A Phase Ib/II Study of Sabizabulin, a Novel Oral Cytoskeleton Disruptor, in Men with Metastatic Castration-resistant Prostate Cancer with Progression on an Androgen Receptor–targeting Agent

Androgen deprivation therapy is the mainstay treatment for men with advanced prostate cancer. Unfortunately, almost all men will develop castration-resistant prostate cancer. Since 2011, several novel androgen receptor (AR)-targeting agents have been approved for metastatic prostate cancer including abiraterone acetate, enzalutamide, apalutamide, and darolutamide. In the first-line treatment of metastatic castration-resistant prostate cancer (mCRPC), novel AR-targeting agent therapy is effective for most patients (1). Unfortunately, sequential use of an alternative AR-targeting agent therapy results in a modest and short-lived benefit suggesting cross-resistance between different agents (2). Chemotherapy is an effective treatment for men with mCRPC; however, toxicities, including neurotoxicity and neutropenia, may limit the utility of taxanes in some patients (3, 4). There remains an unmet clinical need to develop well-tolerated alternatives to infusional chemotherapy for the treatment of patients with mCRPC.

Microtubules are important dynamic intracellular structures composed of polymerized α and β tubulin heterodimers which are critically involved in a wide range of cellular processes including maintenance of cell shape (cytoskeleton), motility, intracellular transport network for macromolecules and receptors (intracellular trafficking), and cell division (3, 4). Microtubule dynamics, that is, the balance between polymerization and depolymerization, are a validated molecular target in cancer therapeutics (3, 4). There are at least three major classes of drug-binding sites on the β tubulin subunit of microtubules: taxanes, vinca alkaloids, and colchicine (3, 4). As a result of utilizing different binding sites, the efficacy and safety profiles of each type of drug class are distinct. For instance, taxanes stabilize and prevent depolymerization of microtubules, whereas drugs that target the colchicine-binding site inhibit microtubule polymerization and induce depolymerization (3, 5). Furthermore, drugs that target different sites on microtubules can be used in combination with each other, producing synergistic anticancer activity (4).

Sabizabulin (VERU-111) is an oral 2-aryl-4-benzoyl imidazole small molecule that is part of a novel class of microtubule-targeting drugs that disrupts the cytoskeleton by targeting the colchicine binding site of β tubulin as well as binding, by strong hydrogen bonds, to a unique site on α tubulin to cross-link α and β tubulin subunits causing low nanomolar inhibition of microtubule formation and induction of microtubule depolymerization (5–11). This unique mechanism of action causes microtubule fragmentation and disruption of the cytoskeleton which is different than what has been observed with other microtubule-targeting agents (8). Sabizabulin also directly represses transcription of β tubulin isoforms including βIII and βIV, a common mechanism of drug resistance (8). In addition, treatment with sabizabulin arrests cell growth in G₂–M of cell cycle and activates caspase 3 and 9 and cleaves PARP to induce apoptosis (8, 12).

In preclinical cell line and xenograft cancer models, sabizabulin demonstrated low nanomolar inhibition of cell proliferation and tumor growth, prevention of cancer cell invasion and metastases, and suppression of angiogenesis across a broad range of tumor types including those with acquired resistance to taxanes (6–8, 12–14). In prostate cancer, sabizabulin inhibited cellular proliferation and xenograft growth of a variety of prostate cancers including those with AR splice mutations (AR-V7), BRCA1 and BRCA2 mutations, acquired taxane resistance, as well as resistance to novel AR-targeting agents (6, 7, 15, 16). Moreover, sabizabulin may overcome common drug resistance mechanisms because it is not substrate for proteins involved in multidrug resistance including P-glycoprotein, multidrug resistance proteins, and breast cancer resistance protein (6, 7, 17).

Sabizabulin's unique targeting of α and β tubulin subunits may explain the different preclinical safety profile when compared with other microtubule-targeting agents. Liver toxicity, neurotoxicity, and myelosuppression (neutropenia) in mice, rats, and dogs were not observed with sabizabulin treatment (6, 7, 17). Sabizabulin has good penetration of the blood–brain barrier and is also not a substrate for CYP3A4, which reduces the likelihood of major drug–drug interactions (6, 7, 15, 17).

As a result of the promising spectrum of antitumor activity and unique safety profile in preclinical castration-resistant prostate cancer models, a first-in-human, phase Ib/II study of sabizabulin treatment in men with mCRPC following disease progression on at least one novel AR-targeting agent therapy was conducted with the following objectives: (i) to determine the MTD and the recommended phase II dose (RP2D); (ii) to characterize the safety profile; (iii) to assess the preliminary evidence of antitumor efficacy; and (iv) to compare the pharmacokinetic profile of two different dose formulations.

This was a multicenter, open-label, phase Ib/II study (NCT03752099) conducted in the United States. The phase Ib study enrolled 39 men using a 3+3 design (18) with escalating oral daily dosing of 4.5 to 81 mg sabizabulin using a 21-day cycle of 7 days on/14 days off sabizabulin. Dose-escalation continued until the required dose-limiting toxicity (DLT) was reached to determine both the MTD and the RP2D. The DLT period was defined as the first 4 weeks of treatment. For patients that tolerated the oral sabizabulin dosing in the first 4 weeks, intrapatient dose escalation and increased frequency of dosing (up to daily dosing) was allowed. For instance, each patient that completed the initial 4 week window without a DLT would then continue at the current dose level but increase the dosing frequency to 2 weeks on, 1 week off. If safe and well tolerated at this schedule, patients would then take sabazabulin daily. All patients that tolerated drug without a DLT at daily dosing could be increased to the next dose level (provided this dose level was determined to be safe). In the phase II portion of the study, 41 men were enrolled and all were dosed at the RP2D (63 mg daily). The primary objective of the phase Ib/II study was to investigate safety of sabizabulin and to determine the MTD and RP2D. Secondary objectives included assessing preliminary evidence of antitumor activity by Prostate Cancer Working Group 3 criteria (19) and RECIST 1.1 (20). This study was Institutional Review Board approved and written informed consent was obtained from all of the patients on the study. The trial was conducted in compliance with the principles of the Declaration of Helsinki and in accordance with the International Conference on Harmonization guidelines for Good Clinical Practice.

Key inclusion criteria included: >18 years of age with mCRPC, defined according to PCWG3 criteria; a PSA concentration ≥2.0 ng/mL at screening; and an Eastern Cooperative Oncology Group (ECOG) performance status ≤ 2. Prior treatment with at least one novel AR-targeting agent (e.g., abiraterone/prednisone, enzalutamide, apalutamide, and darolutamide) was required. In the phase Ib study, one prior taxane chemotherapy for mCRPC was allowed, but not required. In the phase II study, treatment with chemotherapy for mCRPC was not permitted. The full inclusion/exclusion criteria are provided in the Supplementary Data.

The MTD was considered reached when a DLT was observed in either 2 of the first 3 patients or 2 of the first 6 patients enrolled at that dose. Patients were assessed for toxicities at each clinical evaluation. Toxicities were graded according to Common Terminology Criteria for Adverse Events v5.0 standardized grading scales. A DLT was defined as follows: any grade 4 neutropenia or grade 3 febrile neutropenia; grade 4 thrombocytopenia or grade 3 thrombocytopenia with hemorrhage; grade 3–4 non-hematologic toxicity (except for: grade 3 nausea, vomiting, and diarrhea or grade 4 vomiting and diarrhea persisting <72 hours in patients who have received suboptimal prophylaxis; grade 3 fatigue <7 days; grade 3–4 laboratory abnormalities that are asymptomatic and corrected within 72 hours); an increase in alanine aminotransferase/aspartate aminotransferase >3× upper limit of normal (ULN) with a concomitant increase in total bilirubin >2× ULN; any toxicity that in the judgment of the investigators poses a risk to the subjects’ safety at a given dose level.

Clinical efficacy was assessed by radiographic progression-free survival (rPFS) in all patients who received ≥63 mg of sabizabulin in the phase Ib and phase II portions of the study. rPFS was measured from the time of first dose to radiographic tumor progression as defined by PCWG3 for progressive disease or death. Patients who did not have radiographic progression were censored at the data analysis cut-off date. The objective response rate (ORR) was measured as the percentage of patients with measureable disease who achieved an objective response by RECIST 1.1 (i.e., complete response or partial response) on sabizabulin. Serial PSA measurements were also obtained.

The phase Ib study portion was designed as a standard 3+3 dose escalation with 3 patients per dose group as needed to define the MTD and identify the RP2D. For the phase II study, assuming a null hypothesis of 25% and a RECIST-defined radiographic response rate of ≥ 50%, with an α = 0.05 and a power of 80%, the required sample size was 26 subjects. To account for potential early dropouts (for reasons other than disease progression or dose-limiting serious adverse events (SAEs), an additional 9 patients were included in the phase II portion of the study (n = 35). Seven patients were enrolled on study and allowed to continue treatment on a novel AR-targeting agent therapy despite progression. Unadjusted Kaplan–Meier curves were used to estimate rPFS.

In the phase Ib/II study, six subjects were included in a pharmacokinetic substudy. On day 1 of the pharmacokinetic substudy, the pharmacokinetics of the 63 mg powder-in-capsule (PIC) formulation administered fasted was assessed. On day 2, the pharmacokinetics of the 63 mg PIC formulation administered fed was assessed. On days 3–6, the patients were administered 31.5 mg formulated capsule (FC) formulation. On day 7, the pharmacokinetics of the 31.5 mg FC formulation was assessed. On day 8, the pharmacokinetics of the 31.5 mg FC formulation was assessed.

The data generated in this study are available upon request from the corresponding author.

A total of 80 patients in the United States were enrolled from February 2019 to September 2020: 39 patients were enrolled in the phase Ib study at 7 U.S. centers; 41 patients were enrolled in the phase II study across 13 U.S. centers. Baseline characteristics in both the phase Ib and phase II studies are shown (Table 1). In the phase Ib study (n = 39), the median age was 74 years (range, 61–92) and the majority had a performance status of ECOG 0. Most patients had bone-only disease (55%) with an additional 21% having both lymph node and bone involvement. This study population included 23% of patients with prior taxane-based chemotherapy. Moreover, 44% of patients had prior treatment with both abiraterone and enzalutamide. In the phase II study (n = 41), a similar patient population was enrolled with the exception of prior chemotherapy in the metastatic setting which was an exclusion criteria.

Patient disposition for the phase Ib and II are illustrated on the CONSORT diagram (Supplementary Figs. S1A and S1B) and a swimmers plot of patients on the phase Ib is illustrated in Fig. 1. As of the data cut-off date of October 01, 2021, 37 of 39 had patients completed treatment on the phase Ib trial and were discontinued for the following reasons: n = 9 withdrew consent, n = 5 adverse event, n = 21 investigator discretion, n = 1 death (unrelated to study drug), n = 1 radiographic progression. Two patients remain on study in the absence of disease progression for more than 2.75 years. During dose escalation, the adverse events observed at doses less than 54 mg daily were grade < 2 events (Supplementary Tables S1 and S2). At the 54 mg dose level, grade 3 fatigue (n = 1, 11%) occurred. Grade 3 nausea (n = 1; 7%), vomiting (n = 1; 7%), and fatigue (n = 1; 7%) were observed in the 63 mg cohort. In the 72 mg cohort, grade 3 diarrhea (n = 3, 23%), fatigue (n = 1, 8%), nausea (n = 1, 8%) were reported. The highest dose level tested was 81 mg [causing grade 3 anemia (n = 1, 33%) and nausea (n = 1, 33%)]. On the basis of the prespecified DLT criteria, the MTD was not reached. Given the number of grade ≥ 3 adverse events observed in the ≥ 72 mg/day cohorts, the 63 mg/day dosing was selected as the RP2D. Table 2 summarizes all grades ≥3 treatment emergent adverse events in the phase Ib patients treated at the 63 mg dose and the phase II patient population (total N = 55 patients). Most adverse events were low grade (≤2), which included diarrhea (59%), fatigue (33%), and nausea (31.5%). At the time of data analysis, 3 patients remain on study. Reason for study removal are as follows: n = 5 adverse events, n = 16 principal investigator decision, n = 9 withdrew consent, n = 8 radiographic progression. Neutropenia, neurotoxicity, and alopecia were not observed on study.

In the phase II study, 41 patients with mCRPC were enrolled and treated with 63 mg per day taken orally on a continuous schedule. For the efficacy analysis, we included 14 patients from the phase Ib study that were treated with at least one cycle of 63 mg sabizabulin on a daily basis for a total of 55 patients (Table 2). Of the 29 patients with measureable disease, the ORR was 20.7% (n = 6; 5 partial, 1 complete). PSA declines were observed in 14 of 48 evaluable patients including 4 of 48 with ≥50% decline (Table 3; Fig. 2). Although patients remain on study, the median rPFS for the 55 patients that received at least one cycle of 63 mg or greater was 11.4 months [95% confidence interval (CI): 29.63–65.79] (Fig. 3). For the evaluable patients with measurable disease at baseline (n = 29), the median rPFS is 10.8 months (95% CI: 25.80–66.26). Durable responses lasting >12 months occurred in 14.5% (n = 8/55) of the patients receiving 63 mg daily dosing (examples of which are presented in Supplementary Fig. S2).

The outcomes on the 7 patients treated with the alternate AR-targeted drug combined with sabizabulin were similar to the entire group in terms of safety and efficacy. One patient had radiographic progression at 4.8 months, 5 discontinued treatment without evidence of radiographic progression, and 1 had an adverse event. One patient in this group remained on study until discontinuation at 11.9 months. There were no differences in adverse event types or rates in this small subset of patients.

The T_1/2 of sabizabulin is approximately 5 hours and steady state is reached within 5 days of daily dosing (Supplementary Table S2). Administration of sabizabulin with a high-fat meal has minimal effect on the extent of bioavailability (AUC_0–24h) of the sabizabulin in both the PIC and FC formulations. However, with a high-fat meal, the T_max was delayed in both formulations and resulted in a lower C_max with the PIC formulation. The relative bioavailability of the sabizabulin 32 mg FC formulation was similar to the 63 mg PIC formulation suggesting a similar pharmacokinetic profile.

This phase Ib/II clinical trial in men with mCRPC who received extensive prior therapies utilized a dynamic study design allowing for intrapatient dosing changes/escalation along with classic 3+3 principles. During the dose-escalation portion of the study, the MTD was not defined. Because of the grade 3 diarrhea observed at doses of 72 and 81 mg, the once-daily 63 mg dose given on a continuous basis was selected as the RP2D. Moreover, based on pharmacokinetic studies, the AUC observed at the 63 mg daily dosing schedule was predicted to be the effective dose.

By choosing a 63 mg daily dosing strategy, we hypothesized that toxicities would be tolerable while also maintaining efficacy. This is supported, in part, based on the 55 patients who received ≥63 mg dosed daily in the phase Ib and II portions of the study. In these patients, gastrointestinal (GI) toxicities were frequently observed, common toxicities among orally dose drugs. These GI adverse events were low grade and improved with supportive care (i.e., anti-diarrheals) or brief drug interruption. Importantly, neutropenia or neurotoxicity (i.e., peripheral neuropathy), a known toxicity of taxane chemotherapy was not observed with sabizabulin treatment at any dose/schedule level. The overall mild toxicity profile of sabizabulin at 63 mg daily indicates that long term dosing is feasible and well tolerated. With respect to dosing strategies in future studies, we conducted a pharcokinetic substudy using two different dosage formulations—63 mg PIC and 32 mg FC—both dosed on a daily basis. Both formulations produced similar serum drug concentrations, albeit with different dosages. The 32 mg FC formulation resulted in nearly double the oral bioavailability comparable with 63 mg PIC (Supplementary Table S3). We hypothesize that using a 32 mg FC formulation will result in less residual unabsorbed drug in the GI tract which may improve GI tolerability.

As indicated in the Results, 7 patients were treated with an alternative AR-targeting agent combined with sabizabulin at the time of their study entry. Overall, these patients performed similarly to those that received sabizabulin alone. Although the numbers are small, the outcomes from these patients were similar to the single-agent sabizabulin data. More information if necessary to assess efficacy and safety of the combination; however, this would seem to be best evaluated combining a novel AR-targeting agent with sabizabulin in the first-line setting for patients with mCRPC.

In both the phase Ib and II clinical studies, preliminary evidence of efficacy by sabizabulin was observed. At 63 mg dosed daily, both PSA and objective responses were observed, with 5 patients remaining on study for over 1 year. These efficacy signals along with a favorable safety profile warrant further study of sabizabulin in men with mCRPC.

Sabizabulin disrupts the cytoskeleton by targeting the colchicine binding site on β tubulin as well as binding by strong hydrogen bonds to a unique site on α tubulin to cross-link α and β tubulin subunits of microtubules to inhibit microtubule formation and cause microtubule depolymerization (5–9, 11, 16). This unique binding, targeting, and cross-linking action causes microtubule fragmentation and disruption of the cytoskeleton which is different from vinca alkaloids or taxanes (8). Sabizabulin's unique target sites of α and β tubulin subunits appears to result in a different preclinical safety profile than observed for the other microtubule-targeting agents especially neurotoxicity, and myelosuppression (neutropenia) in mice, rats, and dogs (6, 7, 17). This favorable safety profile is consistent with the findings on this phase Ib/II clinical study. While targeting the microtubules, there are several properties that distinguish sabizabulin from taxanes and other microtubules targeting agents. Among these are a distinct chemical structure, mode of action, preclinical and clinical safety profile, in addition to chemical and pharmacologic properties that allow for safe long term daily oral administration. Chronic exposure of cancer cells to this class of targeted agent could potentially enhance effects on the spindle foundation and mitosis, and consequently, optimize cytostatic properties.

Additional testing of sabizabulin in prostate cancer can be carried out in various clinical states of the disease. Further definition of the benefits of sabizabulin would be best served with randomized comparisons against suitable controls considered appropriate for individual clinical states where robust therapeutic benefits remain an unmet need. While frequently used in clinical practice, increasing evidence suggest that sequential administration of AR-targeting agents in men with mCRPC is associated with substantial loss of efficacy of the alternate compound (21–24). In view of the similarities between the two approaches, a comparison between a second-line AR-targeting agent and sabizabulin in the prechemotherapy space in men with mCRPC could represent a reasonable and appropriate test for this novel compound in phase III development.

Chronic daily administration of oral sabizabulin has excellent bioavailability and a reasonable safety profile. At the recommended phase II dose of 63 mg daily preliminary evidence of clinical benefit (including PSA and objective responses, some of which were durable). Further testing of sabizabulin in mCRPC is warranted in a phase III clinical study.

R. Tutrone owns less than $25,000 in Veru Inc. stock. C. Pieczonka reports other support from Veru during the conduct of the study; personal fees from Pfizer, Astellas, Myovant, Bayer, AstraZeneca, Dendreon, and Janssen outside the submitted work. R.H. Getzenberg reports personal fees and other support from Veru Inc. during the conduct of the study. D. Rodriguez reports a patent for PCT/US21/52209 pending; and is employee of Veru Inc. M.S. Steiner reports other support from Veru Inc. during the conduct of the study; other support from Veru Inc. outside the submitted work; in addition, M.S. Steiner has a patent for sabizabulin pending, issued, and licensed to Veru Inc. M.A. Eisenberger reports other support from Veru Inc. during the conduct of the study; other support from Veru Inc. outside the submitted work. E.S. Antonarakis reports personal fees from Amgen, Bayer, Blue Earth, Bristol Myers Squibb, Curium, Eli Lilly, ESSA, Exact Sciences, Foundation Medicine, GlaxoSmithKline, Invitae, Janssen, Johnson & Johnson, Merck, and Tempus; grants from Celgene and Sanofi; and grants and personal fees from AstraZeneca and Clovis during the conduct of the study; in addition, E.S. Antonarakis has a patent for an AR-V7 biomarker technology issued and licensed to QIAGEN. No disclosures were reported by the other authors.

M.C. Markowski: Conceptualization, formal analysis, investigation, writing–review and editing. R. Tutrone: Investigation, writing–review and editing. C. Pieczonka: Investigation, writing–review and editing. K.G. Barnette: Conceptualization, data curation, formal analysis, methodology, writing–review and editing. R.H. Getzenberg: Conceptualization, resources, data curation, supervision, writing–original draft, writing–review and editing. D. Rodriguez: Conceptualization, data curation, methodology, writing–review and editing. M.S. Steiner: Conceptualization, resources, formal analysis, supervision, methodology, writing–original draft, writing–review and editing. D.R. Saltzstein: Investigation, writing–review and editing. M.A. Eisenberger: Conceptualization, formal analysis, methodology, writing–review and editing. E.S. Antonarakis: Conceptualization, investigation, methodology, writing–review and editing.

We would like to extend our deepest appreciation to the patients who participated on this study. We also recognize and appreciate the tireless of work of the clinical trial nurses, coordinators, and research staff. This work was sponsored by Veru Inc.

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.