Purpose:

The Polo-like kinase 1 (PLK1) is a master regulator of mitosis and overexpressed in acute myeloid leukemia (AML). We conducted a phase Ib study of the PLK1 inhibitor, onvansertib, in combination with either low-dose cytarabine (LDAC) or decitabine in patients with relapsed or refractory (R/R) AML.

Patients and Methods:

Onvansertib was administered orally, in escalating doses, on days 1–5 in combination with either LDAC (20 mg/m2; days 1–10) or decitabine (20 mg/m2; days 1–5) in a 28-day cycle. The primary endpoint was to evaluate first-cycle dose-limiting toxicities and the MTD. Secondary and exploratory endpoints included safety, pharmacokinetics, antileukemic activity, and response biomarkers.

Results:

Forty patients were treated with onvansertib (12–90 mg/m2) in combination with LDAC (n = 17) or decitabine (n = 23). Onvansertib was well tolerated with most grades 3 and 4 adverse events related to myelosuppression. In the decitabine arm, the MTD was established at 60 mg/m2, and 5 (24%) of the 21 evaluable patients achieved complete remission with or without hematologic count recovery. Decrease in mutant circulating tumor DNA (ctDNA) during the first cycle of therapy was associated with clinical response. Engagement of the PLK1 target, TCTP, was measured in circulating blasts and was associated with greater decrease in bone marrow blasts.

Conclusions:

The onvansertib and decitabine combination was well tolerated and had antileukemic activity particularly in patients with target engagement and decreased mutant ctDNA following treatment. This combination will be further investigated in the ongoing phase II trial.

Translational Relevance

Patients with relapsed or refractory acute myeloid leukemia (R/R AML) have limited treatment options and dismal outcome. The Polo-like kinase 1 (PLK1) is a master regulator of the cell cycle, and is overexpressed in AML. Onvansertib is an oral and highly selective ATP-competitive PLK1 inhibitor with potent antitumor activity in AML preclinical models. In this phase Ib study, we examined the safety and efficacy of onvansertib in combination treatments for patients with R/R AML. Onvansertib with decitabine demonstrated safety and preliminary efficacy. We identified biomarkers correlated with treatment response. First, early decrease in mutant circulating tumor DNA was associated with clinical response. Second, engagement of the PLK1 target, TCTP, in circulating blasts was observed in a subset of patients and may be associated with greater decrease in bone marrow blasts. Results of the phase Ib study justify further investigation of the onvansertib and decitabine combination in patients with R/R AML.

Acute myeloid leukemia (AML) is characterized by the clonal expansion of myeloid blasts resulting in bone marrow failure. AML is predominantly a disease of older patients with a median age at diagnosis of 68 years (1). For patients with AML who are believed unfit for, or do not desire, intensive treatment, hypomethylating agents (HMA; e.g., azacitidine or decitabine) or low-dose cytarabine (LDAC) have historically been treatment options. However, complete responses are uncommon and often of limited duration (2). In 2018, new agents (venetoclax and glasdegib) were approved in the United States in the first-line setting in combination with HMAs or LDAC for older and unfit patients based on phase II open-label trials (3–5). Recent updates from the ongoing randomized phase III studies showed significant increase in overall survival (OS) for venetoclax in combination with azacitidine, but not LDAC (6, 7). Patients with relapsed or refractory (R/R) AML have very limited effective therapy options, particularly in the absence of targetable mutations such as FLT3 or IDH1/2, and their outcomes are dismal with median survival of less than 6 months (8, 9).

Polo-like kinases (PLK) are a family of five highly conserved serine/threonine protein kinases. PLK1 is a master regulator of mitosis and is involved in several steps of the cell cycle, including mitosis entry, centrosome maturation, bipolar spindle formation, chromosome separation, and cytokinesis (10). PLK1 has been shown to be overexpressed in solid tumors and hematologic malignancies, including AML (11, 12). PLK1 inhibition induces G2–M-phase arrest with subsequent apoptosis in cancer cells, and has emerged as a promising targeted therapy (11).

Several PLK inhibitors have been studied in clinical trials (13). In a randomized phase II study of patients with AML who were treatment naïve yet unsuitable for induction therapy, the pan-PLK inhibitor, volasertib (BI6727), administered intravenously in combination with LDAC showed a significant increase in OS when compared with LDAC alone (14). A subsequent randomized phase III study identified no benefit of the combination and described an increased risk of severe infections (15).

Onvansertib (also known as PCM-075 or NMS-1286937) is a selective ATP-competitive PLK1 inhibitor. Biochemical assays demonstrated high specificity of onvansertib for PLK1 among a panel of 296 kinases, including other PLK members (16). Onvansertib has potent in vitro and in vivo antitumor activity in models of both solid and hematologic malignancies (16). Onvansertib inhibited cell proliferation at nanomolar concentrations in AML cell lines and tumor growth in xenograft models of AML (16, 17). In addition, onvansertib significantly increased cytarabine antitumor activity in disseminated models of AML (16, 17).

A phase I, first-in-human, dose-escalation study of onvansertib in patients with advanced/metastatic solid tumors identified neutropenia and thrombocytopenia as the primary dose-limiting toxicities (DLT; ref. 18). These hematologic toxicities were anticipated on the basis of the mechanism of action of the drug and were reversible, with recovery occurring within 3 weeks. The half-life of onvansertib was established between 20 and 30 hours (18). The oral bioavailability of onvansertib plus its short half-life provide the opportunity for convenient, controlled, and flexible dosing schedules with the potential to minimize toxicities and improve the therapeutic window.

This multicenter phase Ib study (NCT03303339) was performed to assess the safety, pharmacokinetics, and preliminary clinical activity of onvansertib in combination with either LDAC or decitabine in patients with R/R AML. Pharmacodynamics and biomarker studies, including baseline genomic profiling, serial monitoring of mutant allele fractions in plasma, and the extent of PLK1 inhibition in circulating blasts, were performed to identify potential biomarkers associated with clinical response.

Patient selection

Patients 18 years or older with a confirmed diagnosis of AML and Eastern Cooperative Oncology Group (ECOG) performance status of 0–2 were eligible. Patients were allowed up to three prior treatment regimens for their AML, including HMAs. Induction therapy and hematopoietic stem cell transplant were counted as one prior line of therapy. Treatment-naïve patients who were unfit for intensive induction therapy were also eligible. Treatments for preexisting myelodysplastic syndrome (MDS) were allowed and not considered as prior treatment. Patients with treatment-related AML and acute promyelocytic leukemia were excluded, as were patients with aspartate aminotransferase and/or alanine aminotransferase ≥2.5 × upper limit of normal, total bilirubin >2 mg/dL, or serum creatinine ≥2.0 mg/dL. The protocol was approved by the institutional review board or independent ethics committees at each participating center and was in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. Written informed consent was obtained from all patients before screening.

Treatment plan and study design

Onvansertib was administered orally, in escalating doses, on days 1 through 5 in combination with either LDAC (arm A, 20 mg/m2 subcutaneously once daily on days 1 through 10) or decitabine (arm B, 20 mg/m2 intravenously over 1 hour on days 1 through 5) in a 28-day cycle. Investigators had the flexibility to shorten the cycle to 21 days if they judged that more frequent dosing could benefit the patient. The primary objectives were to evaluate DLTs and the MTD or recommended phase II dose (RP2D) of onvansertib. The first dose of onvansertib was 12 mg/m2, which was half of RP2D established in the phase I single-agent study for solid tumors (18). Each arm followed a standard 3+3 dose-escalation design, in which onvansertib dose was escalated by 50% increments. Three patients were treated, and if there were no DLTs in the first cycle, escalation to the next higher dose occurred. If a DLT was reported, an additional 3 patients were treated at that dose. If ≥2 patients experienced a DLT, the dose was considered nontolerated, and lower doses were explored in subsequent cohorts. Subjects who had not received at least 80% of the dose of study drug(s) during the first cycle or who discontinued for any reason other than DLT were replaced. The MTD was defined as the highest dose achieved at which no more than one of six subjects experienced a DLT. The RP2D was determined based on the assessment of safety, pharmacokinetics, and preliminary efficacy in subjects treated at a dose cleared for safety.

Safety

Patients were evaluable for safety if they had received at least one dose of onvansertib. Safety evaluations included physical examinations, laboratory test results, electrocardiograms, and monitoring of adverse events (AE) graded by the NCI Common Terminology Criteria for Adverse Events (NCI CTCAE version 4.03). Investigators assessed causality as either unrelated, unlikely, possibly, probably, or definitely related to study drugs. DLTs (defined in Supplementary Table S1) were evaluated during the first cycle of treatment.

Antileukemic activity

Bone marrow evaluations were performed at screening, between days 15–28 of cycles 1 and 2, and following every other subsequent cycle if considered clinically appropriate by the investigator. Response to treatment was evaluated by the investigator using the modified International Working Group criteria 2003 (19), detailed in Supplementary Table S2. Antileukemic activity was assessed in all patients evaluable for DLTs.

Pharmacokinetics

Blood samples were collected from patients for pharmacokinetic analysis on days 1 and 5 at predose; at 0.5, 1, 2, 3, 4, 8, and 24 hours after administration of onvansertib; and once on days 8, 15, and 22. Onvansertib plasma concentrations were determined by LC/MS-MS at PRA Bioanalytical Laboratory. The pharmacokinetics parameters, including maximum plasma concentration (Cmax), time to maximum plasma concentration (tmax), half-life (t1/2), and AUC, were calculated for each patient using Phoenix WinNonlin.

Correlative studies

Plasma inhibitory activity assay

Onvansertib plasma inhibitory activity was assessed by measuring changes in the phosphorylation of the PLK1 substrate, TCTP, as described previously for FLT3 inhibitors (20). HL-60 cell line used for this assay was obtained from the ATCC and cultured in Iscove's modified Dulbecco's medium supplemented with 20% FBS. The cell line was not tested for Mycoplasma. Briefly, 3 × 106 HL-60 cells (passage 2–5) were incubated with 1 mL of patient plasma at 37°C for 1 hour. Cells were washed, lysed, and changes in phosphorylated TCTP (pTCTP)/TCTP were assessed by immunoblot. Plasma samples from patients treated with onvansertib (12–60 mg/m2) in combination with LDAC or decitabine were used. For each dose level, 3–4 patients were analyzed (a total of 18 patients) on the basis of the availability of plasma samples, and each dose level included at least 1 patient of each arm (LDAC/decitabine). For each patient, pTCTP/TCTP level was quantified in cells incubated with plasma collected on day 1 at predose and postdose (tmax) and normalized to day 1 predose sample to measure the % pTCTP inhibition by onvansertib-containing plasma.

Blood collection and processing for correlative studies

Blood samples were collected from patients into CellSave Preservative Tubes (Silicon Biosystems) and EDTA tubes on days 1 (predose and 3 hours postdose), 5, 8, 15, and 22 of each cycle and processed 24 hours after collection at Cardiff Oncology's laboratories. Plasma was separated from whole blood collected in CellSave tube by centrifugation. Peripheral blood mononuclear cells (PBMC) and bone marrow mononuclear cells (BMMC) were isolated by density gradient centrifugation over Histopaque-1077 (Sigma Aldrich), following the manufacturer's instructions. To quantify AML blasts, cells were labeled with a cocktail consisting of fluorochrome-labeled antibodies to CD45, CD34, CD117, CD13, and HLA-DR (BioLegend). Samples were analyzed with a FACSCanto II (BD Biosciences). AML blasts were identified on the basis of a low side scatter/CD45dim profile and expression of blast markers. Genomic DNA (gDNA) was extracted from PBMCs and BMMCs using the QIAamp DNA Blood Kit (Qiagen) and circulating tumor DNA (ctDNA) from plasma using QIAamp Circulating Nucleic Acid Kit (Qiagen) according to the manufacturer's recommendations. DNA was quantified with a Qubit 3.0 Fluorometer (Thermo Fisher Scientific). Targeted next-generation sequencing (NGS) was performed with the VariantPlex Myeloid Panel (ArcherDx), which includes 75 genes known to be associated with AML (Supplementary Table S3). NGS libraries were sequenced on the MiSeq Platform (Illumina) with MiSeq reagent kit v3. Data analysis was performed using Archer Analysis version 6.0.3.2 (ArcherDX) with custom targets, VariantPlex_Myeloid_GSP5031-v1.0, and targeted mutations, Archer Comprehensive Targets v1.1. For 20 patients, a driver mutation was identified by targeted NGS, and probe-based assays enabling detection of variant and wild-type alleles were commercially obtained (PrimePCR, Bio-Rad Laboratories, or TaqMan SNP Genotyping Assays, Thermo Fisher Scientific). Digital droplet PCR (ddPCR) was performed using the Bio-Rad QX200 Droplet Digital PCR System following the manufacturer's protocols. Each sample was analyzed using at least two technical replicates. A Poisson correction was applied for independent enumeration of mutant alleles and wild-type alleles to calculate mutant allele frequency (MAF). Data analysis was carried out using the QuantaSoft Software, version 1.7 (Bio-Rad).

Protein extract preparation and immunoblot

Protein extracts were prepared from PBMCs isolated from CellSave tube and from HL-60 cells using M-PER buffer with Protease and Phosphatase Inhibitor Cocktails (Thermo Fisher Scientific). Protein concentration was measured with the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). Western blots were performed as Simple Western Assays using the Wes System (ProteinSimple), a combination of capillary electrophoresis and immunodetection techniques, following the manufacturer's protocols. Primary antibodies were purchased from Cell Signaling Technology: phospho-TCTP-Ser46 (#5251) and TCTP (#5128). Quantitative analysis was performed using Compass Software (ProteinSimple). Signal intensity (area) of pTCTP was normalized to the peak area of TCTP and reported as pTCTP/TCTP or %pTCTP.

Statistical analysis

Pearson correlation was used to analyze the relationship between plasma ctDNA MAF and bone marrow MAF. A Fisher exact test was performed to test the association between target engagement and decrease in bone marrow blasts. The AUC of the ROC curve was used to measure the performance of the log2 (C1/C0) to predict clinical response. The ROC curve was used to select the optimal threshold for prediction based on the Youden method (21) and the sensitivity and specificity at this threshold were estimated. Ninety-five percent confidence intervals (CI) were calculated for the AUC, and for all parameters at the threshold, using 2,000 stratified bootstrap replicates (22). All statistical tests were performed using R version 4.0.2 and GraphPad Prism version 8.4.3.

Patient characteristics

A total of 40 patients were enrolled in this phase Ib study in nine centers in the United States between January 2018 and August 2019. They were treated in two noncomparative study arms: arm A (onvansertib + LDAC, n = 17) and arm B (onvansertib + decitabine, n = 23). Patients were assigned to either arm on the basis of investigator discretion and slot availability. At data cutoff, 31 October 2019, 5 patients continued to receive onvansertib on-study: 4 patients in the decitabine arm and 1 patient in the LDAC arm. Baseline characteristics are summarized in Table 1 and detailed in Supplementary Table S4. The median age was 68 years (range, 33–88), and 70% of patients were male. Twenty-seven (68%) patients had adverse risk cytogenetics based on the 2017 ELN recommendations (23). At study entry, 20 (50%) patients had received one prior regimen for AML and 15 (38%) patients had received two or three prior treatments. There were 5 (12.5%) patients with untreated AML, but four of them had received HMAs for a previous diagnosis of MDS. The median proportion of bone marrow blasts was 26% (range, 5–95) with 55% of patients having >20% bone marrow blasts at baseline. Two (9%) patients in the decitabine arm had prior decitabine treatment and 8 (47%) patients in the LDAC arm had prior cytarabine treatment.

Table 1.

Baseline characteristics.

Patients (N = 40)
Age, median (range) 68 (33–88) 
Male/female 28/12 (70%/30%) 
ECOG performance status  
 0 4 (10%) 
 1 33 (82.5%) 
 2 3 (7.5%) 
Prior AML therapies  
 0 5 (12.5%) 
 1 20 (50%) 
 2+ 15 (37.5%) 
Prior cytarabine treatment (all patients) 24 (60%) 
 LDAC arm (n = 17) 8 (47%) 
 Decitabine arm (n = 23) 16 (70%) 
Prior decitabine treatment (all patients) 10 (25%) 
 LDAC arm (n = 17) 8 (47%) 
 Decitabine arm (n = 23) 2 (9%) 
Cytogenetic risk  
 Favorable 2 (5%) 
 Intermediate 11 (27.5%) 
 Adverse 27 (67.5%) 
Percent bone marrow blasts, median (range) 26 (5–95) 
Patients (N = 40)
Age, median (range) 68 (33–88) 
Male/female 28/12 (70%/30%) 
ECOG performance status  
 0 4 (10%) 
 1 33 (82.5%) 
 2 3 (7.5%) 
Prior AML therapies  
 0 5 (12.5%) 
 1 20 (50%) 
 2+ 15 (37.5%) 
Prior cytarabine treatment (all patients) 24 (60%) 
 LDAC arm (n = 17) 8 (47%) 
 Decitabine arm (n = 23) 16 (70%) 
Prior decitabine treatment (all patients) 10 (25%) 
 LDAC arm (n = 17) 8 (47%) 
 Decitabine arm (n = 23) 2 (9%) 
Cytogenetic risk  
 Favorable 2 (5%) 
 Intermediate 11 (27.5%) 
 Adverse 27 (67.5%) 
Percent bone marrow blasts, median (range) 26 (5–95) 

Dose escalation and toxicity

Onvansertib was investigated at five dose levels (12–60 mg/m2) in arm A (LDAC) and at six dose levels (12–90 mg/m2) in arm B (decitabine). At data cutoff, the median number of completed onvansertib cycles for both arms was 2 (range, 0–18), and the median time on treatment was 53 days (range, 11–574). Five patients (1 in arm A and 4 in arm B) had at least one cycle of less than 28 days (24 or 25 days). Treatment with onvansertib was well tolerated in both arms through the first five dose levels (12–60 mg/m2), with no DLTs reported. In arm A, no higher dose levels were explored. In arm B, 2 of 6 patients treated at dose level 6 (90 mg/m2) experienced a DLT during the first cycle of treatment, consisting of grade 3 (G3) mucositis and grade 4 (G4) rash, respectively. Subsequently, onvansertib MTD in combination with decitabine was established at 60 mg/m2.

Treatment-emergent AEs reported in at least 10% of patients are listed in Table 2. The most common G3 and G4 AEs were anemia (35%), febrile neutropenia (30%), neutropenia (25%), thrombocytopenia (25%), leukopenia (12.5%), and stomatitis (12.5%). G3 and G4 AEs considered possibly related to onvansertib were mostly hematologic: neutropenia (15%), anemia (12.5%), thrombocytopenia (12.5%), and leukopenia (5%). G3 stomatitis, reported in 3 patients (8%) at the highest dose levels (one at 60 mg/m2 and two at 90 mg/m2), was the only nonhematologic G3/G4 AE possibly related to onvansertib reported in ≥5% patients. Two patients (one from 18 mg/m2 and one from 27 mg/m2 dose level) had a dose reduction of onvansertib (dose level −1) due to hematologic toxicities. Three patients were discontinued from therapy because of treatment-related AEs, including grade 2 (G2) fatigue, G3 mucositis (DLT), and G4 rash (DLT).

Table 2.

Treatment-emergent AEs reported in ≥10% of all patients (N = 40).

AEGrade 1–2 (%)Grade 3–4 (%)All grades (%)
Anemia 1 (2.5) 14 (35) 15 (37.5) 
Fatigue 14 (35)  14 (35) 
Febrile neutropenia  12 (30) 12 (30) 
Nausea/vomiting 12 (30)  12 (30) 
Thrombocytopenia 2 (5) 10 (25) 12 (30) 
Rash/pruritus 10 (25) 1 (2.5) 11 (27.5) 
Dyspnea 8 (20) 2 (5) 10 (25) 
Neutropenia  10 (25) 10 (25) 
Stomatitis 4 (10) 5 (12.5) 9 (22.5) 
Diarrhea 8 (20)  8 (20) 
Edema 6 (15) 2 (5) 8 (20) 
Constipation 6 (15)  6 (15) 
Cough 6 (15)  6 (15) 
Decreased appetite 6 (15)  6 (15) 
Epistaxis 6 (15)  6 (15) 
Dizziness 5 (12.5)  5 (12.5) 
Pyrexia 5 (12.5)  5 (12.5) 
Leukopenia  5 (12.5) 5 (12.5) 
Blood bilirubin increased 3 (7.5) 1 (2.5) 4 (10) 
Headache 4 (10)  4 (10) 
Lung infection  4 (10) 4 (10) 
Oropharyngeal pain 4 (10)  4 (10) 
AEGrade 1–2 (%)Grade 3–4 (%)All grades (%)
Anemia 1 (2.5) 14 (35) 15 (37.5) 
Fatigue 14 (35)  14 (35) 
Febrile neutropenia  12 (30) 12 (30) 
Nausea/vomiting 12 (30)  12 (30) 
Thrombocytopenia 2 (5) 10 (25) 12 (30) 
Rash/pruritus 10 (25) 1 (2.5) 11 (27.5) 
Dyspnea 8 (20) 2 (5) 10 (25) 
Neutropenia  10 (25) 10 (25) 
Stomatitis 4 (10) 5 (12.5) 9 (22.5) 
Diarrhea 8 (20)  8 (20) 
Edema 6 (15) 2 (5) 8 (20) 
Constipation 6 (15)  6 (15) 
Cough 6 (15)  6 (15) 
Decreased appetite 6 (15)  6 (15) 
Epistaxis 6 (15)  6 (15) 
Dizziness 5 (12.5)  5 (12.5) 
Pyrexia 5 (12.5)  5 (12.5) 
Leukopenia  5 (12.5) 5 (12.5) 
Blood bilirubin increased 3 (7.5) 1 (2.5) 4 (10) 
Headache 4 (10)  4 (10) 
Lung infection  4 (10) 4 (10) 
Oropharyngeal pain 4 (10)  4 (10) 

There were a total of 71 serious AEs (SAE) reported in 33 patients (Supplementary Table S5). The most common SAEs were: febrile neutropenia (25 events in 17 patients), lung infection (5), and pneumonia (4). Nine (13%) SAEs considered as possibly related to onvansertib consisted of G3 febrile neutropenia (n = 1, resolving within 1 week), G4 neutropenia (n = 1), G3 stomatitis (n = 1), G2 rash maculopapular (n = 1), G3 palmar-plantar erythrodysesthesia syndrome (n = 1), and G4 rash (n = 1). Three deaths occurred while on treatment, all of which were related to AML or its complications: 2 patients died because of progressive disease and 1 from intracranial hemorrhage after a fall.

Pharmacokinetics

Table 3 shows mean pharmacokinetics parameters on day 5 of cycle 1 for all patients enrolled. The Cmax and AUC from 0 to 24 hours (AUC0–24) increased proportionally with the administrated dose in both arms. The tmax was achieved between 1.7 and 3.3 hours and the median half-life of onvansertib (t1/2) was 26.4 hours (range, 16–46.5), and these parameters were independent of the administrated dose or combination treatment. Overall, pharmacokinetics parameters were similar with either LDAC or decitabine as those reported for onvansertib single agent in the phase I solid tumor study (18).

Table 3.

Cycle 1 day 5 mean ± SD pharmacokinetic parameters of onvansertib.

Dose (mg/m2)Combination treatmentNumber of patientstmax (hour)Cmax (nmol/L)AUC(0–24) (nmol/L.hour)t1/2 (hour)
12 Decitabine 2.5 ± 0.6 163 ± 90 2,270 ± 1,440 26 ± 16 
12 LDAC 2.0 ± 1.0 153 ± 84 2,150 ± 833 32 ± 21 
18 Decitabine 2.7 ± 0.6 230 ± 129 3,380 ± 1,740 21 ± 7 
18 LDAC 3.0 ± 1.0 109 ± 39 1,730 ± 847 33 ± 18 
27 Decitabine 2.0 ± 1.7 411 ± 118 5,420 ± 402 16 ± 6 
27 LDAC 2.7 ± 1.5 340 ± 219 4,470 ± 1,450 18 ± 5 
40 Decitabine 3.3 ± 1.0 539 ± 228 8,050 ± 3,380 25 ± 15 
40 LDAC 1.7 ± 1.2 350 ± 104 4,270 ± 1,950 16 ± 7 
60 Decitabine 3.3 ± 1.2 1,040 ± 534 16,300 ± 6,690 47 ± 8 
60 LDAC 2.4 ± 1.1 905 ± 330 12,700 ± 4,850 37 ± 13 
90 Decitabine 3.3 ± 0.8 1,310 ± 806 21,800 ± 18,400 30 ± 11 
Dose (mg/m2)Combination treatmentNumber of patientstmax (hour)Cmax (nmol/L)AUC(0–24) (nmol/L.hour)t1/2 (hour)
12 Decitabine 2.5 ± 0.6 163 ± 90 2,270 ± 1,440 26 ± 16 
12 LDAC 2.0 ± 1.0 153 ± 84 2,150 ± 833 32 ± 21 
18 Decitabine 2.7 ± 0.6 230 ± 129 3,380 ± 1,740 21 ± 7 
18 LDAC 3.0 ± 1.0 109 ± 39 1,730 ± 847 33 ± 18 
27 Decitabine 2.0 ± 1.7 411 ± 118 5,420 ± 402 16 ± 6 
27 LDAC 2.7 ± 1.5 340 ± 219 4,470 ± 1,450 18 ± 5 
40 Decitabine 3.3 ± 1.0 539 ± 228 8,050 ± 3,380 25 ± 15 
40 LDAC 1.7 ± 1.2 350 ± 104 4,270 ± 1,950 16 ± 7 
60 Decitabine 3.3 ± 1.2 1,040 ± 534 16,300 ± 6,690 47 ± 8 
60 LDAC 2.4 ± 1.1 905 ± 330 12,700 ± 4,850 37 ± 13 
90 Decitabine 3.3 ± 0.8 1,310 ± 806 21,800 ± 18,400 30 ± 11 

Clinical response and duration

Swimmer plots are shown for patients evaluable for efficacy (Fig. 1A and B). In the decitabine arm, 5 (24%) patients achieved complete remission with or without incomplete hematopoietic recovery (CR/CRi), 4 patients had a CR and 1 a CRi. In the LDAC arm, 1 (7%) of the 15 patients achieved CRi. Overall response rate (ORR), including CR, CRi, morphologic leukemia free–state (MLFS), and partial response (PR), was 33% (7/21 patients) in the decitabine arm and 13% (2/15 patients) in the LDAC arm. Eight (44%) of the 18 patients with evaluable bone marrow biopsy showed a ≥50% reduction in blasts in the decitabine arm and 3 (25%) of 12 patients in the LDAC arm (Fig. 1C and D).

Figure 1.

Clinical course and bone marrow response of patients evaluable for efficacy (N = 36). A and B, Response to treatment, duration of treatment, and dose adjustments are shown on swimmer plots at the different dose levels (onvansertib 12–90 mg/m2) for both arms. C and D, Waterfall plots of percent change in bone marrow blast from baseline (best response) in patients from both arms. Patient identifiers and onvansertib doses (12–90 mg/m2) are indicated.

Figure 1.

Clinical course and bone marrow response of patients evaluable for efficacy (N = 36). A and B, Response to treatment, duration of treatment, and dose adjustments are shown on swimmer plots at the different dose levels (onvansertib 12–90 mg/m2) for both arms. C and D, Waterfall plots of percent change in bone marrow blast from baseline (best response) in patients from both arms. Patient identifiers and onvansertib doses (12–90 mg/m2) are indicated.

Close modal

In the decitabine arm, CR/CRi was documented with onvansertib at doses of 27 (n = 2), 40 (n = 1), and 90 mg/m2 (n = 2). The median number of cycles to achieve CR/CRi was 4 (range, 1–7) and 1 patient proceeded to transplant immediately after achieving CR. The median duration of CR/CRi was 5.5 months (range, 1.5–11.5). Three of the 5 responders remained on treatment at the data cutoff, with a time since CR/CRi of 1.5, 8, and 11.5 months, respectively. Two responders discontinued treatment: 1 patient proceeded to transplant and 1 patient progressed 2.5 months after remission. Responders had a median age of 75 years (range, 51–76). Two patients were treatment naïve for their AML, but had received azacitidine for MDS, and 3 patients had either relapsed (n = 2) from or were refractory (n = 1) to prior induction chemotherapy.

Correlative studies

Plasma-derived ctDNA as an early marker of treatment response

Mutational profiling was performed at baseline for all patients (n = 40) using DNA from PBMCs or BMMCs. The most frequently mutated genes were TP53 (28%), ASXL1 (20%), SRSF2 (18%), NRAS (18%), DNMT3A (18%), TET2 (18%), FLT3 (16%), and SF3B1 (10%; Supplementary Tables S6 and S7). Patients achieving CR/CRi (n = 6) had mutations in splicing factors (n = 4), signaling effectors (n = 3), and DNA methylation–related genes (n = 2; Supplementary Tables S7 and S8).

Recent studies have shown that plasma-derived ctDNA reflects the fractional abundance of somatic mutations detected in bone marrow in AML and MDS (24, 25). For 20 patients enrolled in this study, a driver mutation was identified by targeted NGS using gDNA from bone marrow or blood, and the MAF of the selected variant was measured in ctDNA from plasma by ddPCR (Supplementary Fig. S1A). In agreement with prior studies (24, 25), we observed a strong correlation (r2 = 0.91; P < 0.0001) between the MAF measured by ddPCR in bone marrow and ctDNA across 49 matched samples (Supplementary Fig. S1B), supporting the potential of ctDNA as a noninvasive tool to monitor treatment response.

We investigated the utility of measuring early changes in ctDNA MAF to predict clinical response (CR/CRi). For this purpose, MAF was assessed in ctDNA before beginning treatment (C0) and after 1 cycle of treatment (C1), and change in MAF calculated as log2 (C1/C0; Fig. 2A). Using an ROC analysis, we showed that changes in ctDNA MAF at cycle 1 were predictive of clinical responses (AUC = 0.94; P = 0.0013; Fig. 2B). The optimal log2 (C1/C0) threshold was calculated to be −0.06, at which the change in MAF predicted patient clinical response with 86% specificity (71%–100% CI) and 100% sensitivity (83%–100% CI; Fig. 2C). Positive and negative predictive values were 75% (60%–100% CI) and 100% (93%–100% CI), respectively. All patients with clinical response (n = 6) showed a decrease in ctDNA MAF at cycle 1 [log2 (C1/C0) < −0.06], while only two of the 14 nonresponders showed a similar decrease (Fig. 2A and C). The median number of cycles to achieve CR/CRi was 3 (range, 1–7) with only 1 patient achieving CR after 1 cycle. Altogether, these data suggest that changes in ctDNA MAF after the first cycle are associated with later clinical response and have the potential to help guide patient treatment.

Figure 2.

ctDNA MAF monitoring as a predictive biomarker of clinical response. MAF was measured by ddPCR in plasma ctDNA of 20 patients. A, Changes in MAF after the first cycle were calculated as log2 (C1/C0). Target used for each patient is indicated. B, ROC curve for clinical response prediction, showing the sensitivity and specificity of plasma ctDNA decrease after one cycle of treatment to predict clinical response (CR/CRi). C, Patients with decrease in ctDNA MAF at first cycle, using the optimal threshold of the ROC curve [log2 (C1/C0) = −0.06], were considered responders on the basis of ctDNA plasma test. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the ctDNA plasma test to predict clinical response are reported.

Figure 2.

ctDNA MAF monitoring as a predictive biomarker of clinical response. MAF was measured by ddPCR in plasma ctDNA of 20 patients. A, Changes in MAF after the first cycle were calculated as log2 (C1/C0). Target used for each patient is indicated. B, ROC curve for clinical response prediction, showing the sensitivity and specificity of plasma ctDNA decrease after one cycle of treatment to predict clinical response (CR/CRi). C, Patients with decrease in ctDNA MAF at first cycle, using the optimal threshold of the ROC curve [log2 (C1/C0) = −0.06], were considered responders on the basis of ctDNA plasma test. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the ctDNA plasma test to predict clinical response are reported.

Close modal

Target engagement is associated with greater decrease in bone marrow blasts

The plasma inhibitory activity assay was used to determine whether a sufficient amount of free drug was present in the plasma of patients to inhibit phosphorylation of the PLK1 substrate, TCTP. In preclinical models, onvansertib inhibits pTCTP in vitro and in vivo (16, 26). We confirmed that pTCTP was reduced by on-treatment plasma (days 1 and 5), but not by plasma with undetectable onvansertib levels (day 22; Fig. 3A). In addition, pTCTP inhibition was observed using plasma from all dose levels (12–60 mg/m2) with greater pTCTP inhibition at higher doses, correlating with inhibition of PLK1 activity (Fig. 3B).

Figure 3.

Onvansertib inhibitory activity in plasma and circulating blasts. TCTP and pTCTP protein levels were analyzed by Western blot analysis and pTCTP level was normalized to total TCTP level. A and B, Plasma inhibitory activity assay was performed by incubating HL-60 cells with patient plasma samples. A, Immunoblot and its quantification demonstrate changes in pTCTP level with onvansertib-containing plasma samples. B, Graph represents the % pTCTP inhibition at the different onvansertib dose levels (n = 3 or 4 patients per dose level). C, Immunoblot of pTCTP and TCTP using PBMCs isolated from blood samples collected on day 1 at predose (0 h) and 3 hours postdose (3 h) from target engagement patients (≥50% pTCTP inhibition, top) and nontarget engagement patients (bottom). Onvansertib dose level (12–40 mg/m2) and patient identifiers (in green for LDAC arm and purple for decitabine arm) are indicated. D, Waterfall plot of percent change in bone marrow blast from baseline (best response) in target engagement and nontarget engagement patients. Best clinical response (CR, CRi, and MLFS) and percent peripheral blasts (PB) at baseline are indicated. D, day; X h, number of hours posttreatment.

Figure 3.

Onvansertib inhibitory activity in plasma and circulating blasts. TCTP and pTCTP protein levels were analyzed by Western blot analysis and pTCTP level was normalized to total TCTP level. A and B, Plasma inhibitory activity assay was performed by incubating HL-60 cells with patient plasma samples. A, Immunoblot and its quantification demonstrate changes in pTCTP level with onvansertib-containing plasma samples. B, Graph represents the % pTCTP inhibition at the different onvansertib dose levels (n = 3 or 4 patients per dose level). C, Immunoblot of pTCTP and TCTP using PBMCs isolated from blood samples collected on day 1 at predose (0 h) and 3 hours postdose (3 h) from target engagement patients (≥50% pTCTP inhibition, top) and nontarget engagement patients (bottom). Onvansertib dose level (12–40 mg/m2) and patient identifiers (in green for LDAC arm and purple for decitabine arm) are indicated. D, Waterfall plot of percent change in bone marrow blast from baseline (best response) in target engagement and nontarget engagement patients. Best clinical response (CR, CRi, and MLFS) and percent peripheral blasts (PB) at baseline are indicated. D, day; X h, number of hours posttreatment.

Close modal

Target engagement in blasts was determined by measuring relative pTCTP/TCTP protein levels in PBMCs isolated from patient blood prior to and 3 hours after the first dose of onvansertib (Fig. 3C). Target engagement was defined as a decrease of ≥50% in pTCTP/TCTP at 3 hours postdose versus predose in patients with at least 10% of circulating blasts, and was observed in 8 (33%) of the 24 evaluable patients. Interestingly, target engagement was not dependent on onvansertib dose, pharmacokinetics, or combination treatment (Fig. 3C). However, it seemed to be associated with higher decrease in bone marrow blasts: 4 (67%) of the 6 target engagement patients had a ≥20% decrease in blasts versus 2 (18%) of the 11 nontarget engagement patients (Fig. 3D). Although the difference between the two groups was not significant (Fisher exact test, P = 0.11), these data suggest a potential link between target engagement and antileukemic activity that should be further explored.

Effective treatment options for patients with R/R AML are limited, with generally low CR and limited response durations leading to dismal survival (8). PLK1 has been identified as a promising target for AML in preclinical studies, however, nonspecific PLK1 inhibitors have shown high toxicity profiles and failed in clinic (27). The failure of the pan-PLK inhibitor, volasertib, in a phase III trial may have stemmed from its long half-life (∼5 days) and a consequent inability to adequately control the extended myelosuppression, from the absence of response biomarkers to select patients, and from the lack of specificity for PLK1 resulting in inhibition of PLK2 and PLK3, both mediators of DNA damage or oxidative stress that have tumor suppressor functions (28). Onvansertib with its high specificity for PLK1, its bioavailability, and short half-life, has the potential to mitigate toxicities observed with previous nonspecific PLK1 inhibitors. In this phase Ib clinical trial, patients with R/R AML, or de novo AML considered ineligible for intensive chemotherapy, were treated with onvansertib in combination with either LDAC or decitabine. The objectives were to evaluate the safety, pharmacokinetics, and preliminary antileukemic activity of these combinations. In addition, novel biomarkers reflecting onvansertib activity were explored that could prove useful in predicting clinical response to therapy in future studies.

Patients treated in this trial were characterized predominantly by poor-risk features, including advanced age (median age of 68 years), adverse risk karyotype (68%), and high-risk mutations (28% TP53, 20% ASXL1, 16% FLT3, and 8% RUNX1).

Onvansertib was well tolerated at the first five dose levels (12–60 mg/m2) in combination with both LDAC and decitabine. The most commonly reported G3/G4 AEs were hematologic (neutropenia, thrombocytopenia, and anemia), which are all expected in patients with AML and consistent with the known myelosuppression of decitabine, LDAC, and single-agent onvansertib in patients with solid tumors (18). Onvansertib-related gastrointestinal disorders (nausea, vomiting, and diarrhea) were generally mild to moderate and manageable, with no patient discontinuing therapy due to these events. Although no skin toxicities were seen in the single-agent phase I study, where the maximum onvansertib dose administered was 48 mg/m2 (18); in this study, skin toxicities were observed at the last two dose levels of onvansertib (60 and 90 mg/m2), with a higher incidence and higher grade at 90 mg/m2, suggesting that this should be considered a side-effect of high onvansertib levels.

The onvansertib and decitabine combination showed preliminary antileukemic activity with ORR of 33% across all dose levels (four CR, one CRi, one MLFS, and one PR), while the onvansertib and LDAC combination had less clinical activity (ORR 11%; one CRi and one PR). Historically, the CR/CRi rate of patients with R/R AML to HMAs is 16%, with no significant difference between azacitidine and decitabine (29). In the decitabine arm of our study, the CR/CRi rate of 24% observed across all doses was encouraging, yet further exploration of the combination is warranted to determine its benefit over HMAs. On the basis of these preliminary efficacy data and the fact that standard of care is currently more aligned with the use of decitabine in the United States, the decitabine combination was selected to be further explored in the phase II and the LDAC arm was discontinued.

Molecular disease monitoring using ctDNA was shown to reflect tumor burden during therapy in MDS and AML (24, 25). Similarly, we observed a high correlation between MAF in the bone marrow and ctDNA at matched timepoints. In addition, we found that a decrease in MAF of ctDNA after the first cycle of treatment was associated with clinical response, regardless of the time of response. We recognize that this study has several limitations, including the small sample size to perform the ROC analysis and the lack of an independent cohort to validate the threshold obtained with the ROC analysis. The utility of serial ctDNA measurements as a predictor of clinical response should be further explored as it may help guide treatment for patients with AML.

A decrease in pTCTP was shown to parallel PLK1 inhibition by onvansertib in in vitro and in vivo preclinical models (16, 26). Onvansertib plasma inhibitory activity was correlated with plasma drug levels, albeit not with clinical response. Conversely, target engagement in circulating blasts was observed in a subset of patients and was not associated with onvansertib dose, but with a greater decrease in bone marrow blasts. This observation may be due to the fact that PLK1 activity in leukemic blasts varies between patients resulting in a range of sensitivity and response to PLK1 inhibitors. Further development of the target engagement assay and its evaluation in a larger patient population are needed to determine its true utility as response biomarker.

Recently, venetoclax in combination with HMAs or LDAC has been introduced for older patients ineligible for intensive chemotherapy. However, the median OS of patients after failure of venetoclax in combination with HMAs is low (9), and novel therapies are urgently needed for this population. Onvansertib has shown efficacy in venetoclax-resistant preclinical models, supporting its potential for venetoclax R/R AML patients. A phase II study of 32 patients treated with onvansertib (60 mg/m2) and decitabine is currently ongoing to further assess the safety profile and antileukemic activity of this combination in patients with R/R AML, including patients who failed prior venetoclax therapy. The identification of predictive biomarkers would be key to enrich for patients who are likely to benefit from the onvansertib and decitabine combination. The ongoing phase II correlative studies include further evaluation of the association between target engagement, changes in ctDNA and clinical response, and identification of additional potential biomarkers through genomic and transcriptional analyses.

A.M. Zeidan reports other from Celgene/BMS (consultancy/honoraria/research funding), AbbVie (consultancy/honoraria/research funding), Pfizer (consultancy/honoraria/research funding), Boehringer Ingelheim (consultancy/honoraria/research funding), Trovagene (consultancy/honoraria/research funding), Incyte (consultancy/honoraria/research funding), Takeda (consultancy/honoraria/research funding), Novartis (consultancy/honoraria/research funding), Astex (research funding), Medimmune/AstraZeneca (research funding), Aprea (research funding), ADC Therapeutics (research funding), Otsuka (consultancy/honoraria), Jazz (consultancy/honoraria), Agios (consultancy/honoraria), Acceleron (consultancy/honoraria), Astellas (consultancy/honoraria), Daiichi Sankyo (consultancy/honoraria), Cardinal Health (consultancy/honoraria), Taiho (consultancy/honoraria), Seattle Genetics (consultancy/honoraria), BeyondSpring (consultancy/honoraria), Ionis (consultancy/honoraria), and Epizyme (consultancy/honoraria) outside the submitted work. M. Ridinger reports personal fees from Cardiff Oncology (employment) during the conduct of the study and outside the submitted work, as well as has a patent for PCT/US19/48044 pending. T.L. Lin reports other from Trovagene (support to institution for conduct of clinical trial) during the conduct of the study and other from AbbVie, Astellas, Aptevo, Bio-Path Holdings, Celgene, Celyad, Genetech-Roche, Gilead, Incyte, Jazz, Mateon, Leukeumia & Lymphoma Society, Ono Pharmaceuticals, Pfizer, Prescient Therapeutics, Seattle Genetics, and Tolero (support to institution for conduct of clinical trial) outside the submitted work. P.S. Becker reports grants from Cardiff Oncology (to the University of Washington, site principal investigator) during the conduct of the study P.S. Becker reports grants from AbbVie, Amgen, BMS, GlycoMimetics, Invivoscribe, JW Pharmaceutical, Novartis, Pfizer, and Secure Bio (to the University of Washington for studies, site principal investigator or principal investigator), and personal fees from Accordant Health Services (as a rare disease consultant on the medical advisory board) outside the submitted work, as well as has a patent for high-throughput screening of cancer stem cells pending to Becker, Lee, Blau, Oehler, Martins, Monnat, and University of Washington (submitted, broadly relevant). G.J. Schiller reports grants from Trovagene (research support) during the conduct of the study. P.A. Patel reports other from Agios (honoraria for advisory board), Forty Seven, Inc (travel stipend), and Rafael (travel stipend) outside the submitted work. A.I. Spira reports grants and nonfinancial support from Cardiff Oncology during the conduct of the study. E. Samuëlsz reports personal fees from Cardiff Oncology during the conduct of the study and outside the submitted work. S.L. Silberman reports consultancy with Cardiff Oncology, Inc. (formerly Trovagene, Inc.). M. Erlander reports personal fees from Cardiff Oncology (employment) during the conduct of the study and outside the submitted work, as well as has a patent for PCT/US19/48044 pending. E.S. Wang reports personal fees from Pfizer, Kite Pharmaceuticals, Arog/Dava Oncology, Astellas, MacroGenics, PTC Therapeutics, Stemline, Genentech, AbbVie, Takeda/Millenium, Jazz, BMS/Celgene, Daiichi Sankyo, Celyad, and Agios outside the submitted work. No potential conflicts of interest were disclosed by the other author.

A.M. Zeidan: Conceptualization, resources, formal analysis, supervision, validation, investigation, visualization, writing-original draft. M. Ridinger: Conceptualization, resources, formal analysis, supervision, validation, investigation, visualization, methodology, writing-original draft. T.L. Lin: Conceptualization, resources, supervision, investigation, writing-review and editing. P.S. Becker: Conceptualization, resources, supervision, investigation, writing-review and editing. G.J. Schiller: Conceptualization, resources, supervision, investigation, writing-review and editing. P.A. Patel: Conceptualization, resources, supervision, investigation, writing-review and editing. A.I. Spira: Conceptualization, resources, supervision, investigation, writing-review and editing. M.L. Tsai: Conceptualization, resources, supervision, investigation, writing-review and editing. E. Samuëlsz: Resources, investigation, visualization, methodology, writing-review and editing. S.L. Silberman: Conceptualization, investigation, writing-original draft. M. Erlander: Conceptualization, resources, supervision, funding acquisition, writing-review and editing. E.S. Wang: Conceptualization, resources, supervision, investigation, writing-review and editing.

The authors thank the patients who participated in this trial and their families, as well as the medical team, nursing staff, and the study coordinators at each of the sites. This study was sponsored by Cardiff Oncology which provided onvansertib and worked with investigators to design the study, develop the protocol, monitor the study, and analyze the data. Cardiff Oncology provided logistical support to collect patient samples for the correlative studies. Peter Croucher (Cardiff Oncology) provided his support for the statistical analyses.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Shallis
RM
,
Wang
R
,
Davidoff
A
,
Ma
X
,
Zeidan
AM
. 
Epidemiology of acute myeloid leukemia: recent progress and enduring challenges
.
Blood Rev
2019
;
36
:
70
87
.
2.
Wang
ES
. 
Treating acute myeloid leukemia in older adults
.
Hematology Am Soc Hematol Educ Program
2014
;
2014
:
14
20
.
3.
DiNardo
CD
,
Pratz
K
,
Pullarkat
V
,
Jonas
BA
,
Arellano
M
,
Becker
PS
, et al
Venetoclax combined with decitabine or azacitidine in treatment-naive, elderly patients with acute myeloid leukemia
.
Blood
2019
;
133
:
7
17
.
4.
Wei
AH
,
Strickland
SA
,
Hou
J-Z
,
Fiedler
W
,
Lin
TL
,
Walter
RB
, et al
Venetoclax combined with low-dose cytarabine for previously untreated patients with acute myeloid leukemia: results from a phase Ib/II study
.
J Clin Oncol
2019
;
37
:
1277
84
.
5.
Norsworthy
KJ
,
By
K
,
Subramaniam
S
,
Zhuang
L
,
Del Valle
PL
,
Przepiorka
D
, et al
FDA approval summary: glasdegib for newly diagnosed acute myeloid leukemia
.
Clin Cancer Res
2019
;
25
:
6021
5
.
6.
DiNardo
CD
,
Jonas
BA
,
Pullarkat
V
,
Thirman
M
,
Garcia
J
,
Wei
AH
, et al
Azacitidine and venetoclax in previously untreated acute myeloid leukemia
.
N Engl J Med
2020
;
383
:
617
29
.
7.
Wei
AH
,
Montesinos
P
,
Ivanov
V
,
DiNardo
CD
,
Novak
J
,
Laribi
K
. 
Venetoclax plus LDAC for newly diagnosed AML ineligible for intensive chemotherapy: a phase 3 randomized placebo-controlled trial
.
Blood
2020
;
135
:
2137
45
.
8.
Ferrara
F
,
Lessi
F
,
Vitagliano
O
,
Birkenghi
E
,
Rossi
G
. 
Current therapeutic results and treatment options for older patients with relapsed acute myeloid leukemia
.
Cancers
2019
;
11
:
224
.
9.
Maiti
A
,
Rausch
CR
,
Cortes
JE
,
Pemmaraju
N
,
Daver
NG
,
Ravandi
F
, et al
Outcomes of relapsed or refractory acute myeloid leukemia after frontline hypomethylating agent and venetoclax regimens
.
Haematologica
2020 Jun 4
[Epub ahead of print]
.
10.
Barr
FA
,
Silljé
HHW
,
Nigg
EA
. 
Polo-like kinases and the orchestration of cell division
.
Nat Rev Mol Cell Biol
2004
;
5
:
429
40
.
11.
Degenhardt
Y
,
Lampkin
T
. 
Targeting polo-like kinase in cancer therapy
.
Clin Cancer Res
2010
;
16
:
384
9
.
12.
Renner
AG
,
Dos Santos
C
,
Recher
C
,
Bailly
C
,
Créancier
L
,
Kruczynski
A
, et al
Polo-like kinase 1 is overexpressed in acute myeloid leukemia and its inhibition preferentially targets the proliferation of leukemic cells
.
Blood
2009
;
114
:
659
62
.
13.
Gutteridge
REA
,
Ndiaye
MA
,
Liu
X
,
Ahmad
N
. 
Plk1 inhibitors in cancer therapy: from laboratory to clinics
.
Mol Cancer Ther
2016
;
15
:
1427
35
.
14.
Döhner
H
,
Lübbert
M
,
Fiedler
W
,
Fouillard
L
,
Haaland
A
,
Brandwein
JM
, et al
Randomized, phase 2 trial of low-dose cytarabine with or without volasertib in AML patients not suitable for induction therapy
.
Blood
2014
;
124
:
1426
33
.
15.
Döhner
H
,
Symeonidis
A
,
Sanz
A
,
Deeren
D
,
Demeter
J
,
Anagnostopoulos
A
, et al
Phase III randomized trial of volasertib plus low-dose cytarabine (LDAC) versus placebo plus LDAC in patients aged ≥65 years with previously untreated AML, ineligible for intensive therapy
.
Haematologica
2016
;
101
:
1
881
.
16.
Valsasina
B
,
Beria
I
,
Alli
C
,
Alzani
R
,
Avanzi
N
,
Ballinari
D
, et al
NMS-P937, an orally available, specific small-molecule polo-like kinase 1 inhibitor with antitumor activity in solid and hematologic malignancies
.
Mol Cancer Ther
2012
;
11
:
1006
16
.
17.
Casolaro
A
,
Golay
J
,
Albanese
C
,
Ceruti
R
,
Patton
V
,
Cribioli
S
, et al
The Polo-like kinase 1 (PLK1) inhibitor NMS-P937 is effective in a new model of disseminated primary CD56+ acute monoblastic leukaemia
.
PLoS One
2013
;
8
:
e58424
.
18.
Weiss
GJ
,
Jameson
G
,
Von Hoff
DD
,
Valsasina
B
,
Davite
C
,
Di Giulio
C
, et al
Phase I dose escalation study of NMS-1286937, an orally available Polo-like kinase 1 inhibitor, in patients with advanced or metastatic solid tumors
.
Invest New Drugs
2018
;
36
:
85
95
.
19.
Cheson
BD
,
Bennett
JM
,
Kopecky
KJ
,
Büchner
T
,
Willman
CL
,
Estey
EH
, et al
Revised recommendations of the international working group for diagnosis, standardization of response criteria, treatment outcomes, and reporting standards for therapeutic trials in acute myeloid leukemia
.
J Clin Oncol
2003
;
21
:
4642
9
.
20.
Levis
M
,
Brown
P
,
Smith
BD
,
Stine
A
,
Pham
R
,
Stone
R
, et al
Plasma inhibitory activity (PIA): a pharmacodynamic assay reveals insights into the basis for cytotoxic response to FLT3 inhibitors
.
Blood
2006
;
108
:
3477
83
.
21.
Youden
WJ
. 
Index for rating diagnostic tests
.
Cancer
1950
;
3
:
32
5
.
22.
Robin
X
,
Turck
N
,
Hainard
A
,
Tiberti
N
,
Lisacek
F
,
Sanchez
J-C
, et al
pROC: an open-source package for R and S+ to analyze and compare ROC curves
.
BMC Bioinformatics
2011
;
12
:
77
.
23.
Döhner
H
,
Estey
E
,
Grimwade
D
,
Amadori
S
,
Appelbaum
FR
,
Büchner
T
, et al
Diagnosis and management of AML in adults: 2017 ELN recommendations from an international expert panel
.
Blood
2017
;
129
:
424
47
.
24.
Yeh
P
,
Dickinson
M
,
Ftouni
S
,
Hunter
T
,
Sinha
D
,
Wong
SQ
, et al
Molecular disease monitoring using circulating tumor DNA in myelodysplastic syndromes
.
Blood
2017
;
129
:
1685
90
.
25.
Nakamura
S
,
Yokoyama
K
,
Shimizu
E
,
Yusa
N
,
Kondoh
K
,
Ogawa
M
, et al
Prognostic impact of circulating tumor DNA status post–allogeneic hematopoietic stem cell transplantation in AML and MDS
.
Blood
2019
;
133
:
2682
95
.
26.
Cucchi
U
,
Gianellini
LM
,
De Ponti
A
,
Sola
F
,
Alzani
R
,
Patton
V
, et al
Phosphorylation of TCTP as a marker for polo-like kinase-1 activity in vivo
.
Anticancer Res
2010
;
30
:
4973
85
.
27.
Goroshchuk
O
,
Kolosenko
I
,
Vidarsdottir
L
,
Azimi
A
,
Palm-Apergi
C
. 
Polo-like kinases and acute leukemia
.
Oncogene
2019
;
38
:
1
16
.
28.
Strebhardt
K
. 
Multifaceted polo-like kinases: drug targets and antitargets for cancer therapy
.
Nat Rev Drug Discov
2010
;
9
:
643
60
.
29.
Stahl
M
,
DeVeaux
M
,
Montesinos
P
,
Itzykson
R
,
Ritchie
EK
,
Sekeres
MA
, et al
Hypomethylating agents in relapsed and refractory AML: outcomes and their predictors in a large international patient cohort
.
Blood Adv
2018
;
2
:
923
32
.