Ispinesib and related compounds

Chemo Secrets From a Breast Cancer Survivor

Breast Cancer Survivors

Get Instant Access

2.1.1. Clinical progress on ispinesib and SB-743921


Ispinesib (SB-715992/CK0238273, 2), one member of a KSP inhibitor series bearing a quinazolinone core, was the first mitotic kinesin inhibitor to enter clinical trials. Like its earlier reported analog 1, ispinesib was reported to be an allosteric inhibitor of KSP that binds at the motor domain and inhibits its ATPase activity in an ATP uncompetitive manner. It was also shown to be > 70,000-fold selective for KSP versus other members of the kinesin family. Results have been reported from two Phase I clinical trials with different dosing schedules. In trial KSP10001 [11], ispinesib was dosed i.v. once every 21 days (q21) in 45 patients. The drug was well tolerated without alopecia or prevalent neurotoxicity. At a dose of 21 mg/m2, dose-limiting toxicities (DLT) were prolonged (>5 days) neutropenia and febrile neutropenia. Prolonged disease stabilization, ranging from 7 to 43 weeks, was observed in 15 of 45 patients representing a variety of tumors. In addition, evidence of minimal response (<50% tumor mass decrease) was noted in three patients. Ispinesib had a mean (range) Cl of 98 (48-141) ml/min, Vss of 256 (162-514)1, and t1/2 of 34 (17-56) h at a dose of 18mg/m2, which was the dose level selected for Phase II trials on the same q21-day schedule. In trial KSP 10002 [12], ispinesib was dosed i.v. on days 1, 8, and 15, one cycle every 28 days in 30 patients. The DLT observed at 8 mg/m2 was grade 3 neutropenia preventing administration of the day 15 dose during cycle 1. The maximum tolerated dose (MTD) on this schedule was 7 mg/m2. Prolonged disease stabilization ranging from 16 to over 32 weeks was evident in eight of 30 patients. At the MTD, median Cl was 94 ml/min and median t1/2 was 33 h, very similar to results in the KSP10001 trial. The desired cellular phenotype, mitotic arrest with monopolar spindle formation, was observed in tumor biopsies in both Phase I trials.

Ispinesib has also been examined in Phase Ib clinical combination with docetaxel [13], capecitabine [14], and carboplatin [15]. The DLT for the combination of ispinesib and docetaxel was prolonged (X5 days) grade 4 neutropenia, and the optimally tolerated regimen (OTR) was defined as 10mg/m2 of ispinesib and 60mg/m2 of docetaxel, each administered once every 21 days. Ispinesib and docetaxel plasma concentrations were consistent with those previously reported when each drug was given as monotherapy, suggesting no pharmacokinetic interaction between them. A total of 13/24 patients (11 prostate, 1 renal, and 1 bladder) had a best response of stable disease (duration 2.25-7.5 months) [13].

For the combination of ispinesib capecitabine, the DLT and OTR had yet to be defined, although ispinesib had been administered at 18 mg/m2 for every 21 days in the combination with capecitabine at 2000 mg/m2 daily for 14 of 21 days. One DLT of prolonged (X 5 days) grade 4 neutropenia had been observed at the time of the report. Ispinesib plasma concentrations did not appear to be affected by the presence of capecitabine. A total of eight patients (3 breast, and 1 each of tongue, colorectal cancer (CRC), bladder, thyroid, and salivary gland) had a best response of stable disease (duration 2-6.5 months) [14].

For the combination of ispinesib and carboplatin, DLTs included prolonged (X5 days) grade 4 neutropenia, grade 4 thrombocytopenia, and grade 3 febrile neutropenia. The OTR was ispinesib at 18 mg/m2 (the Phase II dose) and a car-boplatin target AUC of 6 (a commonly used monotherapy target exposure), both administered for q21 days. At the OTR, gastro intestinal (GI) toxicities were limited to grade 1/2 and minimal reports of grade 1 neuropathy were noted. The incidence of grade 3/4 thrombocytopenia and grade 3/4 neutropenia at the OTR were lower relative to full doses of carboplatin and ispinesib, respectively. At the OTR, is-pinesib concentrations did not appear to be affected by carboplatin and systemic exposures of carboplatin were within 11% of predicted values, suggesting no interaction with ispinesib. One patient with breast cancer had the best response of partial response at cycle 2. A total of 13/28 (46%) patients had a best response of stable disease (duration 3-9 months) [15].

Ispinesib has been studied in eight Phase II clinical trials, including studies in patients with locally advanced or metastatic breast cancer, platinum-refractory and sensitive non-small cell lung cancer, ovarian cancer, hepatocellular cancer, colorectal cancer, head and neck cancer, hormone-refractory prostate cancer, and melanoma. In the ongoing breast cancer study, women with locally advanced or metastatic breast cancer, receive ispinesib as monotherapy at 18mg/m2 as a 1-h intravenous infusion for every 21 days. In an interim analysis of Stage 1 data from this two-stage trial, partial responses were reported in three of 33 evaluable patients. Maximum decreases in tumor size ranged from 46% to 68%, and the duration of response from 7.1 to 13.4 weeks. The overall response rate for all 33 evaluable patients was 9% with a median time to progression of 5.7 weeks. The adverse events were manageable, predictable, and consistent with the Phase I clinical trial experience with ispinesib, and ispinesib plasma concentrations were comparable to those observed in the Phase I clinical trial [11,16].

In mice with advanced MX-1 human breast carcinoma, ispinesib (30mg/m2) in combination with docetaxel (30mg/m2) or ispenisib (15mg/m2) with capecitabine (1500 mg/m2) resulted in greater tumor growth delay than with either agent alone. Cisplatin was also shown to enhance the activity of ispinesib against murine P388 lymphocytic leukemia [17], with sub-MTD doses of both in combination being superior to either agent alone at their respective MTDs.

A second KSP inhibitor, SB-743921, entered clinical trials in 2004. Like ispinesib, it has been shown to be a very selective KSP inhibitor that is > 40,000-fold more selective over other kinesins. Interim results were reported from a Phase I trial in 19 patients in which SB-743921 was administered intravenously q21 days [18]. DLT included prolonged grade 4 neutropenia, grade 3 febrile neutropenia, grade 3 elevated transaminases, grade 3 hyperbilirubinemia, and grade 3 hyponatremia. Also like ispinesib, SB-743921 did not cause neurotoxicity or alopecia. A dose of 4 mg/m2 at q21 days was suggested for Phase II. At this dose median Cl was 21 ml/ min, and t1/2 was 28 h.

2.1.2. Quinazolinone core replacement

The potent pharmacophore represented by ispinesib has received a good deal of attention recently, particularly with respect to replacement of the quinazolinone core. Among the fused bicyclic examples to appear were 3-8 that encompass both 6,6 and 6,5 ring systems [19-25]. A potent version of 7, known as BMS-601027, has been reported to have a KSP ATPase IC50 of 86 nM and a cell IC50 of 317 nM [22]. Several monocyclic analogs have also been reported including examples 9-14 that represent 6- and 5-membered replacements of the original pyrimidinone core [26-34].

R3 N

R3 N

2.2. Monastrol analogs

Ksp Chemistry

Monastrol (15) was the first reported specific inhibitor of KSP [5]. It was described as an allosteric inhibitor of motor function (IC50 = 30 mM) that inhibits ADP release both in the presence and absence of microtubules. More recently, X-ray crystal structures of the KSP-ADP complex with monastrol provided insights into the structural basis for the inhibition by this agent [35,36]. Monastrol was found to induce dramatic conformational changes in helices a2 and a3 and the insertion loop (L5) that result in the formation of an induced-fit pocket not visible in the KSP-ADP structure [35]. Based on the subsequent kinetic and thermodynamic studies it was proposed that monastrol first binds weakly to the nucleotide-free state

and/or the KSP-ATP collision complex which has an open conformation of loop L5 [37,38]. Upon tight ATP binding, a conformational change was suggested to occur that causes the inhibitor pocket to close, therefore increasing the binding affinity of monastrol to KSP. A series of spectroscopic probes were used to elucidate the pathway of structural changes in solution, and the results are consistent with the crystallographic model [39].

Analogs of monastrol have been pursued to improve biochemical and cellular potency for this scaffold [40]. Replacement of the phenyl ring or 3-OH group attenuated the activity as did replacement of the thiourea sulfur with oxygen. However, analogs with a fused bicyclic core showed improved potency. Compound 16 with an IC50 of 2 mM was 10 times more potent than monastrol. Introducing a gem-dimethyl group (compound 17) further boosted the potency to 200 nM. Compounds 16 and 17 also had corresponding improvements in cellular potency with the desired cellular phenotype observed.

2.3. Tetrahydro-ß-carbolines



R =



R =



R =



R =

Compound 18, also known as HR22C16, was identified as a KSP inhibitor (IC50 = 800 nM) from a cell-based screen. Subsequent optimization yielded more potent compounds such as 19 (IC50 = 90 nM) [41]. A recent study was reported on the synthesis and biological evaluation of 60 analogs that included both cis and trans-isomers with different R groups [42]. Trans-isomers were found to be more active than cis-isomers in general, with the absolute stereochemistry shown being most prefered. Compound 20 was the most potent among the 60 analogs with an IC50 of 650 nM against KSP. HR22C16 and its analog 21 were shown to be active in both paclitaxel-sensitive (1A9) and paclitaxel-resistant (PTX10) ovarian cell lines [43]. Compound 21 had IC50 values of 0.8 and 2.3 mM in 1A9 and PTX10 cells, respectively. Compared to a 750-fold loss of activity for paclitaxel in PgP-overexpressing ovarian cell line A2780-D10, 21 had a 2.4-fold loss, suggesting that it is not a PgP substrate. The antiproliferative activity of 21 was attributed to mitotic arrest followed by cell death, which was mediated through an intrinsic apoptotic pathway. Interestingly, 21 was found to be antagonistic with paclitaxel

treatment. Whether 21 (1 mM) was administrated concomitantly with paclitaxel (5nM) or paclitaxel was dosed first (24 h), monopolar spindle phenotype was predominant. On the other hand, paclitaxel-type (i.e. multipolar) spindles were predominant when 20 was followed by paclitaxel [43].

2.4. Dihydropyrroles and dihydropyrazoles

3,5-Diaryl-4,5-dihydropyrazoles 22 and 23 were identified as screening hits with ATPase IC50 values of 3.6 and 6.9 mM, respectively [44]. Combining chloro and hydroxy substituents in the same molecule yielded the more potent compound 24, which had an IC50 of 450 nM. Optimization work around the two phenyl groups demonstrated tolerance for 3-phenyl substitution leading to a more potent compound 25 (IC50 = 51 nM), but almost no tolerance for variation on the 5-phenyl ring. Exploration of N1-substitution with larger acyl or alkyl groups generally resulted in analogs with lower potency, whereas compound 26 with a dimethyl urea had similar potency to compound 25. The (S)-antipode of compound 25, separated by chiral phase HPLC, was found to be the active stereoisomer with an ATPase IC50

of 26 nM. It was also found to cause caspase-3 induction, a well-established marker of apoptosis, in A2780 human ovarian carcinoma cells with an IC50 of 15 nM [44].

Replacement of the dihydropyrazole core with a dihydropyrrole also yielded a viable scaffold in which the hybridization of the aryl-bearing carbon atoms was maintained. Introduction of basic amide and urea groups to the dihydropyrrole core led to enhanced potency. Compounds 27 and 28 had ATPase IC50 values of 3.6 and 2.0 nM, respectively. Both compounds caused mitotic arrest in A2780 cells with EC50's of 12 and 8.6 nM [43], respectively. X-ray crystal structures of KSP bound with inhibitors (S)-25, 26, and 28 showed binding to the same allosteric pocket as monastrol [44,45].

Cellular response and the role of the spindle checkpoint in the induction of apoptosis by compound 27 were investigated in cancer cell lines. The results indicated that both activation of the spindle assembly checkpoint and mitotic slippage were required for the induction of apoptosis. This led to the suggestion that agents that promote mitotic slippage could act synergistically with KSP inhibitors in cancer cells with competent spindle checkpoints [46].

Compound 27 showed high aqueous solubility (> 10mg/ml) and moderate clearance (21-40 ml/min/kg) with t1/2 ranging from 1-4 h in rat, dog, and monkey. Moderate human ether-a-go-go related gene (hERG) potassium channel binding was also observed for this dihydropyrrole series. Compounds 27 and 28 had hERG IC50 values of 2.4 and 3.5 mM, respectively [43], hence efforts to minimize hERG binding in this series were pursued [47]. Reintroduction of a 3-OH group to the northern phenyl ring in combination with neutral N1-acyl groups reduced hERG binding while maintaining potency. Compound 29 had an ATPase IC50 of 7nM, cell mitotic EC50 of 22 nM, and hERG IC50 of 33 mM. With diminished solubility in the absence of the basic amino group, phosphate prodrugs of 29 were made to improve solubility. Prodrug 30 was shown to be inactive against KSP (IC50> 1 mM), while its solubility was improved to > 20 mg/ml at pH 7. It was cleaved rapidly in blood to produce parent 29.

Some closely related KSP inhibitors have also been reported where the central dihydropyrrole and dihydropyrazole cores were replaced by tetrahydropyridine [48], dihydroisoxazole [49] and dihydrooxadiazole [50] groups.

2.5. Tetrahydroisoquinolines

A new series of tetrahydroisoquinolines was identified from high-throughput screening. Representative compound 31 had an IC50 of 9.7 mM in an ATPase assay and an IC50 of 2.4 mM in a proliferation assay in A2780 human ovarian carcinoma cells [51]. Nuclear magnetic resonance (NMR) experiments showed compound 31 bound to the same allosteric site as monastrol. A binding model was constructed using NMR data and crystallographic data of co-crystal structures of KSP-monastrol and KSP-HR2216 [51]. The model was then used to help guide structure-activity relationship (SAR) development for the series.

In optimization studies, the fused dihydrofuran ring of 31 was found to be replaceable with 7,8-dimethyl groups without loss of potency. Substitution on N with alkyl groups larger than methyl was not well tolerated, but analogs with amino groups appended to 2-4 carbon atom chains were shown to have similar activity to 31. This raised the possibility that the charged side chain may point toward solvent. In support of this, N-acylation products such as amides, sulfonamides, carbamates, and primary ureas were generally much less active. However, compound 32 (race-mic) with a N,N-dimethyl urea, was slightly more potent than 31 (ATPase IC50 = 2.75 mM). In an attempt to recapitulate the H-bond between monastrol's phenol-OH group and the carbonyl of back-bone amide Glu 118 of KSP, a hy-droxyl group was introduced to the 3-position of 32. This was found to increase potency by about eight-fold in the ATPase assay (IC50 = 306 nM) and three-fold in the cellular assay (IC50 = 376 nM). For comparison, the 4-OH analog was completely inactive [51]. A co-crystal structure of KSP with the (R)-antipode of 32 was obtained, and as expected, the compound binds in the known allosteric site of KSP in the predicted orientation. An H-bond between the oxygen of the phenol and the carbonyl oxygen of Glu 118 and the side chain Arg 119 was evident. Van der Waals interactions between the tetrahydroisoquinoline, dimethyl urea, and phenyl ring with the protein were also observed.

Was this article helpful?

0 0
100 Hair Growth Tips

100 Hair Growth Tips

100 Hair Growth Tips EVERY Balding Person Should Know. This Report

Get My Free Ebook

Post a comment