First-in class selective AXL inhibitor bemcentinib (BGB324) in combination with LDAC or decitabine exerts anti-leukaemic activity in AML patients unfit for intensive chemotherapy: Phase II open-label study

7043

Background: The RTK AXL represents a therapeutic target promoting AML cell proliferation and survival by pleiotropic mechanisms and is a negative regulator of anti-tumour immunity. Bemcentinib is a first-in-class, highly selective, oral AXL inhibitor that has previously shown encouraging anti-leukaemic activity as a monotherapy in r/r AML and hr-MDS. Methods: A monotherapy dose-escalation and expansion part of this trial is complete. In this second, phase II part of the study, 11 and 15 AML pts unfit for intensive chemotherapy received bemcentinib at RP2D (200 mg po/d) in combination with low-dose cytarabine (LDAC) and decitabine, respectively. Median age was 77 yr (range: 50-83), median screen myeloblast count 39% (3-95%) and 2/19 (11%) of pts evaluable for FLT3 were FLT3+. Plasma protein biomarker levels were measured using the DiscoveryMap v3.3 panel (Myriad RBM) at screen and following treatment. Results: The most common TRAEs (≥ 15% of pts) were ECG QT prolongation (35%) and diarrhoea (15%). Among these, 3 were Grade 3, and none 4 or 5. All TRAEs were manageable and/or reversible. As of Feb ‘19, 9 pts (2 de novo, 1 secondary, 6 r/r) in the bemcentinib + LDAC group were evaluable for response and 4 (44%; 2 de novo + 2 relapsed) achieved rapid CRi at C2D1. Responses were durable (range: 7 – 11 cycles) in 3 of the 4 responders. A further 2 pts (22%, 1 secondary + 1 relapsed) achieved durable SD (5 and 6 cycles). mPFS among the 5 pts with durable CRi or SD was 5 months (range: 3.5-7.7). Further, at the time of writing, 11 pts (8 de novo, 3 r/r) in the bemcentinib + decitabine group were evaluable for response of which 4 (36%, all de novo) achieved CRi after ≥ 4 cycles. One additional de novo pt achieved durable SD lasting for 5 cycles. Conclusions: Bemcentinib in combination with LDAC exerted early durable responses in patients with both de novo and relapsed AML whilst the combination of bemcentinib + decitabine exerted comparably fewer and later responses in de novo AML. Soluble biomarker correlations will be presented at the meeting. Both combinations were generally well-tolerated and further exploration is warranted. Clinical trial information: NCT02488408.

The immunomodulatory activity of bemcentinib (BGB324): A first-in-class selective oral AXL inhibitor in patients with relapsed/refractory acute myeloid leukemia or myelodysplastic syndrome

70

Background: The tyrosine kinase AXL acts as suppressive immune checkpoint of the innate immune system. AXL signalling leads to immunosuppression and tumour immune escape by downregulating dendritic cell activity, modulating efferocytosis as well as favouring an immunosuppressive chemokine profile and M-MDSC expansion. Therefore it represents an interesting novel immune oncology target. Bemcentinib (BGB324) is a first-in-class, highly selective and orally bioavailable small molecule AXL inhibitor in phase II clinical development. In pre-clinical models, inhibition of AXL signalling with bemcentinib reversed multiple established immune suppressive mechanisms leading to increased infiltration of CTLs, NK and NKT cells and decreased infiltration of M-MDSCs (Wnuk-Lipinska, 2017). We explored safety, PK, efficacy and the effects of treatment with bemcentinib on the T- and B cell repertoire in patients with relapsed AML or MDS. Methods: BGBC003 (NCT02488408) is an open-label, dose finding study of bemcentinib in patients with R/R AML or MDS. Three dose levels have been explored and a loading dose of 400 mg on days one to three followed by 200 mg daily thereafter has been established as safe and recommended phase 2 dose. The TCRß repertoire was quantified by NGS of DNA isolated from PBMNCs using an Illumina MiSeq sequencer. TCRß genes and the IGH repertoire were analysed with BIOMED2-TCRß-A and –B and BIOMED2-FR1/-FR3 primer pools, respectively. Using genomic DNA as template, the amplicons were tagged with Illumina adapters and indices in two consecutive PCR reactions. Demultiplexing and FastQ formated data output was generated by the MiSeq reporter. Analysis of TCRß and IGH data was performed on a Microsoft Cloud using our in-house analysis pipeline Pippa. Results: Early evidence of antileukemic activity was seen. Diversification of the TCR and IGH repertoire could be detected in peripheral blood at day 21 of therapy compared to pre-treatment in three and two out of five matched samples, respectively. Conclusions: Bemcentinib shows preliminary evidence of immune modulation in AML. Clinical trial information: NCT02488408.

Efficacy of clofarabine and low-dose cytarabine combination for genetically high-risk acute myeloid leukemia prior to allogeneic stem cell transplantation

e18515

Background: The outcome of genetically high risk acute myeloid leukemia (AML) patients (pts) remains poor, with a median overall survival (OS) of a few months after the failure of regimens. Long-term survival requires complete remission followed by allogeneic stem cell transplantation (ASCT).Clofarabine, a deoxyadenosine analog, at 20mg/m2 days 1-5, in combination with low-dose cytarabine (LDAC) has reported efficacy and tolerability in older newly diagnosed AML pts (Cancer 2015;121:2375). We have therefore prioritized this combination for relapsed/refractory(R/R) pts. Aims:Determine the efficacy of Clofarabine and low dose cytarabine in R/R AML pts who are eligible for ASCT. Methods: Using Holden Comprehensive Cancer Center database, we performed a retrospective chart review on 16 pts of R/R AML who were treated with Clofarabine and LDAC between January 2014 and Dec 2016. Each patient received Clofarabine 20mg/m2on days 1-5 IV and LDAC on days 1-10 subcutaneous. The primary endpoint was complete remission (CR) or complete remission with incomplete marrow recovery (CRi). Secondary endpoint was median OS. Results: Sixteen pts were analyzed. The median age was 58 years (range 27-78). Nine pts (56%) had AML with high risk characteristics (complex karyotype and/or presence of FLT3-ITD). Nine pts achieved either a CR or CRi. Responses were observed in 7of 9 pts with genetically high-risk features. 2 pts had mutated TP53, and both of them achieved CR. The median OS time was 13.25 months (range 3.6-19.2 months). 4 pts received ASCT after treatment combination. Conclusions: These results suggest that Clofarabine is well-tolerated and effective in the treatment of R/R AML pts, especially those with high-risk genetics. Clofarabine and LDAC could be used as a bridging therapy prior to ASCT for high-risk R/R AML pts. Future clinical trials are warranted to explore additional modifications to this combination in order to optimize therapy for this group of high-risk and heavily treated R/R AML pts.

A biomarker-directed phase 2 trial of SY-1425, a selective retinoic acid receptor alpha agonist, in adult patients with acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS)

TPS7071

Background: SY-1425 (tamibarotene) is an orally available, synthetic retinoid approved in Japan for the treatment of relapsed/refractory (R/R) APL. SY-1425 is a more potent and selective retinoic acid receptor alpha (RARα) agonist with improved pharmacologic properties compared to all-trans retinoic acid (ATRA) including increased half-life and lack of metabolism by CYP26A1 resulting in extended relative exposures. SY-1425 binding to RARα relieves pathogenic repression of myeloid differentiation. Super-enhancers associated with RARA and upregulation of RARA expression correlate with increased sensitivity to SY-1425 in vitro and predict for response to SY-1425 with induced differentiation and reduced proliferation in RARA-high PDX AML models, but not in RARA-low models. SY-1425 also induces the RARα target gene DHRS3 in RARA-high AML cell lines. This study is designed to demonstrate pharmacodynamic (PD) and clinical effects of SY-1425 in non-APL AML and MDS patients (pts) positive for the RARA super-enhancer associated biomarker or exploratory RARA pathway biomarker, IRF8. Methods: This study is enrolling pts with R/R AML, R/R higher-risk MDS, newly-diagnosed AML ≥60 yrs unlikely to respond to or tolerate standard therapy, and transfusion dependent lower-risk MDS pts without del 5q who are unlikely to respond to or have failed ESAs. Pts must be biomarker positive based on centralized testing of tumor cells from blood. All pts receive SY-1425 at 6 mg/m2/day PO with continuous twice daily dosing. Primary objectives are to characterize the activity of SY-1425 by ORR in AML and higher-risk MDS pts or transfusion independence in lower-risk MDS pts. Secondary objectives include event-free and relapse-free survival, duration of response, overall survival, hematologic improvement and safety. PD evaluation includes induction of DHRS3 and expression of myeloid differentiation markers. Target enrollment is 80 pts. This trial opened in September 2016. Through a protocol amendment, SY-1425 treatment in combination with azacitidine will also be evaluated. ClinicalTrials.gov identifier: NCT02807558.

Comparison of two major intergroup induction intensive regimens in Hispanic adolescent and young adults (AYA) with acute lymphocytic leukemia (ALL): The importance of a complete molecular response

7072

Background: Novel intensive chemotherapy regimens impacted in the prognosis of adolescent and young adult pts with ALL. The application of pediatric inspired therapies demonstrated complete response reported to be > 80%, and long-term survival rates range from 30% to 45%. In the past, ALL treatments had been designed on the presence of risk factors as elevated WBC, time to complete remission or immunophenotype. Recently, MRD monitoring had risen as a risk predictor. Hypothesis: Complete response(CR) associated with a CMR has impact in disease-free survival. Methods: Between March> 2010 and November 2012, 101 adolescent and young adult pts were diagnosed and treated for (Ph)–negative ALL with the C10403 intergroup protocol (N=47) and the 2008 Spanish PETHEMA regimen (N=54). The residual disease status was assessed after induction by multiparametric flow cytometry (MRD <0.1%). Results: The median age of the pts was 17.56 years (range, 14-24); with 69 pts (68%) ≤ 18 year-old. 27 (26.7%) pts were WBC ≥ 50x109/L, 63 pts (62%) B immunophenotype, cytogenetics was unknown in 69 pts. After the induction, the CR was achieved in 81% vs 79.6% pts ( Int C10403 vs Pethema, respectively ). The induction mortality rate was 0% vs 7%. Slow responders to induction chemotherapy were 23.81% and 28.8% respectively (p=0.591). A CMR (CR+MRD ≤0.1%) was seen on 70.21% vs 63.27%. The kaplan-meier 2-year disease free-survival rate calculated for CMR patients on C10403 was 35% vs 17% for the Pethema, which was statistically significant (p< 0.04). A multivariate analysis of pts characteristics in CMR could not identified risk factors for disease prognosis. Conclusions: Two different regimens C10403 vs Pethema, could not demonstrate a statistic difference in morphological CR. Lately,the addition of minimal residual disease ≤0.1% (MRD) in the analysis, demonstrated a plausible monitoring variable of response. We suggest that a CMR (CR +MRD at induction) could be a new standard variable to assess responses regardless of previous risk categories. Further validation may be mandatory.

Breast cancer incidence in a large cohort of Peruvian women affiliated to a pre-paid system

e12570

Background: Breast cancer (BC) is the second most common malignancy and the leading cause of death by cancer in Peruvian women (age-standarized rate [ASR] of 34 new cases/100,000 women estimated by GLOBOCAN 2008). The purpose of this study was to assess the incidence of BCin acohort ofwomenat Oncosalud, an oncologic pre-paid system that currently has 600,000 affiliates. Methods: We evaluated a dynamic cohort (period 1989 to 2011) of women affiliatedat Oncosalud – AUNA, an oncologic prepaid system.The crude incidence rate per year (number of new cases/women at risk), the specific rate according to age (number of new cases / persons-year) and cumulative risk were calculated. Results: Overall, during the assessment period, the BC incidence rate per year was 175.6 and the ASR incidence was 111.9 per 100,000 affiliates respectively. In our cohort of affiliates there were no BC cases before 1993 (with 907 women at risk for that year). The highest incidence rate was 177.6 registered in 1997 (11,822 women at risk). Incidence rates started decreasing in 2003 (169.2 with a population at risk of 39,593 women). The lowest incidence was 71.5, registered in 2011 (279,680 women at risk).According to age-groups, there were no BC cases under20 years old. Specificincidence ratesper age-group increases from the 30 year old-group (55.8). The peak of BC incidence was between 70 to 74 years old (407.4). In the same way, the cumulative risk increases after 30 years old. Conclusions: In our cohort of affiliates, the incidence of BC is greater than the general population, it could be due to the process of negative selection; however, specific incidence rates per age-group and cumulative risk are increased after 30 years, as seen in the general population.

Abstract P6-07-20: Variables measured at Central Nervous System (CNS) relapse, but not Immunophenotype, identify groups of breast cancer patients with shorter post CNS-relapse survival

Objective: We evaluated breast cancer immunophenotype and variables measured at CNS-relapse as prognostic factors in a cohort of Hispanic patients.

Methods: We reviewed data from 2602 breast cancer women in stages I-III diagnosed between 2000 and 2005 at the Instituto Nacional de Enfermedades Neoplásicas, Lima, Perú. According to immunophenotype, tumors were categorized in luminal A (RE+ and/or RP+, HER2−), luminal B (RE+ and/or RP+, HER2+), HER2 (RE−, RP−, HER2+) and triple negative (RE−, RP−, HER2−). Clinicopathological data at diagnosis and at CNS relapse were evaluated and used to calculate prognostic scores according to Recursive Partitioning Analysis (RPA) score, Graded Prognostic Assessment (GPA) Score, and Basic Score for Brain Metastases. Endpoints were time to CNS metastasis and Post CNS-relapse survival.

Results: With a median follow-up of 7.5 years, 818 (31.4%) patients had locoregional or distant recurrence. In total, 159 (6.1%) patients developed CNS metastases, of whom 92 (3.5%) as the first site of recurrence and 22 meningeal metastases were detected. In triple negative, 51 (32.1%) developed SNC metastases, 46 (28.9%) in luminal A, 37 (23.3%) in HER2 and 25 (15.7%) in luminal B patients. Median time since breast cancer diagnosis to SNC metastases was 28.1 months (mo) and the variables influencing it were Clinical T (P &lt; 0.001), clinical stage (P &lt; 0.001), histologic grade (p = 0.005), estrogen and progesterone receptors ((p = 0.026 for both) and inmmunophenotype (P &lt; 0.005). None of these factors was associated with post SNC-recurrence survival. In regard to variables measured at SNC relapse, we couldn't find differences in age groups (&lt;65 vs ≥65 years old; p = 0.662), extracraneal metastases (present, 4.2mo vs absent, 5.3mo), number of brain metastases (&lt;1 vs ≥ 2; p = 0.550) or volume of CNS metastases (&lt;3cm; 7.7mo vs ≥3cm, 5.2mo; p = 0.298). Statistical differences were observed when patients were stratified by control of primary tumor (controlled, 6.3mo; vs uncontrolled, 3.6 mo; P &lt; 0.001), leptomeningeal involvement (present, 3.0mo vs absent, 5.2mo; p = 0.25) and karnofski performance status (≥ 7, 6.5mo vs &lt;7, 3.5mo; p = 0.002). In the multivariate analysis, significant variables related with a shorter post-SNC recurrence survival were a Karnofski performance status &lt;7 (HR 1.46; p = 0.048); uncontrolled primary tumor (HR 1.92; p = 0.001); and leptomeningeal involvement (HR 1.94; p = 0.025). When survival was evaluated stratifying patients according to prognostic scores, we found differences when cases were grouped according to RPA score (Class I, 6.8 mo vs Class II, 5.5 mo vs Class III, 3.5 mo; p = 0.014). GPA scores could not identify groups with significant difference (p = 0.070). The basic score for BM detected groups with significant differences (scores 0–1, 3.7 mo vs score 2, 6.9 mo vs score 3, 6.8 mo).

Conclusions: Prognostic scores calculated from variables measured at CNS relapse are useful to identify groups with shorter post CNS recurrence survival while immunophenotype of breast cancer was unable to identify groups of different prognosis.

Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P6-07-20.

A phase II study of clofarabine and daunorubicin in patients age 60 or older with newly diagnosed adult acute myeloid leukemia

6602

Background: The outcome of acute myeloid leukemia (AML) patients (pts) ≥60 years (yrs) old remains poor, with a median overall survival of a few months. Neither the addition of other drugs during induction of the traditional 7+3 regimen nor an increase of cytarabine or daunorubicin doses has improved their overall survival. We have therefore designed a combination of clofarabine and daunorubicin. Aims: Determine the safety and efficacy of clofarabine and daunorubicin in pts ≥ 60 yrs of age. Methods: Induction-clofarabine 20 mg/m2 /d in a 1-hour IV for 5 days followed 3 hrs later by daunorubicin 50 mg/m2 IV over 5 minutes on days 1, 3, and 5. Pts who did not achieve complete remission (CR) or CR without platelet recovery (CRp) were eligible for a second round of induction as above. Consolidation - 2 cycles of clofarabine 20mg/m2 /d for 3 days and daunorubicin 50 mg/m2/day on days 1 and 3. Results: Twenty-one pts were enrolled. The median age was 69 (range 60-85) yrs. Seven pts (34%) had secondary AML and 52% had complex karyotype. Twenty-one pts completed one induction cycle, one pt received a second induction due to residual disease and 6 pts underwent consolidation therapy. The principal toxicity was grade ≥3 infections and prolonged thrombocytopenia and neutropenia, which occurred in 18% and 21% of pts, respectively. Three deaths were documented from treatment complications (sepsis and fungal infections). Eight (38.1%) pts achieved either a CR or CRp. Responses were observed in 5/11 (45.5%) pts with high risk features (complex karyotype and/or presence of FLT3). The median disease free-survival was 1.9 (range, 0.9-8) months and the median overall survival was 6.5 (range, 4.2-14.2) months. The 1-yr overall survival was 35%. Conclusions: These preliminary results suggest that clofarabine in combination with daunorubicin is safe, well-tolerated with minimal toxicity, and active for the upfront treatment of ≥60 yrs old AML pts, including those with adverse features. Future trials will explore additional modifications to this regimen in order to optimize therapy for this group of pts with high risk features and frequent treatment failure.

Abstract P4-09-18: Triple Negative Breast (TNBC) Cancer in Hispanic Population: Clinicopathological and Recurrence Features

Background: TNBC is characterized by the lack of hormonal receptors and HER2 expression and occurs most frequently in Black and Hispanic than in Caucasian population. We compare clinicopathological and outcome variables between TNBC and non-TNBC Hispanic patients (pts). Methods: We reviewed 1044 cases of operable breast cancer treated at INEN, Lima-Perú (2000 — 2002). Phenotypes were categorized by IHC into TNBC and non-TNBC. Variables in both groups were compared with the X2 test. We used Kaplan-Meier method for comparing disease free survival (DFS), overall survival (OS), post recurrence survival by phenotype (TNBC v non-TNBC) and the hazard rate (HR) for different times. Results: 212 pts (20.3%) were TNBC and 832 non-TNBC (79.7%); median for age was 48 and 49 years (y), for tumor size 3.5 and 2.9cm; TNBC was more likely to have involved nodes (55.7 v 45.6%; P=0.009), tumors &gt; 5cm (14.2 v 8.1%; P=0.011) and histological grade 3 (62.7 v 29.0%; P&lt;0.001). Recurrence rates (RR) were 25.9 v 22.5% (TNBC v non-TNBC). TNBC and non-TNBC present different RR.

Recurrence sites and phenotype

Median of follow-up was 7.63 and 7.57y; there was not significant differences for DFS between TNBC and non-TNBC (72.8 v 76.6%;P=0.057), but in OS had differences (73.2 v 79.9%; P=0.014). The HR for recurrence was calculated to 1y (7.9 v 2.7%); 2y (8.8 v 5.8%); 3y (8.0 v 6.5%); 4y (2.7 v 3.3%) and to 5y (3.5 v 3.9%) and the HR for death to 1y (4.3 v 0.8%); 2y (6.7 v 2.6%); 3y (8.4 v 3.8%); 4y (5.5 v 2.8%) and to 5y (2.6 v 4.5%) for TNBC and non-TNBC, respectively. HR for recurrence and death was higher in TNBC until the first 3y and 4y, respectively, then HR were similar for both groups. Post recurrence survival show differences between TNBC and non-TNBC (HR=1.7; 95% IC 1.2 - 2.3; P=0.002). Conclusion: In early breast cancer, TNBC is more aggressive than non-TNBC; present a higher RR in brain, distant nodes and other soft tissues; strongly time-dependent HR for recurrence and death, and a worse Post recurrence survival.