Search

Ripretinib is a tyrosine kinase inhibitor (TKI) indicated for the treatment of adult patients with advanced gastrointestinal stromal tumor (GIST) who have received prior treatment with 3 or more kinase inhibitors, including imatinib. In a phase 1 study (NCT02571036), the maximum tolerated dose was not reached at doses up to 200 mg twice daily (BID). Ripretinib 150 mg BID intrapatient dose escalation regimen was offered to patients in the phase 1 and 3 trials (INVICTUS, NCT03353753) after radiologic progression at 150 mg once daily (QD). The BID dose was well tolerated with a similar safety profile to the 150 mg QD (approved dose regimen). In this analysis, the population pharmacokinetics (popPK) of ripretinib and DP-5439, a metabolite TKI with in vitro activity and exposure similar to ripretinib, were characterized and potential covariates influencing exposure were evaluated. The popPK models were developed for ripretinib and DP-5439 using 5284 and 5160 quantifiable concentrations, respectively, from 350 patients pooled from the phase 1 and INVICTUS phase 3 studies. Models were estimated using first-order conditional estimation with interaction method in NONMEM® (version 7.3; ICON, Hanover, MD, USA) and evaluated based on standard goodness-of-fit metrics. A covariate analysis was conducted to assess the sources of variability in ripretinib pharmacokinetics (PK) using a full model approach with backward elimination (significance level = 0.005). Tested covariates included age, body weight, sex, race, tumor type, prior gastrectomy, and mild hepatic impairment and mild to moderate renal impairment. Ripretinib oral PK was well described by a 2-compartment model with zero-order drug release at the absorption site followed by first-order absorption with a modest, linear dose-dependent decrease in relative bioavailability and linear elimination. A high-fat meal was predicted to increase the ripretinib AUC by 36% at 150 mg QD relative to the fasted state. The only covariate effect retained in the final popPK model for ripretinib was 29% lower apparent clearance in females compared with males, resulting in a 40% higher AUC in females. No other covariates were identified for ripretinib. The resulting effect on DP-5439 was similar. PK exposure in patients with mild hepatic impairment or mild to moderate renal impairment was similar to that in patients with normal organ function. There were no clinically meaningful differences in the PK of ripretinib based on age, race, body weight, or tumor type. Based on the safety profile of ripretinib observed in patients with advanced GIST and PK variability, the effects of sex, fasted state, and prior gastrectomy on ripretinib PK were not considered clinically significant. Therefore, ripretinib can be taken with or without food by all patients and no dose adjustment is recommended for patients with mild hepatic impairment or patients with mild to moderate renal impairment. Citation Format: Suzanne George, Filip Janku, Ping Chi, Margaret von Mehren, Neeta Somaiah, Hans Gelderblom, Robin L. Jones, Martine Allard, Adekemi Taylor, Xiaoyan Li, Julie Meade, Vienna Reichert, Rodrigo Ruiz-Soto, Jing Wang, Michael C. Heinrich. Population pharmacokinetics of ripretinib in patients with advanced malignancies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr CT123.

About 50% of inflammatory myofibroblastic tumors (IMFTs) harbor rearrangements of anaplastic lymphoma kinase (ALK), the most common driver in this rare tumor. EORTC 90101 “CREATE” is so far the only prospective, disease-specific phase II clinical trial performed in IMFT. It showed high antitumor activity of crizotinib, an ALK/ROS1/MET-inhibitor. This activity was mainly seen in ALK-rearranged IMFT, but the trial results suggested that other molecular events may also contribute to crizotinib sensitivity. Here we present results of an in-depth molecular analysis of archival IMFTs collected from patients within the CREATE trial with the aim to identify other molecular alterations, which could be responsible for response to crizotinib. DNA from archival IMFTs (primary tumor or metastatic lesion) was used for low-coverage whole genome sequencing to identify genomic regions of gains or losses. Whole exome sequencing (WES) was used to assess the mutational landscape, and mutations affecting cancer consensus genes (CCGs) were analyzed in more detail. Correlative and statistical analysis were used to assess the association between molecular findings and clinical outcomes of IMFT patients. In 24 IMFTs analyzed, the recurrent whole-arm copy number losses were: 22q (58% of cases), 16p (33%), 16q (29%), 13q (29%), 19q (29%), 19p (25%), 6p (25%), 6q (21%) and 18q (21%). Loss of chromosome 19 was found to be associated with shorter progression-free survival in patients receiving crizotinib (p = 0.008). Moreover, 2p21 (including EPAS1, SIX2, EML4 from the CCG set) was frequently amplified (54% of cases), while recurrent losses were observed at 22q12.3 (71%, ISX), 7q36.3 (54%, MNX1), 10q26.3 (54%, MGMT, DUX4), 1p36.32 (50%, RPL22, SKI, TNFRSF14, CAMTA1, PRDM16), 8p23.3 (50%, ARHGEF10), 12q24.33 (46%, POLE), and 11q13.4 (33%, CCND1). No significant correlations were found between focal copy number alterations and crizotinib response. A complex genetic rearrangement involving chromosome 2, identified in 2 ALK-rearranged cases, was likely consistent with chromothripsis, a form of genomic instability related to cancer development. WES was performed in 22 cases with an average of 392 alterations per sample. When considering non-synonymous mutations in CCG, a total of 178 mutations were identified, affecting 143 genes with an average of 7 per case. Mutations in 30 CCGs, identified in ≥ 2 cases, were mainly related to DNA damage/repair mechanism, WNT, RB1-TP53, and RTK-RAS-PI3K signaling. In a case with no ALK rearrangement and ALK immunopositivity, a single base substitution in ALK (p.N571K) was identified. In conclusion, we identified multiple molecular alterations in archival IMFTs and provided further insight in the molecular profile of this ultra-rare malignancy, which may potentially lead to the identification of novel therapeutic targets for IMFT patients. Citation Format: Che-Jui Lee, Patrick Schöffski, Elodie Modave, Bram Boeckx, Diether Lambrechts, Maria Debiec-Rychter, Raf Sciot, Jozef Sufliarsky, Hans Gelderblom, Jean-Yves Blay, Agnieszka Wozniak. Molecular analysis of archival inflammatory myofibroblastic tumor tissue samples from EORTC 90101 “CREATE” and correlation with response to crizotinib [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 794.

Background. Inflammatory myofibroblastic tumor (IMFT) is a rare mesenchymal neoplasm, mainly driven by anaplastic lymphoma kinase (ALK) rearrangement, which is present in 50% of cases and target for ALK inhibition. For the current project we used tumor material from IMFT patients treated with the ALK/ROS1/MET inhibitor crizotinib in the frame of EORTC 90101, where IMFT patients were attributed to ALK+/- sub-cohorts based on the presence (>15% of cells) or absence ( Material and Methods. Archival material from 24 IMFT cases, both primary or metastatic samples, was analyzed by IHC using ALK antibodies: ALK1 (DAKO) and ALK-D5F3 (Cell Signalling), and pan-NTRK (Ventana, clone EPR17341). ALK and ROS1 rearrangement were studied by FISH and fusion genes were detected using the Archer CTL Fusion Panel. Results. ALK immunopositivity by ALK1 and ALK-D5F3 was observed in 14 out of 24 cases and 11/21, respectively. ALK rearrangement by FISH was found in 13/23 samples. With Archer, fusion transcripts were identified in 13/20 specimens, all but one involving ALK with 11 different fusion partners. In one case we detected an ETV6-NTRK3 fusion; NTRK positivity was confirmed by IHC. Interestingly, the NTRK fusion case responded to crizotinib, which shows cone anti-NTRK activity [Okamura et al. JCO Precis Oncol 2019]. We did not detect ROS1 rearrangement by FISH in any sample. Two samples with ALK rearrangement by Archer were negative by FISH, with 11% and 0% of cells with ALK split signal, but both showed protein expression with both antibodies used. They responded to crizotinib with either a RECIST 1.1 complete response or stable disease, for 43 and 9 months, respectively. All but one case with ALK rearrangement by Archer were positive by ALK IHC with ALK1 antibody. The only immunonegative sample had an EML4-ALK1 fusion. Two cases were positive with ALK-D5F3 antibody but no ALK positivity was detected using ALK1, and no ALK rearrangement was detected by FISH or Archer panel, suggesting false positive results. Conclusions. ALK rearrangement is the most common driver in IMFT and the fusion can involve multiple partners. ALK1 but not the ALK-D5F3 antibody shows high ALK specificity in IMFT. The Archer CTL Fusion panel is a reliable and sensitive method for detecting fusion transcripts in IMFT, also allowing to detect alternative fusion genes, which may be responsible for the sensitivity to kinase inhibitors. Citation Format: Agnieszka Wozniak, Che-Jui Lee, Tom van Wezel, Jozef Sufliarsky, Hans Gelderblom, Jean-Yves Blay, Maria Debiec-Rychter, Raf Sciot, Judith V.M.G. Bovee, Patrick Schöffski. Detection of molecular drivers in inflammatory myofibroblastic tumor: study on archival tissue from EORTC 90101 “CREATE” phase II clinical trial [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 3191.

Background: The incidence of brain metastases (BM) in breast cancer (BC) patients (pts) has increased over the past decade. The brain is regarded as a sanctuary site for metastatic cells which are partially protected from drugs by the blood-brain barrier (BBB). 2B3-101 is a doxorubicin (DOX) liposomal formulation that uses glutathione transporters on the BBB to penetrate the brain. Non-clinical studies have shown a 5-fold enhanced delivery of DOX to the brain after IV administration of 2B3-101 compared to liposomal DOX, without signs of cardio- or neurotoxicity. Methods: This phase 1/2a open label study assessed the safety, tolerability, MTD, PK, and anti-tumor activity of single agent 2B3-101 in pts with BM of solid tumors, or high-grade gliomas. For this analyses BCBM pts (n=25) were included. These pts received 2B3-101 at a starting dose of either 40 (n=3) or 50 (n=22) mg/m2 IV every 3 weeks, until disease progression or unacceptable toxicity. Anti-tumor activity was assessed by RECIST 1.1. Patients with HER2-positive BCBM were treated with concurrent trastuzumab. Results: As of May 30, 2014, 88 cycles (median 2, range 1-10) of 2B3-101 alone or with trastuzumab were administered to 25 heavily pretreated BC pts, 3 pts are still on treatment. Median age was 47 (33–61) years and 18 (72%) pts had HER2+ disease. Pts had received a median of 7 (4–15) prior regimens; 18 (72%) had received prior radiation therapy to the brain. In phase 1, 2B3-101 alone or with trastuzumab was well tolerated up to a dose intensity of 15 mg/m2/wk. Cycle 1 MTD was not reached. Phase 2a dose of 50 mg/m2 was selected based upon tolerability after repeated dosing. The most frequent reported treatment emergent AEs are qualitatively consistent with conventional PEGylated liposomal DOX and were (≥ grade 2): neutropenia (59%), palmar plantar erythrodysesthesia (PPE) (50%), fatigue (45%), stomatitis (22%), and infusion reaction (18%). Notable Grade 3–4 AEs were neutropenia (35%) and PPE (13%). All AEs were transient and manageable with dose delays, reductions and standard medication. 57% of pts had 2B3-101 dose delays and 39% of pts required dose reductions. 2B3-101 showed no neuro- or cardiotoxicity. PK data showed non-linear exposure of 2B3-101 without signs of accumulation upon repeat dosing, a mean half-life of 69h (range 43-120h) and independent of trastuzumab co-treatment. Best overall and intracranial tumor responses of 2B3-101 in 23 evaluable BCBM pts (92% of total) are summarized in Table 1. Table 1: Anti-tumor activity results of 2B3-101 in BCBM patientsPopulationNumber of evaluable ptsOverall partial responseOverall stable diseaseIntracranial tumor response of ≥ 20%.12-weeks overall PFS rateOverall232 (9%)11 (48%)4 (17%)48%"Luminal"40 (0%)2 (50%)1 (25%)50%HER2 positive162 (13%)9 (56%)3 (19%)56%TNBC30 (0%)0 (0%)0 (0%)0% Conclusions: 2B3-101 alone or with trastuzumab is safe and well tolerated and shows intra- and extracranial anti-tumor activity in heavily pretreated BC pts. A 12-week PFS rate of 56% in HER2+ BCBM was observed, which warrants further clinical studies. An international, multicenter, randomized, controlled phase IIb study in HER2+ BCBM is being planned. NCT01386580, sponsored by to-BBB technologies BV. Citation Format: Philippe G Aftimos, Bojana Milojkovic-Kerklaan, Véronique Diéras, Sevilay Altintas, Carey Anders, Monica Arnedos, Hans Gelderblom, Patricia Soetekouw, Werner Gladdines, Pieter Gaillard, Carlos de Sousa, Agnes Jager, Myra van Linde, Ahmad Awada, Jan Schellens, Dieta Brandsma. Phase 1/2a study of glutathione PEGylated liposomal doxorubicin (2B3-101) in breast cancer patients with brain metastases [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P6-16-04.

Background: MBG453 and spartalizumab, humanized IgG4 mAbs, block binding of TIM-3 to PtdSer and PD-1 to PD-L1/2, respectively. Preclinical studies show synergistic anti-tumor activity of TIM-3 and PD-1 co-blockade. Here we report the dose escalation results from a Ph I-Ib/II study of MBG453 ± spartalizumab in metastatic solid tumors (NCT02608268). Methods: Metastatic cancer pts received intravenous MBG453 alone at 80-1200 mg Q2W or Q4W, or combination therapy (Q2W/Q4W) with MBG453 (20-800/80-1200 mg) + spartalizumab (80-240 mg/80-400 mg). The recommended Ph II dose (RP2D) was determined using an adaptive Bayesian logistic regression model guided by escalation with overdose control together with Ph I-Ib endpoints (dose-limiting toxicity [DLT] at 8 weeks). Results: As of July 26, 2018, 87 pts received MBG453 alone (14% were anti-PD-1/PD-L1 pre-treated; common tumors [≥9%]: pancreatic cancer [11%], sarcoma [10%], and colorectal cancer [CRC; 9%]) and 86 pts received MBG453 + spartalizumab (27% were anti-PD-1/PD-L1 pre-treated; common tumors [≥6%]: melanoma [8%], NSCLC [7%], CRC [7%], and ovarian cancer [6%]). One DLT (myasthenia gravis grade [G] 4) was reported in a pt with thymoma treated with MBG453 (240 mg Q4W) + spartalizumab (80 mg Q4W). Treatment-related adverse events (AEs) were reported in 40% and 59% of pts treated with MBG453 and MBG453 + spartalizumab, and G3/4 AEs in 0% and 11% of pts; the most common AE (≥10%) was fatigue in 10% and 15% of pts, respectively. MBG453 exposure generally increased in a dose-proportional manner. Maximum tolerated doses were not identified with the tested dose/schedule. MBG453 800 mg Q4W (n=9 pts treated) and MBG453 800 mg + spartalizumab 400 mg Q4W (n=6 pts treated) were declared as RP2Ds. Stable disease (SD) was seen in 25/87 (29%) pts treated with MBG453 alone (common tumors [≥2 pts]: sarcoma [n=5], and breast cancer, CRC, ovarian cancer, cholangiocarcinoma, and NSCLC [n=2 pts each]); 4 of 25 pts with SD were anti-PD-1/PD-L1 pre-treated. Of 86 MBG453 + spartalizumab pts, partial responses were seen in 4 pts (5%); 1 anti-PD-1/PD-L1 pre-treated pt (out of 6 with NSCLC; DOR: 392 d), 3 anti-PD-1/PD-L1 naïve pts (CRC [n=2 out of 6, DOR: 223 d, 109 d]; SCLC [n=1 out of 3; DOR: 112 d]). SD was seen in 34/86 (40%) MBG453 + spartalizumab pts (common tumors [≥3 pts]: melanoma [n=5; 2 cutaneous, 2 uveal, and 1 non-cutaneous], ovarian cancer [n=3], and urothelial carcinoma [n=3]); 10 of 34 pts with SD were anti-PD-1/PD-L1 pre-treated. RNAseq analysis of screening and on-treatment biopsies revealed a pharmacodynamic trend of increased IFN-γ-associated gene signatures following MBG453 + spartalizumab treatment. Conclusions: MBG453 + spartalizumab was well tolerated with preliminary signs of anti-tumor activity. MBG453 800 mg Q4W and MBG453 800 mg + spartalizumab 400 mg Q4W were selected as the RP2Ds; dose expansion is ongoing in pts with melanoma or NSCLC resistant to anti-PD-1/PD-L1. Citation Format: Giuseppe Curigliano, Hans Gelderblom, Nicolas Mach, Toshihiko Doi, Wai Meng David Tai, Patrick Forde, John Sarantopoulos, Philippe L. Bedard, Chia-Chi Lin, Stephen Hodi, Sofie Wilgenhof, Armando Santoro, Catherine Sabatos-Peyton, Tyler Longmire, Kitty Wan, Panagiotis Nikolopoulos, Luigi Manenti, Aung Naing. Phase (Ph) I/II study of MBG453± spartalizumab (PDR001) in patients (pts) with advanced malignancies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr CT183.

Objectives: Ripretinib is a novel, oral kinase switch control inhibitor of KIT and PDGFRα. Encouraging clinical benefit from the phase 1 dose escalation and expansion study as measured by ORR, DCR and PFS in 2nd, 3rd, and >4th line GIST patients with a favorable tolerability profile at doses >100 mg/day has been previously reported (ESMO 2018, abstract #1603O). It has been reported that more than 40% of GIST patients use acid-reducing agents. PPIs are the most potent acid-reducing agents that may impair the absorption of kinase inhibitors 1,2. This retrospective analysis aims to explore whether ripretinib can be used regardless of concomitant PPI use. Methods: The analysis assessed the impact of PPIs on the plasma concentration of ripretinib using PK data from the expansion cohort at the recommended Phase 2 dose of 150 mg QD. Plasma concentrations of ripretinib and its active metabolite DP-5439 obtained from patients who used or did not use PPIs were compared on Cycle 1 Day 1 (C1D1, n=106) and Day 15 (C1D15, n=102). Patients using PPIs were defined as those who continuously took PPIs for at least 4 days prior to C1D1 or C1D15. Patients who did not use PPIs were defined as those who did not take PPIs or any other acid-reducing agents during the study. Results: PK profiles were consistent between patients using and not using PPIs (Table 1), indicating a low likelihood of a clinically significant drug interaction between PPIs and ripretinib. Conclusions: This retrospective PK analysis provides supporting evidence that restriction of co-administration of PPIs with ripretinib may not be necessary. A dedicated drug interaction study is planned to provide a definitive assessment. References: 1. Smelick et al, Mol. Pharmaceutics 2013, 10, 4055−4062 2. Budha et al, Clin Pharmacol Ther. 2012, 92(2):203-13 Table 1.PK Exposure of Ripretinib and DP-5439 in Patients Using or not Using PPIsPK ConcentrationsRipretinib in ng/mL [mean (CV%)]DP-5439 in ng/mL [mean (CV%)]Ripretinib + DP-5439 in ng/mL [mean (CV%)]Using PPIsNot using PPIsUsing PPIsNot using PPIsUsing PPIsNot using PPIsC1D1 6 hr566 (58%) n=24670 (53%) n=82302 (64%) n=23297 (59%) n=82862 (54%) n=24975 (48%) n=82C1D15 pre-dose364 (79%) n= 24344 (63%) n=78960 (80%) n=24889 (86%) n=781350 (72%) n=241260 (75%) n=78C1D15 6 hr834 (51%) n=24871 (47%) n=731170 (69%) n=241060 (67%) n=732040 (53%) n=241960 (49%) n=73Notes: Table 1 is based on data from patients without a history of gastrectomy. A separate analysis in patients with gastrectomy will be published after sufficient data is obtained. Citation Format: Filip Janku, Michael Heinrich, Ping Chi, Albiruni Abdul Razak, Margaret von Mehren, Michael Gordon, Kristen Ganjoo, Jonathan Trent, Robin L. Jones, Hans Gelderblom, Kelli Running, Jing Wang, Rodrigo Ruiz-Soto, Suzanne George. Ripretinib (DCC-2618) pharmacokinetics (PK) in a Phase I study in patients with gastrointestinal stromal tumors (GIST) and other advanced malignancies: A retrospective evaluation of the PK effects of proton pump inhibitors (PPIs) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr CT058.

Background: The multi-tumor phase 2 trial EORTC 90101 (NCT01524926) assessed the activity and safety of the ALK/MET/ROS1 inhibitor C in IMFT, an orphan malignancy associated with ALK rearrangement or copy number changes. Methods: Pts with local diagnosis of advanced/metastatic IMFT consented for shipment of a tumor tissue block and were screened for treatment after central confirmation of the diagnosis by reference pathology. Eligible ALK positive (+) and negative (-) pts received oral C 250 mg bid until RECIST 1.1 progression (PD). ALK+ was defined as at least 15% of tumor cells with rearrangement on FISH (Vysis LSI ALK Dual Color Break Apart Probe, Abbott Molecular) and/or immunohistochemical positivity (ALK MAb Clone CD246, DAKO). A Simon's optimal 2 stage design was implemented. If at least 2 out of the first 12 eligible and evaluable ALK+ pts achieved a confirmed RECIST 1.1 partial or complete response (PR, CR), a maximum of 35 pts were enrolled. If at least 6 had a confirmed PR/CR, the trial would be deemed successful. Results: Between 10/2012 and 04/2017, 13 sites in 8 European countries recruited 35 pts with a local diagnosis of IMFT, of whom only 20 had centrally confirmed IMFT and were treated with C. Among 12 eligible and evaluable ALK+ pts, 6 achieved a confirmed PR or CR, 1 a non-confirmed PR and 5 had stable disease (SD) as best response (ORR 50.0%, 95% confidence interval: 21.1-78.9%). Further efficacy endpoints in ALK+ pts: disease control rate (DCR = CR/PR/SD as best response) 100.0% (73.5-100.0%), 1-year (y) progression free rate (PFR) 73.3 % (37.9-90.6%), 1-y OS rate (OSR) 81.8% (44.7-95.1%). Among 7 eligible and evaluable ALK- pts, 1 achieved a PR, 5 had SD and 1 PD (ORR 14.3%, 0.0-57.9%). Further data in ALK- cases: DCR 85.7% (59.8-100.0%), 1-y PFR 53.6 % (13.2-82.5%), 1-y OSR 83.3% (27.3-97.5%). One additional ALK- case was non evaluable (no measurable disease at baseline). Common related adverse events were nausea (11/20 [55%]), fatigue (9/20 [45%]), blurred vision (9/20 [45%]), vomiting (7/20 [35%]), diarrhea (7/20 [35%]). Conclusions: EORTC can perform precision medicine phase II trials in ultra-rare cancers such as IMFT, with mandatory collection of tissue, real time reference pathology and genetic profiling. With an ORR of 50% and a DCR of 100% in ALK+ disease, C met pre-specified response rate criteria in this trial. The drug achieves long-lasting disease control in the vast majority of ALK+ pts. Sporadic responses and disease stabilization in ALK- cases either suggest limitations of the assays and their cut-offs for target positivity, or the presence of other oncogenic drivers/alternative fusions that may be sensitive to C. Based on the findings of this prospective trial, C should be considered as systemic treatment standard of care for this orphan disease. Citation Format: Patrick Schoffski, Jozef Sufliarsky, Hans Gelderblom, Jean-Yves Blay, Sandra J. Strauss, Silvia Stacchiotti, Piotr Rutkowski, Lars H. Lindner, Michael G. Leahy, Antoine Italiano, Nicolas Isambert, Maria Debiec-Rychter, Raf Sciot, Thomas Van Cann, Sandrine Marréaud, Axelle Nzokirantevye, Sandra Collette, Agnieszka Wozniak. Prospective precision medicine trial of crizotinib (C) in patients (pts) with advanced, inoperable inflammatory myofibroblastic tumor (IMFT) with and without ALK alterations: EORTC phase II study 90101 "CREATE" [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr CT045.

Introduction Gastrointestinal stromal tumors (GISTs) are rare malignancies of the gastrointestinal tract. In the Netherlands, most patients with GIST are treated in five sarcoma expertise centers. GIST is known to have driver mutations in the tyrosine kinase receptors KIT and PDGFRα in respectively 80% and 10% of the patients. Mutations (single nucleotide variations, insertions and deletions) are mostly found in KIT exon 9 and 11. GIST patients with a primary KIT exon 11 mutation respond generally well to tyrosine kinase inhibitors (TKIs). The aim of this study was to evaluate treatment response by the detection of KIT exon 11 mutations in cell free DNA (cfDNA) of patients with GIST at multiple time points. Materials and methods Since Dec 2014, plasma samples of patients have been collected at every hospital visit and were put in a bio-databank for analysis of mutation detection in cfDNA. Eligibility criteria were histological diagnosis of GIST and treatment with a TKI. Patients with a known KIT exon 11 mutation, of which a baseline (before start of any TKI) sample or a recurrence/progression sample was available, were analyzed. A custom single-tube digital droplet PCR (ddPCR) drop-off assay, designed at the UMCG, was used. This assay detects 80% of the known KIT exon 11 mutations. Results Until Nov 2016, >620 samples of 175 patients have been collected. From 43 patients with a KIT exon 11 mutations in the pretreatment tumor biopsy, a baseline/progression sample was available. Mutations in exon 11 were detected in cfDNA from baseline plasma of 22/43 patients, 20 with metastasized and 2 with localized disease. Of the 21 patients without detectable mutations, 11 had localized and 10 metastatic disease. Three of these patients with metastatic disease had mutations that could not be detected by our probe. Two weeks after start of TKI, an increase in mutant allelic frequency was seen in 5 (of 9 available 2 week samples) patients, while 4 weeks later the mutant allelic frequency had decreased, in parallel with decrease in tumor load on radiological imaging. Eight patients developed radiological disease recurrence (2 patients, after 4 and 20 months) or progression (6 patients, median 10 (6-24) months). This was paralleled by an increase in mutant allelic frequency in cfDNA in 6 of them (baseline median 0% (0-0.9%); at progression median 3.5% (0-30%); p= 0.028; related samples Wilcoxon signed rank test). Summary and conclusion KIT exon 11 mutations in cfDNA could be detected in 74% of the patients with metastatic disease. Disease progression or recurrence, coincides with a rise of mutant alleles detected in the plasma. In conclusion, mutation detection in cfDNA of GIST patients with metastatic disease is feasible. Our study will be extended to include the monitoring of early progression based on cfDNA, which may guide early treatment adaptations. Citation Format: Pieter A. Boonstra, Arja ter Elst, Marco Tibbesma, Ron H. Mathijssen, Florence Atrafi, Frits van Coevorden, Sheima Farag, Neeltje Steeghs, Ingrid M. Desar, Winette T. van der Graaf, Hans Gelderblom, Boudewijn van Etten, Jelle Overbosch, Albert J. Suurmeijer, Jourik A. Gietema, Ed Schuuring, Anna K. Reyners. Dynamics of KIT exon 11 mutations in cell free plasma DNA of patients treated for advanced gastrointestinal stromal tumors: Results from the Dutch GIST bio-databank [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 4951. doi:10.1158/1538-7445.AM2017-4951

Background: GRID is a phase III study for patients with advanced gastrointestinal stromal tumors (GIST) following failure of imatinib (I) and sunitinib (S) who were randomized to receive either the multikinase inhibitor regorafenib (R) or placebo (P). R demonstrated a highly significant improvement in progression-free survival compared with P (HR 0.27, p Methods: DNA was isolated from archival tumor tissue and analyzed for KIT mutations via Sanger sequencing. The expectation was that primary KIT mutations would be detectable in archival tissue but that secondary KIT mutations may be undetectable in tissues obtained before treatment with I or S. To overcome this potential limitation, plasma samples drawn at GRID study entry, post I and S failure, were used as a source of circulating DNA for evaluation of GIST oncogenic mutations (KIT, PDGFRA, BRAF) via BEAMing technology. Results: KIT mutations were detected in 83 of 138 (60%) plasma samples and 64 of 99 (65%) tumor tissue samples analyzed. Primary KIT exon 11 and 9 mutations were identified in approximately 42% and 18% of the tissue samples, respectively. The frequency of the canonical exon 9 mutations was similar for plasma and tissue samples, showing consistency between mutation-detection technologies. With limitations of tumor-based assays, a lower incidence of secondary KIT resistance mutations was detected in patient-matched archival tumor tissue compared with plasma samples: resistance mutations were detected in 12% of tissue samples vs 48% of plasma samples. Most (76%) secondary KIT mutations detected in plasma DNA were located in the KIT activation loop encoding structural alterations known to mediate resistance to I and S. Nearly half of the plasma samples in which secondary KIT mutations were identified harbored multiple secondary mutations, consistent with the results of previous studies on fresh tumor biopsies taken following resistance to both I and S. R was clinically active compared with P in all KIT mutational subgroups evaluated (HR 0.27 in patients with KIT exon 9 mutations; HR 0.25 in patients with secondary KIT mutations identified via plasma DNA). Conclusions: In GIST patients from the GRID trial, driver oncogenic mutations and secondary oncogenic mutations leading to I and S resistance are readily detectable via BEAMing of circulating DNA from plasma. BEAMing may provide a real-time assessment of tumor genotype in GIST and other cancers using blood-derived circulating DNA, that may be more comprehensive than tumor sampling. GIST patients with a wide spectrum of primary and secondary mutations in oncogenic kinases benefit from treatment with R. Citation Format: George D. Demetri, Michael Jeffers, Peter Reichardt, Yoon-Koo Kang, Jean-Yves Blay, Piotr Rutkowski, Hans Gelderblom, Peter Hohenberger, Michael Leahy, Margaret von Mehren, Heikki Joensuu, Giuseppe Badalamenti, Martin Blackstein, Axel Le Cesne, Patrick Schöffski, Robert G Maki, Sebastian Bauer, Binh Bui Nguyen, Jianming Xu, Toshirou Nishida, John Chung, Chetan D. Lathia, Christian Kappeler, Iris Kuss, Dirk Laurent, Paolo G Casali. Detection of oncogenic kinase mutations in circulating plasma DNA and correlation with clinical benefit in the phase III GRID study of regorafenib vs placebo in TKI-refractory metastatic GIST. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr LB-295. doi:10.1158/1538-7445.AM2013-LB-295

Chondrosarcomas (CS) are malignant cartilage-forming tumors of bone. High grade chondrosarcomas show metastasis formation in 71% of cases, for which currently no treatment strategies exist. Kinome profiling revealed high GSK3β, AKT, and Src pathway activity. These kinases play a role in either cell survival, migration, or both. Our aim was to identify the roles of these pathways in chemoresistance and migration and investigate downstream effects while pinpointing the most important kinase in chondrosarcoma. We used 5 conventional chondrosarcoma cell lines and 3 dedifferentiated chondrosarcoma cell lines for the experimental procedures. Kinase inhibition was performed with enzastaurin (GSK3β and AKT inhibition) and dasatinib (Src inhibition). WST and Live cell imaging with AnnexinV staining (BD Pathway 855 Bioimager) were performed to investigate cell viability and apoptosis formation. Proliferation and migration assays were performed with the RTCA xCelligence System (Roche). Tissue microarrays (TMAs) were constructed containing 8 EC, 92 central CS (grade I n=42, grade II n=35, grade III n=14), 11 OC and 45 peripheral CS (grade I n=31, grade II n=11, grade III n=3). Overexpression of Src family members can lead to excess Nrf2 translocation from the nucleus. Immunohistochemistry was performed for the Src family members and GPX3, the target of the transcription factor Nrf2. Western blot was performed for Src family members, Nrf2, and GPX3. Inhibition of GSK3β and Akt with enzastaurin in chondrosarcoma cell lines was ineffective, also when combined with Src inhibition (dasatinib) or doxorubicin. Combination treatment of dasatinib with doxorubicin, however, showed synergistic loss in cell viability and apoptosis formation, with cleaved caspase3 activation, which was not observed after dasatinib single treatment. Migration assays in cell lines (n=6) revealed that at low concentrations of dasatinib (200nM), migration was successfully inhibited. TMA staining revealed sporadic expression of Yes and Lck (5%), expression of Src (57%) and abundant expression of Fyn (78%) throughout our panel of CS tissues. Fyn is described to be associated with metastasis formation and is a good candidate to play a role in chondrosarcoma progression. Overall pSrc (active Src) approached 100% in all grades. Western blot analysis showed nuclear localization of Nrf2 and restoration of GPX3 transcription after dasatinib treatment of cell lines. We show that Src family kinases rather than GSK3β and Akt kinases contribute to chemoresistance in chondrosarcoma. Src kinase inhibition successfully prepared cells for chemotherapy and acts synergistically with doxorubicin, and moreover was able to completely inhibit chondrosarcoma cell migration and restore GPX3 transcription. These results strongly indicate Src family kinases, and in particular Fyn, to be a potential target for the treatment of inoperable and metastatic chondrosarcomas. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 828. doi:1538-7445.AM2012-828