A phase 2 study of KX2-391, an oral inhibitor of Src kinase and tubulin polymerization, in men with bone-metastatic castration-resistant prostate cancer

Abstract

Purpose

KX2-391 is an oral non-ATP-competitive inhibitor of Src kinase and tubulin polymerization. In phase 1 trials, prostate-specific antigen (PSA) declines were seen in patients with advanced prostate cancer. We conducted a single-arm phase 2 study evaluating KX2-391 in men with chemotherapy-naïve bone-metastatic castration-resistant prostate cancer (CRPC).

Methods

We treated 31 patients with oral KX2-391 (40 mg twice-daily) until disease progression or unacceptable toxicity. The primary endpoint was 24-week progression-free survival (PFS); a 50 % success rate was pre-defined as clinically significant. Secondary endpoints included PSA progression-free survival (PPFS) and PSA response rates. Exploratory outcomes included pharmacokinetic studies, circulating tumor cell (CTC) enumeration, and analysis of markers of bone resorption [urinary N-telopeptide (uNTx); C-telopeptide (CTx)] and formation [bone alkaline phosphatase (BAP); osteocalcin].

Results

The trial closed early after accrual of 31 patients, due to a pre-specified futility rule. PFS at 24 weeks was 8 %, and median PFS was 18.6 weeks. The PSA response rate (≥30 % decline) was 10 %, and median PPFS was 5.0 weeks. Additionally, 18 % of men with unfavorable (≥5) CTCs at baseline converted to favorable (<5) CTCs with treatment. The proportion of men with declines in bone turnover markers was 32 % for uNTx, 21 % for CTx, 10 % for BAP, and 25 % for osteocalcin. In pharmacokinetic studies, median C _max was 61 (range 16–129) ng/mL, and median AUC was 156 (35–348) ng h/mL. Common toxicities included hepatic derangements, myelosuppression, fatigue, nausea, and constipation.

Conclusion

KX2-391 dosed at 40 mg twice-daily lacks antitumor activity in men with CRPC, but has modest effects on bone turnover markers. Because a C _max of ≥142 ng/mL is required for tubulin polymerization inhibition (defined from preclinical studies), higher once-daily dosing will be used in future trials.

Appendix: Derivation of 142 ng/mL as the plasma threshold for KX2-391 activity

KX2-391 is 83 % protein-bound in the serum. This correlates with an approximate fourfold reduction in potency for both mechanisms of action (Src inhibition, and tubulin polymerization inhibition) measured in whole cells in the presence of human plasma. The IC₅₀ for Src inhibition in human tumor cells is about 25 nM in the absence of human plasma and about 100 nM in the presence of human plasma. However, the IC₅₀ for tubulin polymerization inhibition in human tumor cells is about 125 nM in the absence of plasma and about 500 nM in the presence of plasma. 500 nM of KX2-391 corresponds to a plasma concentration of 216 ng/mL. In mouse HT29 xenograft studies, it was determined that KX2-391 partitions into tumor tissue versus plasma at a ratio of 1.52. This reduces the required plasma concentration by 1.52-fold (i.e. 216/1.52 = 142 ng/mL). Based upon these assumptions, both mechanisms of action of KX2-391 will be engaged in patients when plasma levels reach approximately 142 ng/mL or higher (which should result in tumor tissue levels of at least 216 ng/mL). In addition, the drug level needs to remain above 142 ng/mL for long enough to cause significant apoptosis in tumor cells. An 80 mg dose of KX2-391 produces a C _max of 242 ng/mL, and plasma levels stay above the 142 ng/mL threshold for about 3 h [14]. This suggests that significant efficacy may be obtained with KX2-391 when it is dosed at or above 80 mg in patients. Preclinical toxicity and efficacy studies in mice indicated that once-daily dosing of KX2-391 was less toxic, allowing at least threefold higher drug doses to be administered, and appeared more efficacious than twice-daily dosing. Once-daily KX2-391 dosing in ongoing trials (NCT01397799) suggests that the maximum tolerated dose will be above 80 mg.