. Refametinib was administered 50 mg bid po in combination with gemcitabine (1000 mg/m2 IV weekly for 7 of 8 weeks in C1; 3 of 4 weeks in subsequent cycles). The primary objective was overall response rate (ORR); secondary objectives were duration of response (DOR), disease control rate (DCR), time to progression (TTP), progression-free survival (PFS), overall survival (OS) and safety. All responses were confirmed by central independent radiological review. Genetic biomarker analysis was conducted on circulating tumor DNA from plasma samples (BEAMing). Results: Sixty patients were treated: median age 63 yrs, 53% male, 40% PS 0. KRAS mutations were detected in 39 patients (65%). Best overall response was PR in 35% [median DOR 3.8 mo (117 days; 95% CI: 83-265)]; SD in 38%, PD in 10%, and not evaluable in 17% The DCR was 73%. Median TTP was 7.4 mo (224 days; 95% CI: 188, UL ND); median PFS 6.2 mo (190 days; 95% CI: 112-225); OS 8.9 mo (270 days; 95% CI: 200-355). In the KRAS subgroups (mut/WT respectively) the ORR, mPFS and OS were 28%/48% (p=0.136), 4.6/9.0 mo (HR 0.26) and 6.6/18.2 mo (HR 0.27). Full biomarker analysis is being reported separately (Riess et al, this meeting). The most common grade 3-4 TEAEs were neutropenia (43%), thrombocytopenia (22%) anemia (12%), elevations of AST (12%) or ALT (13%), DVT (10%), hypertension (12%), fatigue (15%) and rash acneiform (10%). Conclusions: The combination of refametinib and gemcitabine in advanced PC is active, with an acceptable safety profile. The reported high prevalence of KRAS mutations in PC patients was confirmed by BEAMing. A trend towards improved outcomes (ORR, PFS and OS) in patients with KRAS wild type PC was observed. Clinical trial information: NCT01251640.