Southwest Oncology Group Phase II Study of Ispinesib in Androgen-Independent Prostate Cancer Previously Treated
with Taxanes
Tomasz M. Beer,1 Bryan Goldman,2 Timothy W. Synold,3 Christopher W. Ryan,1 Lakshmi S. Vasist,4 Peter J. Van Veldhuizen Jr,5 Shaker R. Dakhil,6 Primo N. Lara Jr,7 Anibal Drelichman,8 Maha H. A. Hussain,9
E. David Crawford10
1Oregon Health & Science University Cancer Institute, Portland
2Southwest Oncology Group Statistical Center, Seattle, WA 3City of Hope National Medical Center, Duarte, CA 4GlaxoSmithKline Clinical Pharmacokinetics/Modeling & Simulation, Research Triangle Park, NC
5University of Kansas Medical Center, Kansas City
6Wichita Community Clinical Oncology Program, Wichita, KS
7University of California, Davis, Sacramento 8Providence Hospital, Southfield, MI 9University of Michigan, Ann Arbor
10University of Colorado Health Sciences Center, Denver

Clinical Genitourinary Cancer,
Vol. 6, No. 2, 103-109, 2008

Keywords: Hela cells, Kinesin spindle protein, Mitosis, Peripheral blood, Prostate-specific antigen

DOI: 10.3816/CGC.2008.n.016

Submitted: Sep 7, 2008; Revised: Feb 27, 2008;
Accepted: Mar 20, 2008
Address for correspondence:
Tomasz M. Beer, MD, Department of Medicine, Oregon Health & Science University, Mail Code CH14R, 3303 SW Bond Ave, Portland, OR 97239
Fax: 503-494-6197
E-mail: [email protected]

Introduction
While docetaxel has a demonstrated effect on survival in androgen-indepen- dent prostate cancer (AIPC),1,2 there is no established therapy for AIPC after docetaxel. The success of docetaxel suggests that the mitotic spindle is an impor- tant target for the development of prostate cancer therapies.
Mitotic kinesins play exclusive and essential roles in assembly and func- tion of the mitotic spindle. They are preferentially expressed in neoplastic cells, and represent novel targets for cancer treatment.3 Ispinesib is the first mitotic kinesin spindle protein (KSP) inhibitor4 to enter clinical trials and has a broad spectrum of activity in preclinical models of cancer, including models that are refractory to chemotherapy.5,6 Because previous work with taxanes provided clinical validation for targeting the mitotic spindle in AIPC, we sought to determine if targeting another component of the mitotic spindle would prove useful.
The safety of ispinesib administered on a 3-weekly schedule was established in a phase I trial. Neutropenia was the dose-limiting toxicity at a dose of

21 mg/m2, and 18 mg/m2 was the recommended dose for phase II evaluation.7 Monopolar mitotic spindles were observed in a tumor biopsy from a patient with squamous cell carcinoma of the head and neck treated at 16 mg/m2 suggesting that biologically relevant doses of the drug are feasible. Based on these results, we selected 18 mg/m2 every 21 days for this study. In this study, we sought to evaluate the safety and activity of ispinesib in patients who had previously been treated with a taxane. In addition to these clinical goals, we sought to evaluate the population pharmacokinetics of ispinesib in these patients as well as examine the presence of the target in archival tumor specimens and evaluate peripheral blood mononuclear cells for evidence of a biologic drug effect in human subjects.
Patients and Methods
Eligibility Criteria
Eligible patients had histologically confirmed adenocarcinoma of the prostate, with clinical evidence of metastases (N1 or M1) and a life expectancy of > 12 weeks. Current evidence of disease progres- sion by imaging studies or by prostate-specific antigen (PSA)8 was required. Previous treatment with 1 taxane-containing chemothera- py regimen was required and had to be completed ≥ 28 days before enrollment. Previous radiation therapy (RT) to < 30% of the bone marrow (BM) or previous use of samarium had to be completed
≥ 28 days before enrollment, but previous strontium use was not allowed. In addition to the elapsed time requirements, patients had to have recovered from clinically significant adverse effects of che- motherapy or RT before they could participate in this study.
Patients had to have a Zubrod performance status (PS) of 0-2, serum PSA > 5 ng/mL, total serum bilirubin ≤ upper limit of normal (ULN), serum creatinine ≤ 1.5 × ULN or a calculated creatinine clearance ≥ 40 mL/min, an aspartate aminotransferase (AST) and alanine aminotransferase (ALT) ≤ 2.5 × ULN, absolute granulocyte count ≥ 1500 cells/mm3, hemoglobin (Hb) ≥ 9 mg/dL, and a platelet count ≥ 100,000 cells/mm3.
Patients with brain metastases, grade ≥ 2 peripheral neuropathy, uncontrolled intercurrent illness, HIV infection, or with another malignancy (except adequately treated basal cell or squamous cell skin cancer, adequately treated stage I/II cancer from which the patient is currently in complete remission, or any other cancer from which the patient has been disease-free for 5 years) were not included in the study. The protocol was approved by the institu- tional review boards of all participating institutions and all patients provided written informed consent.

Concomitant Medications
Standard androgen-deprivation therapy with orchiectomy, go- nadotropin-releasing hormone agonists or antagonists was contin- ued. Flutamide or ketoconazole had to be discontinued ≥ 28 days before to enrollment whereas bicalutamide or nilutamide must have been stopped ≥ 42 days before enrollment.
No other concurrent anticancer treatment was allowed. The use of the following within 14 days of enrollment and throughout the study was prohibited: clarithromycin, erythromycin, troleandomy- cin, rifampin, rifabutin, rifapentine, itraconazole, ketoconazole, fluconazole (> 200 mg per day), voriconazole, nefazodone, flu- voxamine, verapamil, diltiazem, grapefruit juice, or bitter orange.

Amiodarone use within 6 months of enrollment and throughout the study was prohibited.
Treatment
Ispinesib 18 mg/m2 was given as a 1-hour intravenous infu- sion administered every 21 days for a maximum of 12 cycles. Colony-stimulating factor (CSF) support was prohibited during the first cycle. In the event of febrile neutropenia, CSF use was al- lowed in accordance with American Society of Clinical Oncology guidelines.9 Treatment was discontinued because of: progression of disease, symptomatic deterioration, unacceptable toxicity, seri- ous intercurrent illness, a delay of treatment > 3 weeks, or at the patient’s request.

Dose Modifications
The following dose levels were used: starting dose (18 mg/m2), dose level –1 (14 mg/m2), and dose level –2 (11 mg/m2). Grade 4 neutropenia lasting ≥ 5 days, grade 3/4 neutropenia associated with
a fever, or grade 4 thrombocytopenia prompted a dose reduction by 1 level. Treatment was stopped for any grade 3 or greater non- hematologic toxicity and resumed upon resolution of that toxicity to grade ≤ 1 at a dose reduced by 1 level. Patients who suffered grade ≥ 3 neurotoxic events were removed from the study, whereas a grade 2 neurotoxicity prompted a delay until recovery to a grade
≤ 1 followed by treatment at a dose reduced by 1 level. A maximum of 2 dose reductions was allowed.

Assessments on Study
Pre-enrollment. Pretreatment evaluation included medical history, physical examination, weight and PS, serum PSA, complete blood count (CBC), serum bilirubin, ALT, AST, creatinine, radionuclide bone scan, computed tomography (CT) or magnetic resonance imaging (MRI) of the abdomen and pelvis, and brain CT or MRI if clinically indicated.

During Treatment. Physical examination and assessment of ad- verse events as well as a CBC, serum PSA, ALT, AST, bilirubin, and creatinine levels were obtained at the beginning of each cycle. All patients were evaluated for toxicity according to the National Cancer Institute Common Toxicity Criteria version 3.0.

Imaging. In patients with measurable disease, imaging studies were completed every 6 weeks by the same method used at baseline until progression. Bone scans were repeated every 12 weeks.

Follow-up. When off protocol treatment, disease assessment and PSA collection continued for every 6 weeks until disease progres- sion. Patients were assessed for survival for 3 years or until death, whichever occurred earlier.

Statistical Design
The primary endpoint of this trial was the proportion of patients who meet criteria for PSA response defined as a PSA decline of 50% confirmed by a second measurement ≥ 3 weeks later.8 Secondary endpoints included response rate (RR) in measurable disease, PFS, overall survival (OS), and toxicity.

Table 2 Grade  2 Treatment-Related Toxicity (N = 21)
Toxicity Maximum Grade
2 3 4 2-4
Laboratory
Neutropenia 3 2 11 16
Leukocytopenia 7 8 1 16
Anemia 5 1 0 6
Lymphopenia 1 0 1 2
Thrombocytopenia 1 0 1 2
Hyperglycemia 0 1 0 1
Hypoalbuminemia 1 0 0 1
Hyponatremia 0 1 0 1
Increased alkaline phosphatase 2 0 0 2
Clinical
Anorexia 2 0 0 2
Constipation 3 0 0 3
Cystitis 1 0 0 1
Dizziness 1 0 0 1
Dyspnea 0 1 0 1
Fatigue 7 1 0 8
Febrile neutropenia 0 1 0 1
Fever 0 1 0 1
Urinary tract infection 1 0 0 1
Hot flashes 1 0 0 1
Muscle weakness (whole body) 0 2 0 2
Bone pain 1 0 0 1
Arthralgia 0 1 0 1
Urinary retention 1 0 0 1

Abbreviations: PS = performance status; PSA = prostate-specific antigen

A 2-stage design was used for patient accrual. An RR of ≥ 20% would be of interest, whereas further testing would not be pursued if the response probability is ≤ 5%. One response among the first

20 patients was required to continue accrual to 40 patients. This design had a significance level of 4.7% and power of 92%.
Correlative Studies
Pharmacokinetics. Sparse blood samples for plasma pharmacoki- netic analysis were collected during cycle 1 for each subject. Cycle 2 pharmacokinetic samples were collected only if the dose of ispinesib was different from cycle 1. Samples were obtained immediately before and after the infusion, and at the following time points after the infusion: 30-60 minutes, 1.5-2.5 hours, 4-6 hours, 20-29 hours, and 36-48 hours. All samples were collected in potassium ethylene diamine tetraacetic acid tubes and centrifuged at 2500 × g at 5°C
for 10 minutes. Plasma was then frozen at –20°C until analysis for
ispinesib levels using a validated liquid chromatography (LC)/mass spectometry (MS) method by the GlaxoSmithKline Division of Drug Metabolism and Pharmacokinetics. The method for the de- termination of ispinesib in human plasma has been validated over the range 0.1-100 ng/mL using high-performance LC (HPLC)– MS/MS. Ispinesib was extracted from 50 μL of human plasma by protein precipitation using 75/25 acetonitrile/10 mM ammonium

formate (pH, 3) containing an isotopically labeled internal standard ([2H4]-ispinesib). Extracts were analyzed by HPLC-MS/MS using a TurboIonSpray™ interface and multiple reaction monitoring.
A population pharmacokinetic analysis was conducted using NON- MEM (non-linear mixed effects modeling software; University of California, San Francisco) on phase I ispinesib data after an 18 mg/m2 dose on an every-21-day schedule. When a final model was selected, it was validated using the posterior predictive check method. Upon validation of the final population model, it was used to simulate plasma ispinesib concentration-time profiles resulting from a dose of 18 mg/m2 which was utilized in this study. A post-hoc median profile with 90% confidence interval (CI) were simulated using the population pharmacokinetic model developed using phase I data. Observed study data were overlaid on the simulated profile to determine if the observed data exhibited the central tendency and variability of the phase I data obtained at the same dose level.

Molecular Correlates. Archival tumor tissue biopsy was inter- rogated for KSP expression by immunohistochemistry (IHC). Formalin-fixed, paraffin-embedded tissue were cut in 4-μm sections and placed onto positively charged slides. Slides were deparaffinized in xylene, and rehydrated in decreasing concentrations of ethanol: water. After antigen retrieval and blocking, sections were incubated for 30 minutes with an anti-KSP monoclonal antibody (provided by Yun-Fu Hu, PhD, GlaxoSmithKline) at a dilution of 1:5000. Immunohistochemical staining was performed using an Envision+ polymer (rabbit) and the Dako Autostainer Universal Staining System. Slides were counterstained with 50% Mayer’s hematoxylin. Human tonsillar tissue was used as positive control for hEg5 stain- ing. The hEg5-stained slides were read as positive or negative. Stain- ing intensity was scored by 2 independent pathology reviewers in a semiquantitative manner.
For analysis of cytoskeletal morphologic changes after drug treatment, peripheral blood mononuclear cells (PBMC) were isolated from whole blood drawn in CPT® tubes at the following times around dose 1 of

cycle 1: pretreatment, immediately before the end of the infusion, and then 1, 4, and 24 hours after the end of the infusion. Cytospins were made and fixed at each of the treating institutions before being shipped to the central laboratory for analysis.
Analysis consisted of simultaneous fluorescent IHC assessment of -tubulin, phosphohistone H3, and chromatin. Briefly, PBMCs were incubated with primary antibodies (eg, -tubulin, and phos- phohistone H3) for 1 hour at room temperature. After thorough washing, slides were incubated with fluorescent-labeled secondary antibodies for 1 hour at room temperature. After another round of washing, slides were incubated with propidium iodide for labeling of chromatin. Finally, the partially dried slides were covered with glass cover slips using VectaShield mounting media.
Stained slides were imaged using an Olympus IX81 inverted digital fluorescent microscope fitted with a Spot real time digital camera, and image analysis was performed using Image Pro software. Separate images were acquired for the specific target protein (eg, tubulin, and phosphohistone H3) and DNA, and the final image was obtained by merging the separate data files. With each set of slides to be analyzed, human cultured cell line controls (Hela cells), both positive (ie, with 1° antibody) and negative (ie, without 1° antibody), were included to assess the adequacy and specificity of the staining procedure.
Results
Patients
Twenty-five patients were registered to the study between May and November 2005. Three patients were ineligible because of low Hb, high bilirubin, and docetaxel received < 28 days before registra- tion. One patient never received protocol treatment. Characteristics of the 21 eligible and evaluable patients are summarized in Table 1. Notably, 19 of 21 patients had progressed during previous docetaxel therapy indicating that this group was not only extensively pre- treated but largely had treatment-resistant disease.

Treatment Delivery
The median duration of therapy delivered was 10 weeks (range, 2-26 weeks). Four patients required 1 dose reduction and none required 2 dose reductions, whereas the remaining 17 patients re- quired no dose reductions.
Response to Therapy
Prostate-specific antigen response was not observed in any of the patients. Only 7 patients remained on treatment > 12 weeks. Mea- surable disease meeting Response Evaluation Criteria in Solid Tumors criteria was present in 6 patients. No patient had a confirmed response in measurable disease. Median PFS was 2.1 months (95% CI, 1.6-2.7 months). Median OS was 14 months (95% CI, 10-15 months).
Toxicity
There were no treatment-related deaths. One patient discontin- ued therapy because of toxicity that included mouth sores, neutro- penia, and fatigue. Two patients died of prostate cancer while on therapy. The remaining 18 patients were removed from protocol therapy because of disease progression.
Hematologic toxicity, particularly involving white blood cells was frequent and not infrequently severe. The majority of patients (11 of

Figure 2 Moderate Kinesin Spindle Protein Staining Observed in a Subject in the Epithelial Cells Lining the Ductal Structures of the Tumor Compared with Normal Human Tonsillar Tissue
Prostate Cancer (× 10) Prostate Cancer (× 40)

Tonsil Control (× 10) Tonsil Control (× 40)

Figure 3 Monopolar Spindles Demonstrated Fluorescent IHC in Cultured Hela Cells Grown In Vitro in the Presence of Ispinesib
Untreated Ispinesib 10 nM × 16 Hours

Cells grown in the absence of ispinesib display normal mitotic figures. Abbreviation: IHC = immunohistochemistry

Figure 4 Monopolar Spindles Demonstrated Fluorescent IHC in PHA-Stimulated Lymphocytes Grown in the Presence of Ispinesib*
Untreated Ispinesib 10 nM × 16 Hours

*But not in nonstimulated lymphocytes from a patient treated with the drug.
Abbreviations: IHC = immunohistochemistry; PBMCs = peripheral blood mononuclear cells; PHA = phytohemaglutinin

21) experienced grade 4 neutropenia with an additional 2 patients experiencing grade 3 neutropenia (Table 2). This relatively high incidence of neutropenia may have been a reflection of age, PS, and previous treatment history. Grade 3 febrile neutropenia occurred in only 1 patient; therefore, this laboratory toxicity neither translated into frequent clinical morbidity, nor did it compromise significantly the delivery of treatment. Anemia and fatigue were also notable.

Pharmacokinetics
Using a population pharmacokinetic model derived from phase I data of the maximum tolerated dose of 18 mg/m2, a comparison of the sparsely sampled observed data in this study was made. Ispinesib plasma concentrations after an 18-mg/m2 dose were comparable with those observed in phase I studies, as shown in Figure 1. The fixed and random effects of the population pharmacokinetic model adequately describe the central tendency and variability of the observed data.
Molecular Correlates
Immunohistochemistry on Archival Specimens. Archival tissue was collected from 16 patients and analyzed for KSP expression by IHC

(data not shown). Kinesin spindle protein expression was absent to very low in most of the cases. In some of the cases, clusters of positive cells were seen in areas determined to be lymphoprolif- erative centers and not invasive tumor. In the few prostate cancer specimens where positive tumor cells were seen, the staining was diffusely cytoplasmic and not nuclear. As shown in Figure 2, moder- ate KSP staining was detected in 1 case. In this particular specimen, KSP was expressed in the epithelial cells lining the ductal structures of the tumor. Also shown in Figure 2, normal human tonsillar tissue shows areas of moderate-to-high KSP expression. At higher power, a few intensely staining cells that are undergoing mitosis can be seen in the tonsil tissue, but not in the prostate tumor.

Peripheral Blood Mononuclear Cell Studies. Serial PBMCs were col- lected from 11 patients for determination of the effects of ispinesib on cytoskeletal morphology. Eight sets of cytospins were analyzed. We did not observe any monopolar spindles in any of the PBMCs collected from patients (data not shown). However, the expected results were seen in cultured Hela cells grown in vitro in the pres- ence and absence of ispinesib (Figure 3). To further investigate the

negative results of the PBMC studies, additional experiments were performed in lymphocytes that were mitogen stimulated in the pres- ence or absence of ispinesib. As shown in Figure 4, phytohemaglu- tinin (PHA)-stimulated lymphocytes grown in the presence of the drug have the distinct morphologic appearance that is seen in the positive controls.

Discussion
Ispinesib was inactive in this study of patients with androgen-in- dependent, and largely docetaxel-resistant prostate cancer. The lack of activity cannot be attributed to inadequate delivery of treatment as the substantial majority of patients received all treatment at the starting dose and no patients required > 1 dose reduction. Further, pharmacokinetic studies indicate that drug concentrations were simi- lar to those seen in phase I studies at the same dose.
Our effort to document that the drug engaged the target using the relatively noninvasive approach of testing PBMCs was not successful. The absence of an effect on PBMCs might be explained by the fact that these cells were not dividing during the period of time when they were exposed to the drug in vivo and, therefore, might not indicate that the drug did not successfully engage its target. The ex-vivo stud- ies we carried out to confirm that the assays detect drug activity when dividing cells are examined. Future studies should consider examining dividing tissues, possibly mucosa or the BM in an effort to document the biologic activity of ispinesib in human subjects.
The absence of clinical activity could be explained by the lack of target. In archival primary tumors, we found very little KSP ex- pression by IHC. This was an unexpected finding. Kinesin spindle protein is expressed in PC-3 cells and ispinesib-induced growth ar- rest and apoptosis in this androgen-independent cell line.10 Kinesin spindle protein has been shown to be a promising cancer treatment target in a number of other cell lines. To the best of our knowl- edge, ours is the first study to examine KSP expression by IHC in human cancer specimens. Our findings have limitations. Archival tumor specimens allowed us to examine the primary tumor at some previous date and not metastases synchronously with treatment. However, the negative IHC findings are consistent with the lack of clinical activity.
Pharmacodynamic studies evaluating the effect of ispinesib on circulating lymphocytes provide further evidence of the critical im- portance of mitotic cell division for the activity of the drug. We found that lymphocytes, in circulation and in the presence of ispinesib, did not display the expected drug-induced morphologic changes unless they were mitogen stimulated to divide. By extrapolation, one can conclude that nondividing cells are unaffected by ispinesib, and the lack of antitumor activity seen in the current trial likely reflects the low mitotic index of primary prostate tumors.

Conclusion
Prostate cancer is a clinically and biologically heterogeneous dis- ease. We tested ispinesib in the most advanced stage of this disease and in patients with disease largely refractory to docetaxel. Thus, we set a high bar for this novel agent. Our results clearly suggest that this agent will not have a role in patients with such advanced prostate cancer. Further, the apparent absence of KSP staining in archival tumor specimens raises a broader question about the rel- evance of this therapeutic target in prostate cancer. Expression of KSP in human prostate cancer should be evaluated further before additional clinical trials are undertaken.

Acknowledgements
This investigation was supported in part by the following Public Health Service Cooperative Agreement grant numbers awarded by the National Cancer Institute, Department of Health and Human Services: CA38926, CA32102, CA35431, CA46441, CA45807,
CA42777, CA45377, CA20319. The correlative studies were sup- ported under an National Cancer Institute/Cancer Therapy Evalu- ation Program–sponsored Translational Research Initiative contract (#24XS146).

References
1. Tannock IF, de Wit R, Berry WR, et al. Docetaxel plus prednisone or mitoxantrone plus prednisone for advanced prostate cancer. N Engl J Med 2004; 351:1502-12.
2. Petrylak Dp, Tangen CM, Hussain MH, et al. Docetaxel and estramus- tine compared with mitoxantrone and prednisone for advanced refrac- tory prostate cancer. N Engl J Med 2004; 351:1513-20.
3. Wood KW, Cornwell WD, Jackson JR. Past and future of the mitotic spindle as an oncology target. Curr Opin Pharmacol 2001; 1:370-7.
4. Lee Y, Zhiheng J, Sakowicz R. Inhibitors of the mitotic kinesin KSP: biochemical mechanism of action. Proc Am Assoc Cancer Res 2002; 43:65.
5. Johnson RK, McCabe FL, Caulder E, et al. SB-715992, a potent and selective inhibitor of the mitotic kinesin KSP, demonstrates broad-spec- trum activity in advanced murine tumors and human tumor xenografts. Proc Am Assoc Cancer Res 2002; 43:269.
6. Gonzales P, Boehme M, Bienek A, et al. Breadth of anti-tumor activity of CK0238273 (SB-715992), a novel inhibitor of the mitotic kinesin KSP. Proc Am Assoc Cancer Res 2002; 43:269.
7. Chu QS, Holen KD, Rowinsky EK, et al. Phase I trial of novel kinesin spindle protein (KSP) inhibitor SB-715992 IV q 21 days. Proc Am Soc Clin Oncol 2004; 23(14 suppl):146 (Abstract 2078).
8. Bubley GJ, Carducci M, Dahut W, et al. Eligibility and response guide- lines for phase II clinical trials in androgen-independent prostate cancer: recommendations from the Prostate-Specific Antigen Working Group. J Clin Oncol 1999; 17:3461-7.
9. Outcomes of cancer treatment for technology assessment and cancer treatment guidelines. American Society of Clinical Oncology. J Clin Oncol 1996; 14:671-9.
10. Davis DA, Sarkar SH, Hussain M, et al. Increased therapeutic potential of an experimental anti-mitotic inhibitor SB715992 by genistein in PC- 3 human prostate cancer cell line. BMC Cancer 2006; 6:22.