PF also a dual Aurora

PF-03814735, also a dual Aurora-A and Aurora-B inhibitor, was administered in a phase I dose escalation study in a 3-weekly alpha msh for 5 or 10 consecutive days [69], [70]. To date 25 patients have been included in 7 dose levels (5–100mg/day for 5 days). Using the 5-day schedule the MTD was defined as 80mg/day and DLT consisted of febrile neutropenia in two patients at 100mg/day. Other reported adverse events included grade 1/2 diarrhea, vomiting, anorexia, fatigue and nausea. Using this dosing schedule two patients presented disease stabilization. Pharmacokinetic evaluation showed a total PF-03814735 clearance of 1.25±0.39L/h and a median terminal half-life of 20.2h after a single dose of PF-03814735. To date, dose escalation on the 10-day dosing schedule is still ongoing and no results have yet been reported. SNS-314, a selective inhibitor of Aurora kinases A, B and C, was administered as a 3h i.v. infusion once a week (days 1, 8 and 15 in a 28-day cycle) [71]. Dosing started at 30mg/m2 with subsequent dose escalation doubling. To date no DLTs could be observed. Terminal half-life was calculated to be 7h and PK parameters were similar when comparing the 1st- and the 3rd-weekly dose. Skin biopsies revealed inhibition of histone H3 phosphorylation when doses of >240mg/m2 were administered. Dose escalation is still ongoing. Several phase I studies have been conducted to date using different dosing schedules of either oral or intravenous Aurora-A and/or Aurora-B inhibitors in heavily pre-treated patients. The different drugs used were generally well tolerated and apart from grade 3 somnolence [65], grade 2 hypertension, grade 3 fatigue [65] and grade 3 diarrhea [66], neutropenia was the most observed DLT (Table 2). Other grades 1 and 2 toxicities included gastro-intestinal toxicities like nausea, vomiting, diarrhea, and anorexia. In general the tolerability of this group of compounds can be considered favorable. In the studies available to date a low inter individual variability in pharmacokinetic parameters (Table 3) has been reported for this class of drugs, thereby increasing their reliability and predictability when used in the individual patient. Furthermore, all studies report several clinically relevant disease stabilizations some of which even last for more than 6 [66] or 7 months [65], making further development of these drugs even more promising.
Conclusion It is presently not known whether inhibition of Aurora kinase A or B or both should be pursued. Pre-clinical data are emerging indicating that the cellular events observed after exposure to non-selective Aurora kinase inhibitors are due to Aurora-B inhibition. Also in clinical studies with PHA-739358 inhibition of histone H3 phosphorylation could be observed in skin biopsies reflecting Aurora-B inhibition [65]. A clinical marker for Aurora-A inhibition is presently not available. Both in the phase I studies with PHA-680632 and MK-0457, both inhibiting Aurora kinase A, B and C, and the phase I study with AS703569, inhibiting Aurora kinase A and B but not C, several disease stabilizations have been reported. The only more selective Aurora kinase inhibitor with reported disease stabilization in a phase I study is AZD1152, a selective Aurora-B kinase inhibitor. Clinical activity data are not yet reported for the selective Aurora kinase A inhibitors. Selective inhibition might reduce toxicity of the compounds and allow higher dosing. Since clinical development is still in its early stages no definitive clinical conclusion can be drawn on a preference for selective or broader Aurora kinase inhibition. Presently several of the Aurora kinase inhibitors have entered phase II testing. Unanswered but very important issues will be the possible predictive factors for response. An important factor for tumor cell sensitivity could be the presence of intact cell cycle checkpoints in the cell. Cells lacking CHFR, a mitotic checkpoint protein, or cells with a non-functional p53 are more sensitive to Aurora kinase inhibition, thereby providing an advantage for healthy tissue over tumor cells. If this could be confirmed in clinical studies a biomarker guided development strategy would become conceivable.