Prediction of Volasertib Sensitivity in Acute Myeloid Leukemia Relapsed or Refractory to Venetoclax Plus Azacitidine

INTRODUCTION. A randomized phase 2 trial in acute myeloid leukemia (AML) in the first line (1L) setting with the polo-like kinase 1 (PLK1) inhibitor volasertib (Vola) in combination with low-dose cytarabine (LDAC) yielded a 31.0% CR + CRi rate and a median overall survival (OS) of 8.0 months in contrast to 13.3% and 5.2 months, respectively, for LDAC without volasertib. However, a phase 3 study, also in the 1L setting, did not demonstrate a survival advantage, believed to be largely due to high rates of infections (OS of 6.5 months with LDAC + placebo and 5.6 months with LDAC + Vola). Both studies used a fixed/flat dose of Vola of 350 mg twice per 28-day cycle and enrolled patients not fit for intensive chemotherapy. Responses were seen across all prognostic groups, including those with complex karyotype and 17p/TP53 abnormalities. The difference between the phase 2 and 3 trial results was largely attributed to inconsistent antibiotic prophylaxis, fixed dosing which resulted in overdosing of smaller patients, and trial regions. Meanwhile, venetoclax-based 1L treatments have boosted response rates to 60-70% in elderly chemo-ineligible patients. Unfortunately, relapse to such treatments is the norm. Notable Labs has received US FDA clearance to test volasertib in combination with decitabine in a phase 2 clinical trial in relapsed/refractory (R/R) AML. Adjustments in the Vola dose (body surface area-based instead of fixed dose), frequency (Vola infusion once per 28-day cycle), and mandated infection prophylaxis will be implemented to address previously reported toxicities. Furthermore, at a later stage of the trial, we plan on using a companion diagnostic (CDx) to selectively enroll patients believed to have the highest chance of favorable response to volasertib. Here, we demonstrate the feasibility of using cryopreserved bone marrow (BM) or peripheral blood (PB) AML samples for response prediction to volasertib treatment in the context of the CDx.

METHODS. White blood cells (WBCs) from fresh and cryopreserved BM and PB baseline samples from patients treated with venetoclax (Ven) and 5-azacitidine (Aza) were treated ex vivo with vehicle (dimethyl sulfoxide, DMSO), Ven, Aza, and Vola for three days, followed by flow cytometric assessment for blast reduction. Normalized blast counts are the ratios of post-treatment blast counts on drug vs. DMSO, i.e., for a given sample, the number of blasts surviving ex vivo drug treatment divided by the number of viable cells treated with DMSO.

RESULTS. For cryopreserved samples, the mean ± standard deviation (SD) WBC viability prior to the ex vivo drug treatment step was 67 ± 17 % (n = 15) with a recovery of 19 ± 14 % (n = 14). Even though this is lower than the typical >90% viability we have observed with fresh samples, areas under dose-response curves were proportional between fresh and cryopreserved samples, suggesting that lower cell viabilities do not impact assay performance. Mean ± SD EC₅₀ values for cryopreserved samples were 1099 ± 641 nM (n = 6) for Aza, 128 ± 72 nM (n = 5) for Ven, and 10 ± 4 nM (n = 6) for Vola. A preliminary model for Ven-Aza treatment outcome prediction correctly classified 7 of 8 (87.5%) clinical responders and 3 of 4 (75.0%) clinical non-responders (n = 12, accuracy = 83.3%, positive predictive value = 87.5%, sensitivity = 87.5%, specificity = 75.0%). Ex vivo treatment with Vola reduced the normalized blast counts in samples both from Ven-Aza sensitive and resistant patients. Using the median normalized blast count to distinguish potential responders from potential non-responders, 4/8 (50%) Ven-Aza responders and 2/4 (50%) Ven-Aza non-responders were below this cutoff and were thus considered potential volasertib responders.

CONCLUSIONS. We have developed a method based on cryopreserved BM or PB samples to predict Vola sensitivity in R/R AML. In contrast to fresh samples, cryopreserved samples alleviate the time constraints and logistics associated with the shipment of fresh samples. Ven-refractory AML has no standard of care for treatment and portends dismal outcomes in patients. The encouraging performance of our ex vivo predictive medicine platform could pave the way for improving Vola response rates in patients in the upcoming phase 2 volasertib study.

Patil:Notable Labs: Current Employment, Other: Current holder of stock options in a publicly-traded company . Gu:Notable Labs: Current Employment, Other: Current holder of stock options in a publicly-traded company . Tognon:Notable Labs: Other: Collaborative Research funding; AstraZeneca: Other: Collaborative Research funding. Savona:AbbVie; Bristol Myers Squibb; CTI BioPharma Corp.; Geron; Karyopharm; Novartis Pharmaceuticals Corporation; Ryvu Therapeutics; and Sierra Oncology, Inc.: Consultancy; Astex Pharmaceuticals for travel grant.: Other: Financial or Material Support; ALX Oncology Inc.; Astex Pharmaceuticals; Incyte Corporation; and Takeda Pharmaceutical Company Limited.: Research Funding; Empath Biosciences; Karyopharm and Ryvu Therapeutics: Current holder of stock options in a privately-held company. Leonardi:Notable Labs: Current Employment, Other: Current holder of stock options in a publicly-traded company . Wagner:Notable Labs: Current Employment, Other: Current holder of stock options in a publicly-traded company . Michelson:Notable Labs: Consultancy. Lacher:Notable Labs: Current Employment, Other: Current holder of stock options in a publicly-traded company .