A Phase I, Pharmacokinetic, and Biological Study of The

Vol. 9, 4761–4771, October 15, 2003 Clinical Cancer Research 4761

A Phase I, Pharmacokinetic, and Biological Study of the Farnesyltransferase Inhibitor Tipifarnib in Combination with Gemcitabine in Patients with Advanced Malignancies

Amita Patnaik,1 S. Gail Eckhardt, the pharmacokinetic behavior of each agent administered Elzbieta Izbicka, Anthony A. Tolcher, alone and together. The proportions of unfarnesylated and Lisa A. Hammond, Chris H. Takimoto, farnesylated HDJ2, a chaperone protein that undergoes farnesylation, were measured in peripheral blood mononuclear Garry Schwartz, Heather McCreery, cells. Andrew Goetz, Masataka Mori, Kazutoyo Terada, Results: Nineteen evaluable patients were treated with Lou Gentner, Mary-Ellen Rybak, 74 courses of tipifarnib/gemcitabine (mg/mg/m2). Myelosup- Henry Richards, Steven Zhang, and pression was the principal toxicity. Dose-limiting myelosup- Eric K. Rowinsky pression occurred in 2 of 5 patients at the 300/1000 dose level, whereas 2 of 11 evaluable patients at the 200/1000 dose Institute for Drug Development, Cancer Therapy and Research Center, San Antonio, Texas 78229 [A. P., S. G. E., E. I., A. A. T., level experienced dose-limiting toxicity. There was no evi- L. A. H., C. H. T., H. M., A. G., E. K. R.]; Brooke Army Medical dence of clinically relevant pharmacokinetic interactions Center, San Antonio, Texas 78234-6200 [G. S.]; Kumamoto between tipifarnib and gemcitabine. Inhibition of farnesyla- University School of Medicine, Kumamoto, Japan 860-8556 [M. M., tion of HDJ2, a potential surrogate for Ras and/or other K. T.]; and Johnson and Johnson Pharmaceutical Research and potentially relevant farnesylated proteins, was demon- Development, Titusville, New Jersey 08560 [L. G., M-E. R., H. R., S. Z.] strated in peripheral blood mononuclear cells at all dose levels. Partial responses were noted in patients with advanced pancreatic and nasopharyngeal carcinomas. ABSTRACT Conclusions: On the basis of the results of this study, Purpose: To assess the feasibility of administering tipi- the tipifarnib/gemcitabine dose level of 200/1000 is recom- farnib, an oral nonpeptidomimetic competitive inhibitor of mended for disease-directed studies. At this dose level, bio- farnesyltransferase, in combination with gemcitabine and logically relevant plasma concentrations of tipifarnib that recommend doses for disease-directed clinical trials. The consistently inhibit protein farnesylation in vitro are study also sought to identify drug-drug pharmacokinetic achieved and drug-induced inhibition of protein farnesyla- interactions, evaluate effects on protein farnesylation, and tion is measured in most patients. seek preliminary evidence for clinical activity. Experimental Design: Patients with advanced solid ma- INTRODUCTION lignancies were treated with tipifarnib at doses of 100, 200, Ras proteins are activated downstream of receptor tyrosine 2 and 300 mg twice daily continuously and 1000 mg/m gem- kinases and trigger a cascade of phosphorylation events through citabine i.v. on days 1, 8, and 15 every 4 weeks. To identify sequential activation of multiple effector pathways, including pharmacokinetic interactions, the treatment and plasma the mitogen-activated protein kinase, and phosphoinositide- sampling schemes were designed to permit comparisons of 3-OH kinase pathways, which are critical for cell proliferation, differentiation, and survival (1–5). After synthesis as cytosolic precursors, Ras proteins undergo a series of posttranslational modifications, the first and most important being the addition of Received 3/27/03; revised 7/7/03; accepted 7/7/03. a 15-carbon farnesyl isoprenoid group catalyzed by FTase,2 The costs of publication of this article were defrayed in part by the enabling association with the cell membrane (1, 5, 6). When Ras payment of page charges. This article must therefore be hereby marked is stimulated by receptor activation to bind GTP, it activates advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. several downstream effectors, thereby promoting cell prolifera- Some patients were treated at the Frederic C. Barter Clinical Research tion. The intrinsic GTPase activity of Ras then serves to revert Unit of the Audie Murphy Veterans Administration Hospital (San An- the molecule back to its inactive state. Mutations in the ras tonio, TX), supported in part by NIH Grant MO1-RR01346. Presented genes that are most commonly found in human malignancies in part at the 36th Annual Meeting of the American Society of Clinical Oncology, 11th Annual Meeting of National Cancer Institute–American Association for Cancer Research, Inc./European Organization for Re- search and Treatment of Cancer, and 92nd Annual Meeting of the American Association for Cancer Research, Inc. 2 The abbreviations used are: FTase, farnesyltransferase; FTI, FTase 1 To whom requests for reprints should be addressed, at Institute for inhibitor; GGTase I, geranylgeranyltransferase type I; MTD, maximum Drug Development, Cancer Therapy and Research Center, 7979 Wurz- tolerated dose; BID, twice daily; ANC, absolute neutrophil count; DLT, bach, Suite Z400, San Antonio, TX 78229. Phone: (210) 949-5069; Fax: dose-limiting toxicity; PBMC, peripheral blood mononuclear cell; (210) 692-7502; E-mail: [email protected]. HPLC, high-performance liquid chromatography.

ical and pharmacological foundation for disease-directed trials, including a Phase III study in advanced pancreas cancer (22). The objectives of this study were to: (a) characterize the principal toxicities of tipifarnib administered BID in combination with gemcitabine on a conventional dose schedule (1000 mg/m2 i.v. on days 1, 8, and 15 every 4 weeks) in patients with advanced solid malignancies; (b) determine the MTD of tipifarnib in combination with gemcitabine on this dose schedule and recommend doses for subsequent disease-directed trials; (c) describe the pharmacokinetics and pharmacodynamics of tipifarnib and gemcitabine in combination and determine whether there are major effects of each agent on the clearance of the other; (d) seek preliminary evidence of antitumor activity in Fig. 1 Structure of tipifarnib. patients with advanced malignancies; and (e) determine whether farnesylation of HDJ2, a chaperone protein and potential surrogate for Ras and/or other critical farnesylated proteins, is inhibited at clinically relevant doses of tipifarnib and gemcitabine. favor Ras in its GTP-bound active state, in which it constitu- tively enables transmission of proliferative and survival signals (1, 2, 4, 7). Abnormal expression and conformation changes in MATERIALS AND METHODS Ras, conferred by ras gene mutations, have been demonstrated Patients with histologically confirmed advanced solid ma- in ϳ30% of all human malignancies including ϳ50% of colo- lignancies refractory to standard therapy or for whom no effec- rectal, 70–90% of pancreatic, and 30% of non-small cell lung tive therapy existed were candidates for this study. Eligibility cancers (7). criteria included: (a) age of Ն18 years; (b) Eastern Cooperative FTIs were developed on the premise that FTase inhibition Oncology Group performance status Յ2 (ambulatory and capa- would prevent Ras processing and, hence, transduction of crit- ble of self-care); (c) no chemotherapy, radiotherapy, or investi- ical proliferative and survival signals (8). The coupling of Ras to gational therapy within the previous 4 weeks (6 weeks for protein tyrosine kinases also raises the possibility that malig- nitrosoureas or mitomycin C); (d) adequate hematopoietic nancies driven by overactivity upstream of Ras might be inhib- (ANC, Ն1,500/␮l; platelets, Ն100,000/␮l; hemoglobin, Ն9 ited by therapeutics against this target (1, 9). Additionally, a g/dl), hepatic (total bilirubin within the institutional normal large number of multifunctional proteins, as well as Ras, un- limits; aspartate aminotransferase and alanine aminotransferase dergo farnesylation, and several of these proteins, including Յ2.0 times the institutional upper normal limits, unless caused RhoB, Akt2, and the centromere-associated protein-E and cen- by hepatic metastases, in which case elevations of Յ5.0 times tromere-associated protein-F centromeric proteins, seem to be the upper normal limits were permitted), and renal (serum affected critically by FTIs (10–13). creatinine within institutional normal limits) functions; (e)no Tipifarnib (R115777, Zarnestra; Johnson & Johnson Phar- prior extensive myelotoxic therapy [defined as more than six maceutical Research and Development, Titusville, NJ; Fig. 1), courses of alkylating agent-containing chemotherapy (except an oral quinolone analogue of imidazole-containing heterocyclic low-dose cisplatin); more than four courses of carboplatin; two compounds, is a nonpeptidomimetic competitive inhibitor of or more courses of mitomycin C or a nitrosourea; irradiation to FTase (14, 15). Tipifarnib is highly specific for FTase and does Ն25% of hematopoietic reserves; high-dose chemotherapy, fol- not inhibit GGTase I, which can alternatively prenylate both lowed by hematopoietic stem cell rescue]; (f) no concurrent K-Ras and RhoB (6, 14). Tipifarnib inhibits the growth of a radiation therapy (except palliative radiation therapy), chemo- broad spectrum of human malignancies in vitro and in human therapy, hormonal therapy, or immunotherapy; and (g) no co- tumor xenografts that possess either wild-type or various types existing medical conditions likely to interfere with study proce- of ras mutants (6, 14–16). The feasibility of administering dures. Written informed consent was obtained according to tipifarnib has also been evaluated in patients with advanced federal and institutional guidelines. solid malignancies and leukemia, resulting in the characteriza- Dosage and Dose Escalation. The starting dose of tipi- tion of the toxicity profile and pharmacokinetic behavior of the farnib was 100 mg BID p.o. from day 2 of the first course agent (17–21). Myelosuppression has been the principal DLT. without interruption. The dose of tipifarnib was escalated in The MTD of tipifarnib, when administered as a single agent on 100-mg BID increments in successive cohorts of new patients, a continuous schedule, was found to be 300 mg BID (19). whereas the dose of gemcitabine was fixed at 1000 mg/m2 i.v. The rationale for evaluating the administration of tipifarnib on days 1, 8, and 15 every 28 days. A course was defined as 4 in combination with gemcitabine (tipifarnib/gemcitabine) in- weeks in length. Toxicity was graded according to the National cludes the established role of gemcitabine in malignancies as- Cancer Institute Common Toxicity Criteria (version 2.0). A sociated with high incidences of ras mutations (e.g., carcinomas minimum of three new patients was treated at each dose level. of the pancreas, lung, and urothelium), the distinct spectrum of If DLT was not observed in the first three patients, then the dose anticancer activity demonstrated with tipifarnib in preclinical of tipifarnib was increased to the next level. If DLT occurred in studies, and the widely disparate mechanisms of action of these any of the first three new patients in the first course, at least agents. This study was also designed to provide the toxicolog- three additional new patients were treated. If no further DLT

was encountered, dose escalation proceeded. Alternately, if and 90 min and 2, 2.5, 4.5, and 7.5 h after the end of the DLT was noted in one or more of three additional subjects, dose infusion. Daily administration of tipifarnib commenced on day escalation was to be terminated, and the MTD was defined as 2, and on day 15, pharmacokinetic sampling of both drugs was the highest dose level at which less than one-third of at least six initiated. On day 15, tipifarnib was administered at time zero new patients experienced DLT in course 1. Intrapatient dose and 30 min before the start of the gemcitabine infusion, and escalation was not permitted. The following events were con- blood sampling for tipifarnib plasma concentrations was ob- sidered dose limiting: (a) ANC Ͻ500/␮l for longer than 5 days tained before treatment and at 1, 2, 3, 5, 8, and 12 h. Plasma or associated with fever (temperature, Ն38°C) or infection gemcitabine concentrations were monitored on day 15 before requiring parenteral antibiotics; (b) platelets Ͻ50,000/␮l for treatment and 45, 60, 65, and 75 min and 1.5, 2, 2.5, 3, 5, and longer than 5 days or any platelet count Ͻ25,000/␮l; (c) any 8 h after the administration of tipifarnib. On day 22, pharma- drug-related grade 3 or grade 4 nonhematological toxicity, ex- cokinetic monitoring for tipifarnib alone was repeated at the cept alopecia, nausea, and vomiting in the absence of optimal same times as on day 15. For each sample, at least 5 ml of whole antiemetic medication, or asymptomatic elevations of hepatic blood was collected in heparinized tubes, kept on ice, and transaminases thought to be secondary to gemcitabine; and (d) centrifuged within1htoisolate plasma. Plasma was frozen missing two consecutive weekly doses of gemcitabine because immediately and stored at Ϫ20°C or less for later analysis. of unresolved toxicity. Tipifarnib concentrations were measured in thawed plasma Drug Administration. Tipifarnib, supplied by Johnson samples using a previously validated analytical method (17). & Johnson Pharmaceutical Research and Development as Briefly, 1-ml aliquots of plasma were alkalinized with 0.1 M 100-mg capsules, was administered after meals every 12 h. NaOH, extracted with heptane-isoamyl alcohol (90:10, v/v), and Gemcitabine (Gemzar; Eli Lilly and Company, Indianapolis, analyzed using reverse-phase HPLC with UV detection at 240 IN) was obtained commercially as a lyophilized powder in nm. Samples were separated using a C18 BDS-Hypersil HPLC sterile vials containing either 200 mg or 1 g of gemcitabine as column (10 cm ϫ 4.6 mm; particle size, 3 ␮; Alltech, Deerfield, the hydrochloride salt. Gemcitabine was stored at room temper- IL). The mobile phase consisted of 0.01 M ammonium acetate- ature and reconstituted with normal saline to a concentration of acetonitrile (52:48, v/v). The lower limit of quantitation for 40 mg/ml, which was further diluted to a final volume of 250 ml tipifarnib was 2 ng/ml. The chromatographic peaks were quan- before administration over 30 min. tified using UV detection at 240 nm. The mean overall coeffi- Pretreatment Assessment and Follow-Up Studies. A cient of variation was 6.7% at 14.9 ng/ml, 7.1% at 124 ng/ml, history, physical examination, and routine laboratory studies and 7.1% at 2,064 ng/ml. were performed before treatment and at least weekly during Gemcitabine plasma concentrations were analyzed using a treatment. Routine laboratory studies included serum electro- HPLC-based assay (23). Briefly, acetic acid (2 ml, 1 M) was lytes, chemistries, renal and liver function tests, complete blood added to 0.2-ml aliquots of patient plasma and mixed at 4°C. cell and differential WBC counts, coagulation studies, and uri- The aliquot was spiked with 200 ng of internal standard (2Ј,3Ј- nalysis. Ophthalmological evaluations, which included an oph- dideoxycytidine) contained in 100 ␮l of methanol at room thalmical history and slit lamp biomicroscopy, were performed temperature. Thereafter, another 100 ␮l of methanol and 1 ml of before treatment and then before each subsequent course. 1 M acetic acid were added to the aliquot. The mixture was PBMCs were collected before treatment and weekly during the poured very slowly over a solid-phase extraction column (Oasis first 8 weeks of treatment. Relevant radiological studies to MCX solid phase extraction column, 3 ml/60 mg; Waters). The assess measurable and evaluable disease were repeated after solid-phase extraction column was then washed with 3 ml of every other course. Patients were able to continue treatment if Milli-Q water, 1 ml of 1 M acetic acid, and 3 ml of methanol. they did not develop progressive disease, which was defined as Finally, gemcitabine was eluted with 3 ml of methanol-NH4OH an increase in the sum of the bidimensional product of measur- 25% (98:2, v/v). The aliquot was dried under nitrogen at 65°C. able disease by at least 25% or the appearance of any new The extraction residue was redissolved in 200 ␮l of HPLC lesion. A complete response was scored if there was disappear- solvent mixture containing 0.01 M ammonium acetate (pH 9 ance of all active disease on two measurements separated by a with 25% NH4OH) and methanol (90:10, v/v). The extractant minimum period of 4 weeks, whereas a partial response required (10 ␮l) was injected into a HPLC system that included a at least a 50% reduction in the sum of the product of the reverse-phase HPLC column (10 cm ϫ 4.6 mm inner diameter) bidimensional measurements of all index lesions, separated by packed with 3 ␮m of C18 BDS-Hypersil (Alltech). The HPLC at least four weeks. mobile phase initially consisted of 0.01 M ammonium acetate

Pharmacokinetic Sampling and Analytical Methodol- (pH 9 with 25% NH4OH) and methanol (93.5:6.5, v/v), which ogy. To study relevant pharmacokinetic parameters reflecting was pumped isocratically at 0.8 ml/min for 5.6 min. The per- both gemcitabine and tipifarnib exposure, blood was sampled centage of methanol was then increased to 70% from 6.5% at during the first course of treatment. The purpose of the sampling 5.6 min and finally reduced to 6.5% at 8.6 min after starting the schedule was to ascertain the pharmacokinetic information of HPLC run. The mobile phase, consisting of 0.01 M ammonium each drug alone and in combination and specifically to deter- acetate and methanol (93.5:6.5, v/v), was held constant for 12 mine whether there were any major effects of each drug on the min. The overall HPLC run time was 12 min. With UV detec- clearance of the other. Whole blood samples were collected for tion at 275 nm, the retention times of gemcitabine and 2Ј,3Ј- analysis of gemcitabine plasma concentrations on day 1 at the dideoxycytidine, the internal standard, were 4.2 and 5.6 min, following times: before treatment, 15 min, and at 30 min (im- respectively. The concentration of gemcitabine in patient plasma mediately before the end of the infusion), and at 5, 15, 30, 60, samples was calculated by determining the ratio of the area of

peak gemcitabine in each sample to that of the internal standard Table 1 Patient characteristics in each sample and comparing this ratio to a calibration curve Characteristic No. of patients obtained from known samples in the same HPLC running batch. ␮ No. of patients (evaluable) 22 (19) The lower limit of quantitation for gemcitabine was 100 g/ml. Median no. of courses/patient (range) 3 (1–11) Farnesylation Assay. To assess the serial effects of tipi- Median age (range) 59 (22–80) farnib on the proportions of farnesylated and unfarnesylated Gender (male:female) 14:8 a forms of the chaperone protein HDJ2, which may reflect the Median performance status (ECOG ) 05 effect of drug on protein farnesylation, PBMCs were obtained 115 from approximately 8 ml of whole blood collected in sodium 22 citrate vacutainer CPT tubes (Becton-Dickinson, Franklin Previous treatment Lakes, NJ) and processed within 10 min of collection. The Chemotherapy 6 Ϫ Radiation 1 PBMCs were stored at 70°C. The cells were homogenized on Chemotherapy and radiation 5 ice in 10 mM Tris (pH 7.4) and 0.3% SDS with 1% protease None 10 inhibitors (Sigma Chemical Co., St. Louis, MO) and centrifuged Primary tumor sites at 12,000 ϫ g at 4°C. The supernatants were treated with 50 Pancreas 8 ␮ ␮ Lung (non-small cell) 4 g/ml DNase and 100 g/ml RNase. The protein concentrations Colorectal 3 assessed by 260/280-nm absorbance ranged from 1–5 mg/ml Ampullary, lung (small cell), nasopharynx, 1 each (24). An equal amount of protein (20 ␮g) was boiled in SDS adrenocortical, ovary, unknown primary loading buffer containing 0.1 M DTT and electrophoresed in a ECOG, Eastern Cooperative Oncology Group. 7.5% SDS polyacrylamide gels (Bio-Rad Laboratories, Hercu- les, CA). Recombinant human HDJ2, generated by Drs. Ma- sataka Mori and Kazutoyo Terada (Kumamoto University, Pharmacokinetic and Pharmacodynamic Analysis. Kumamoto, Japan) or purchased from Neomarkers (Fremont, Gemcitabine and tipifarnib pharmacokinetics were examined CA), were used as positive controls. The gels were transblotted using noncompartmental analytical methods with WinNonLin to Hybond-C nitrocellulose membranes (Amersham Bio- software (version 3.1; Pharsight Corp., Cary, NC). The sciences, Piscataway, NJ), and the membranes were blocked in AUC0–12 h was calculated using the linear trapezoidal method 0.1% Tween 20 and 10% SuperBlock (Pierce Chemical Co., ␭ (26). The terminal rate constant ( z), was determined from the Rockford, IL). After incubation with either rabbit polyclonal slope of the terminal log-linear portion of the plasma concen- anti-HDJ2 (generated by Drs. Mori and Terada) or monoclonal tration-time curve, and the terminal half-life was calculated as mouse anti-HDJ2 clone KA2A5.6 (Neomarkers), then goat an- ␭ ln (2)/( z). Maximal plasma concentrations (Cmax) were deter- timouse (IgM) peroxidase conjugate (Calbiochem-Novabio- mined by direct observation of the data. chem Corporation, San Diego, CA) or goat antirabbit IgG (Ab- The relationships between indices of tipifarnib and gem- cam, Cambridge, United Kingdom), HDJ2 was visualized by citabine exposure (Cmax and AUC) and myelosuppression in the enhanced chemiluminescence with SuperSignal West Pico rea- first course of therapy were explored. In these pharmacody- gent (Pierce Chemical Co.). Image analysis was performed with namic analyses, myelosuppression was quantified as the per- a personal densitometer (Amersham Biosciences). Relative centage decrements in the ANC and platelet counts using the quantities of farnesylated and unfarnesylated HDJ2, the latter following formula: detected as a slower migrating band, were determined using ImageQuant (Amersham Biosciences), and a percentage of un- Percentage decrement prenylated HDJ2 was calculated (25). Percentage (absolute) Pretreatment blood cell counts change in unfarnesylated HDJ2 was calculated by subtracting Ϫ Nadir blood cell counts pretreatment values obtained on day 1 from day 8 or day 15 ϭ ϫ 100% Pretreatment blood cell counts values. Comparisons were made between the day 1 and day 8 values where possible; however, if the day 8 sample was not Pharmacodynamic data were analyzed using regression available or could not be analyzed, then the day 15 value was methods to fit the data to linear, maximum effect, and sigmoidal

used. Given that tipifarnib concentrations were expected to have maximum effect (Emax) models using WinNonLin software achieved steady state by day 8, it was hypothesized that both (version 3.1; Pharsight Corp.). Model selection and goodness- days 8 and 15 would provide a reasonable index of the farne- of-fit were assessed by examining the coefficients of determi- sylation status of HDJ2 after exposure to therapeutic levels of nation (r2), graphical plots of the observed versus predicted tipifarnib. Descriptive statistics were used to obtain the median data, and the SEs of the estimated parameters. percentage change in the proportion of unfarnesylated HDJ2 at Statistical Analysis. The tipifarnib pharmacokinetic pa- the 100-, 200-, and 300-mg BID dose levels of tipifarnib. The rameters on day 15 (in the presence of gemcitabine) and day 22 percentage of unfarnesylated HDJ2 before treatment was com- (in the absence of gemcitabine) were compared using an pared with values on day 8 or day 15 at each dose level using a ANOVA test. Similarly, an ANOVA test was also used to paired t test. The relationships between the percentage change in compare the gemcitabine pharmacokinetic parameters on day 1 unfarnesylated HDJ2 protein and tipifarnib area under the con- (in the absence tipifarnib) and day 15 (in the presence of

centration-time curve from zero to 12 h (AUC0–12 h) on days 15 tipifarnib). Thus, the clearance of each agent was assessed alone and 22 were also explored. and together. A significance level of 0.05 was used for all

Table 2 Dose escalation scheme No. of patients with No. of patients DLT/evaluable patients Tipifarnib dose New Reduced to No. of courses First All (mg BID) (evaluable) this dose Total (evaluable) course courses 100 4 (3)a 0 4 16 (15) 0/3 0/3 200 13 (11)b 1 14 50 (48) 2/11 2/11 300 5 0 5 11 (11) 2/5 2/5 a One patient was considered inevaluable because of a history of extensive prior myelotoxic therapy. b Two patients requested removal from the study after receiving only one infusion of gemcitabine and were considered inevaluable.

analyses. The results obtained from the ANOVA were used to being dose limiting at 200 mg of tipifarnib BID. The dosage of generate 90% confidence intervals about the geometric mean tipifarnib was required to be reduced for prolonged grade 4 ratios of the pharmacokinetic parameters with (test) and without neutropenia in only one patient (300/1000 dose level). The first (reference) interacting drugs. All statistical analyses were per- dose level (100/1000) was associated with brief grade 3 neutro- formed using the SAS statistical software program (version penia in 1 (1%) of 15 evaluable courses. A total of 48 courses 6.21; SAS Institute, Inc., Cary, NC). were administered at the second dose level (200/1000), of which seven (9%) were associated with grade 3 neutropenia, two (3%) RESULTS with brief grade 4 neutropenia, one (1%) with grade 4 neutro- General. Twenty-two patients, whose pertinent charac- penia with fever, and three (4%) with grade 3 thrombocytope- teristics are summarized in Table 1, were treated with 77 total nia. Unacceptable grade 4 neutropenia lasting more than 5 days courses of tipifarnib/gemcitabine, with 19 patients and 74 was observed in 2 (3%) of 11 courses at the 300/1000 dose level, courses fully evaluable for toxicity assessment. Twelve patients and, in one of these, grade 3 thrombocytopenia lasting 13 days had previously received chemotherapy and/or radiation therapy. was also observed. The total numbers of new patients and fully evaluable courses, Nonhematological Toxicity. Nonhematological toxici- at each tipifarnib dose level, as well as the rates of DLT as a ties were mild to moderate in severity in most patients. Grade 1 function of dose level, are detailed in Table 2. and/or grade 2 nausea or vomiting occurred in 28 (38%) of 74 In summary, tipifarnib dose escalation proceeded in the and 16 (22%) of 74 courses, respectively, and was typically following manner. Three new patients tolerated tipifarnib/gem- managed with oral phenothiazines. Fatigue of grade 1 or grade citabine (mg/mg/m2) at the first dose level of 100/1000 without 2 severity was observed in 37 (50%) of 74 of courses, whereas DLT, and, thus, dose level escalation proceeded to 200/1000, isolated grade 3 fatigue was observed in a single patient during which was also well tolerated in the first three patients. How- a first course at the 200/1000 dose level. The latter event, which ever, the next higher dose level, 300/1000, resulted in an unac- was brief and experienced after the first gemcitabine treatment, ceptably high rate of hematological DLT. Therefore, additional did not recur with subsequent treatments. Given its transient, patients were treated at the preceding dose level of 200/1000, nonreproducible nature and the lack of distinct association with and this resulted in dose-limiting hematological toxicity in 2 of the tipifarnib/gemcitabine regimen, the event was not classified 11 fully evaluable new patients. Therefore, the recommended as dose limiting. Mild to moderate anorexia occurred in 11 Phase II dose level was 200 mg of tipifarnib BID in combination (15%) of 74 courses, whereas grade 3 anorexia was observed in with 1000 mg/m2 gemcitabine i.v. on days 1, 8, and 15 every 28 a single patient at the 300/1000 dose level in association with days. When reviewing all dose levels, 9 of 19 evaluable patients dose-limiting hematological toxicity. Mild to moderate diarrhea, required omission and/or reduction in the dose of gemcitabine at fever, and skin rash were observed in Ͻ20% of courses. One some time during treatment. Gemcitabine dose reductions oc- instance of grade 3 rash was observed at the 200/1000 dose level curred in 19% of all courses (15% because of hematological in association with dose-limiting hematological toxicity. toxicity and 4% because of nonhematologic toxicity), with 13% Asymptomatic grade 1 or grade 2 hepatic transaminase eleva- and 6% occurring at the second and third dose levels, respec- tions were observed in 16 (22%) of 74 courses and were thought tively. to be caused by gemcitabine. A single patient at the second dose Hematological Toxicity. Myelosuppression, particularly level developed a moderate (grade 2) sensory neuropathy after neutropenia, was the principal toxicity of the tipifarnib/gemcit- six courses of treatment. The pertinent past medical history of abine combination. The median, range, and distribution accord- this patient, a 52-year-old male with previously untreated met- ing to National Cancer Institute grade of the nadir ANC and astatic pancreatic cancer, included extensive tobacco use, hy- platelets as functions of both dose level and the extent of prior perlipidemia, and noninsulin-dependent diabetes mellitus. Tipi- therapy are shown in Table 3. The ANC nadir during the first farnib treatment was omitted during the eighth course, but the course typically occurred between days 15 and 22, with neutro- patient continued to receive gemcitabine. The neuropathy im- penia resolving within 7 days. Neutropenia of grade 3 or grade proved to grade 1 level, and, thus, tipifarnib was restarted at a 4 severity occurred in 13 (18%) of 74 courses and was dose dose of 200 mg BID for the ninth course. The recurrence of limiting in 3 (4%) courses. Thrombocytopenia of grade 3 sever- grade 2 sensory neuropathic symptoms ultimately resulted in ity was observed in 4 (5%) of 74 courses, with one episode dose reduction of tipifarnib to 150 mg BID for two additional

Downloaded from clincancerres.aacrjournals.org on September 26, 2021. © 2003 American Association for Cancer Research. 4766 Phase I Trial of Tipifarnib and Gemcitabine b 2/5 2/11 0/3 DLT/total patients No. of patients with heme Ͼ 5 days) Platelets (/ ␮ l)

( Ͻ 50,000; Fig. 2 Mean (ϮSD) steady-state plasma tipifarnib concentration-time profiles after treatment with 200 mg of tipifarnib BID with (day 15) and

4 without (day 22) concomitant administration of gemcitabine (1000 2 ϭ Grade mg/m ; days 1, 8, 15, every 28 days; n 8). 3 Grade

a courses, in which partial resolution of symptoms to grade 1 severity was observed. Antineoplastic Activity. Two partial responses were observed. A 73-year-old male with previously untreated metastatic pancreatic cancer (100/1000 dose level) experienced a partial 151 (23 – 479) 0 (0) 0 (0) 1 (1) 181 (18 – 689) 3 (3) 0 (0) 0 (0) 191 (68 – 423) 0 (0) 0 (0) 0 (0) Median nadir response lasting 6 months. The second partial response (300/

platelets (range; / ␮ l) 1000 dose level), which lasted for 5 months, was observed in a 34-year-old patient with metastatic nasopharyngeal carcinoma who had received prior chemotherapy and radiation. Further- Hematologic toxicity more, two previously untreated patients with metastatic pancre-

fever atic cancer experienced 2.5- to 15-fold reductions in serum CA 19-9, as well as improvement in pain and performance status Table 3 Grade 3 – 4 and for 7–9 months. Pharmacokinetics. Plasma samples for pharmacokinetic

No. of courses (first courses) with toxicity studies were obtained from 15 and 12 patients for tipifarnib and

Grade 4 gemcitabine, respectively. Fig. 2 shows the mean steady-state ( Ͼ 5 days) plasma tipifarnib concentration-time profiles at the 200-mg BID dose level in the presence and absence of gemcitabine, Fig. 3 4 shows paired tipifarnib AUC values for individual patients Grade 0–12 h

Neutropeniain the Thrombocytopenia presence and absence of gemcitabine, and tipifarnib pharmacokinetic parameters in the presence and absence of 3

Grade gemcitabine are shown in Table 4. There were no significant differences in pharmacokinetic parameters reflecting tipifarnib exposure when the agent was administered alone or in the a presence of gemcitabine. As shown in Table 5, which displays plasma tipifarnib and gemcitabine pharmacokinetic parameters for each drug administered alone and in combination, there was

(range; / ␮ l) no major effect of gemcitabine on the clearance of tipifarnib

Median nadir ANC across all three tipifarnib dose levels. Mean plasma gemcitabine concentration-time profiles after treatment with gemcitabine in the presence and absence of

No. of tipifarnib and paired gemcitabine AUC (area under concen- courses last evaluable tration-time curve from time zero until the last measured time point) values for individual patients in the presence and absence

Median values (ranges) forHeme, all hematological. courses. of tipifarnib are shown in Figs. 4 and 5, respectively. As shown a b 300 11 2200 (7 – 22,649) 0 (0) 0 (0) 2 (2) 0 (0) 200 48 3082 (45 – 7,690) 7 (0) 2 (1) 0 (0) 1 (1) 100 15 3090 (765 – 10,272) 1 (0) 0 (0) 0 (0)in these figures, 0 (0) there were no significant effects of tipifarnib on (mg BID) Tipifarnib dose level mean plasma gemcitabine concentrations and AUClast, which

Fig. 3 Paired tipifarnib AUC0–12 h values for individual patients in the presence (day 15) and absence (day 22) of gemcitabine.

Table 4 Summary of mean tipifarnib plasma pharmacokinetic parameters in the presence and absence of gemcitabine Pharmacokinetic variable Units Tipifarnib alonea (day 22) Tipifarnib and gemcitabinea (day 15) Tipifarnib (100 mg BID; n ϭ 4) b ⅐ AUC0–12h ng h/ml 1985 (832) 2452 (648) Cmax ng/ml 409 (222) 493 (213) Tipifarnib (200 mg BID; n ϭ 8) ⅐ AUC0–12h ng h/ml 2552 (1294) 2171 (1229) Cmax ng/ml 525 (260) 468 (355) Tipifarnib (300 mg BID; n ϭ 3) ⅐ AUC0–12h ng h/ml 2833 (1387) 3369 (1911) Cmax ng/ml 536 (260) 805 (490) a Values represent mean (SD). b AUC0–12h, area under the time-concentration curve from zero to 12 h; Cmax, maximum plasma concentration.

Table 5 Comparison of tipifarnib and gemcitabine pharmacokinetic parameters for each agent administered alone and in combination Variable Units Pa Ratiob 90% confidence intervalc Tipifarnib pharmacokinetic parameters (100–300 mg BID, n ϭ 15) with or without gemcitabine (1000 mg/m2) d ⅐ AUC0–12h ng h/ml 0.932 99.3 (86.5; 114.0) Cmax ng/ml 0.999 100.0 (79.4; 125.9) Gemcitabine pharmacokinetic parameters (1000 mg/m2, n ϭ 12) with or without tipifarnib (100–300 mg BID) ⅐ AUClast ng h/ml 0.989 99.7 (67.7; 146.8) Cmax ng/ml 0.640 109.2 (78.6; 151) T1/2 min 0.201 73.5 (38.5; 108.5) a P (two-sided) in log scale for testing a zero difference between two treatments. b Based on least square means; calculated as the ratio of test (tipifarnib and gemcitabine) to reference (tipifarnib or gemcitabine alone). c Confidence limits in log scale are expressed in percentages of tipifarnib alone. d AUC0–12h, area under the time-concentration curve from zero to 12 hours; Cmax, maximum plasma concentration; AUClast, area under the time-concentration curve from zero to the last measurable time point; T1/2, elimination half-life. reflect drug exposure. This is further illustrated in Tables 5 and platelet counts on day 15 (presence of gemcitabine) and day 22

6, which compare pharmacokinetic parameters of gemcitabine (absence of gemcitabine) using linear and Emax models. Simi- in the presence and absence of tipifarnib and fail to demonstrate larly, the relationship between tipifarnib AUC0–12 h and the major differences across the dose levels of tipifarnib evaluated percentage decrements in ANC on days 15 and 22 were de- in this study (100–300 mg BID). The results indicate that there scribed by neither linear nor Emax models. No pharmacody- are no major interactions of gemcitabine on tipifarnib clearance namic relationships between gemcitabine exposure and the per- and exposure, and visa versa. centage decrement in platelet or ANC were evident. Pharmacodynamics. Pharmacodynamic relationships Farnesylation Studies. Blood was serially sampled in 16 between hematological effects and pertinent pharmacokinetic patients to assess the effects of the tipifarnib/gemcitabine regi- parameters reflecting both tipifarnib and gemcitabine exposure men on farnesylation of the chaperone protein HDJ2. In course were explored. No significant relationship was demonstrated 1, the proportion of unfarnesylated HDJ2 increased in 12 (75%) between tipifarnib AUC0–12 h and the percentage decrements in of 16 patients. The median percentage changes in the proportion

ble from both toxicological and pharmacological perspectives could be further evaluated in pivotal first-line clinical trials. Indeed, the results of the present study indicate that the com- bined administration of biologically and clinically relevant doses of both tipifarnib and gemcitabine is feasible. The principal toxicities of the tipifarnib/gemcitabine combination were neutropenia, with neutropenia of grade 3–4 severity occurring in 18% of all courses. The incidence of grade 3 thrombocytopenia was even lower, occurring in 5% of all courses, whereas grade 4 thrombocytopenia was not observed. At the highest dose level of tipifarnib/gemcitabine (300/1000) evaluated in the study, the incidence of intolerable myelosuppression was unac- ceptably high, occurring in two of five patients in first courses. In contrast, the next lower tipifarnib/gemcitabine dose level (200/1000) was better tolerated. At this dose level, which was considered the MTD level and recommended for subsequent Fig. 4 Mean (ϮSD) plasma gemcitabine concentration-time profiles after treatment with 1000 mg/m2 gemcitabine i.v. with (day 15) and disease-directed studies, dose-limiting hematological events without (day 1) concomitant administration of tipifarnib (200 mg BID; were much less frequent, occurring in 2 of 11 (11%) evaluable n ϭ 7). patients. Furthermore, few patients treated at this dose level required either dose modification, reduction, or omission, even after cumulative, long-term treatment. In fact, gemcitabine doses were reduced at some time during treatment because of ϭ of unfarnesylated HDJ2 were 7% (range, 0–14; n 2), 25% insufficient blood cell counts on the day of treatment in only 11 Ϫ ϭ Ϫ ϭ (range, 8 to 43.6; n 9), and 27% (range, 15.9 to 62.8; n (15%) of 74 total courses in the study. In addition, both first- 5) at the 100-, 200-, and 300-mg BID dose levels, respectively. course and cumulative nonhematological toxicities were gener- Paired sets of individual patient HDJ2 farnesylation data before ally mild to moderate at the MTD level. Although a single indi- and after treatment with tipifarnib/gemcitabine are shown in Fig. vidual developed a grade 2 peripheral neuropathy, which has been 6. A significant difference in the percentage of unfarnesylated reported in other clinical evaluations of tipifarnib administered as protein was demonstrated during the first course, with a median both a single agent and in various combination regimens, manifes- ϭ increase of 23% (P 0.001). tations of neurotoxicity were of maximal severity after repetitive drug administration and the patient had several major risk factors DISCUSSION for developing peripheral neuropathy (17, 19, 20). Ras and other small G proteins acquire functional biolog- Because the pharmacokinetic behaviors of both tipifarnib ical activity only after posttranslational modifications facilitate and gemcitabine have been extensively characterized in previ- anchorage to cellular membranes. The discovery that the inhi- ous clinical investigations, the principal objective of the phar- bition of Ras farnesylation blocks Ras function was the primary macokinetic assessments in the present trials was to determine impetus for developing the FTIs (8). In contrast to the more the propensity for major pharmacokinetic interactions with a conventional nonspecific chemotherapeutic agents currently in our principal focus on drug clearance. Congruent with the lack of therapeutic armamentarium, FTIs induce antitumor activity in vitro major toxicological interactions, which is supported by the and in vivo without producing appreciable nonspecific toxicity in feasibility of administering both agents together at full single- animals and in early clinical investigations (3, 6, 9). As would be agent doses, there was no evidence for major pharmacological expected based on the multifunctionality of critical enzymes, the interactions between tipifarnib and gemcitabine. Although the preclinical activity of FTIs is not limited to cells harboring mutated results of the present study do not refute the possibility of ras, and prominent activity in early clinical trials in breast, central pharmacokinetic interactions of a lesser magnitude, the exist- nervous system, and other malignancies that have low incidences ence of functionally relevant effects of tipifarnib on the clear- of ras mutations suggest that the FTIs may broadly affect the ance of gemcitabine, and visa versa, is improbable. Further- farnesylation of many proteins in addition to Ras. more, the range of tipifarnib plasma concentrations at the Given the current ambiguity surrounding the mechanism recommended dose level of 200 mg of tipifarnib BID in com- by which the FTIs exert antitumor activity, this Phase I, phar- bination with 1000 mg/m2 gemcitabine on days 1, 8, and 15 macological, and biological study was designed to determine the every 28 days (minimum plasma concentration, 14–305 ng/ml) MTDs and overall feasibility of administering the nonpeptido- is similar to that noted in single-agent studies and exceeds mimetic tipifarnib in combination with gemcitabine, as well as tipifarnib concentrations capable of inhibiting farnesylation of to identify major pharmacokinetic interactions between the Ras and other critical proteins, and, importantly, tumor growth agents and evidence of inhibition of HDJ2 chaperone protein in preclinical studies (6, 14). In preclinical studies with tipi- farnesylation. The rationale for evaluating this particular regi- farnib, cell lines bearing H-ras or N-ras mutations demonstrated

men was largely based on the high incidence of ras mutations in IC50 for inhibition of proliferation below 5 ng/ml (14). carcinomas of the pancreas and lung and other malignancies in The effects of the combination of tipifarnib and gemcitab- which gemcitabine is a mainstay of treatment, particularly be- ine on the farnesylation of HDJ2, a chaperone protein that cause the role of a tipifarnib/gemcitabine regimen that is feasi- undergoes farnesylation, was also assessed in the present study.

Fig. 5 Paired gemcitabine AUClast values for individual patients in the absence (day 1) and presence (day 15) of tipifarnib at the 100-, 200-, and 300-mg BID dose levels (P ϭ 0.56, paired t test for all paired data sets).

Table 6 Summary of mean gemcitabine pharmacokinetic parameters in the presence and absence of tipifarnib Variable Units Gemcitabine alone (day 1)a Tipifarnib and gemcitabine (day 15)a Tipifarnib (100 mg BID; n ϭ 3) ⅐ AUClast ng h/ml 5,200 (3224) 5,825 (2142) Cmax ng/ml 10,748 (7710) 13,025 (3638) T1/2 terminal min 22 (11.0) 14 (1.0) Tipifarnib (200 mg BID; n ϭ 7) ⅐ AUClast ng h/ml 13,237 (13512) 8,403 (2547) Cmax ng/ml 19,777 (9459) 17,574 (18215) T1/2 terminal min 22 (17) 15 (3) Tipifarnib (300 mg BID; n ϭ 2) ⅐ AUClast ng h/ml 9,890 (3814) 14,032 (5340) Cmax ng/ml 22,476 (8209) 34,371 (22330) T1/2 terminal min 16 (3) 19 (2.0)

Abbreviations: Cmax, maximum plasma concentration; AUClast, area under the time-concentration curve from zero until last plasma sample; T1/2, elimination half-life. a Values represent mean values (standard deviation).

Fig. 6 Paired values for percentage unfarnesylated HDJ2 protein in individual patients during the first course for tipifarnib at dose levels of 100, 200, and 300 mg BID (P ϭ 0.001, paired t test for all paired data sets).

The proportion of unfarnesylated HDJ2 protein increased in 12 concordance between the farnesylation status of the HDJ2 chap- of the 16 patients who had serial sampling performed; however, erone protein, Ras, and/or other farnesylated proteins that are these results and similar studies assessing the farnesylation linked to clinical end points is not known. Therefore, serial status of other readily quantifiable proteins, such as lamin A, evaluations of HDJ2 farnesylation may be used as general must be interpreted with caution for several reasons. First, pharmacodynamic indices of protein farnesylation, however, although the inhibition of HDJ2 farnesylation may indicate that concordant studies relating the farnesylation of HDJ2 to that of tipifarnib is truly affecting its intended target, FTase, which is in proteins linked to tumor proliferation are necessary to determine line with its purported biochemical mechanism, the degree of the overall use and significance of such assessments. Interest-

ingly, the results of preclinical and clinical studies of the effects REFERENCES of L-778,123, an inhibitor of FTase, as well as GGTase I, to a 1. Rowinsky, E. K., Windle, J. J., and Von Hoff, D. D. Ras protein lesser degree, on various prenylated proteins clearly demon- farnesyltransferase: a strategic target for anticancer therapeutic devel- strated that the agent was highly capable of inhibiting HDJ2 opment. J. Clin. Oncol., 17: 3631–3652, 1999. farnesylation, whereas the prenylation of K-Ras was not af- 2. Patnaik, A., and Rowinsky, E. K. Farnesyltransferase and Gera- nylgeranyltransferase I: Targets for Cancer and Cardiovascular Therapy. fected (27). Although these results are likely explained by the Towtowa, NJ: Humana Press Inc., 2001. alternate prenylation of N- and K-Ras by residual GGTase I, 3. Feldkamp, M. M., Lau, N., and Guha, A. Growth inhibition of they illustrate the lack of correlation between the inhibition of astrocytoma cells by farnesyl transferase inhibitors is mediated by a FTase activity and Ras functionality. Another factor that may combination of anti-proliferative, pro-apoptotic and anti-angiogenic ef- potentially confound the interpretation of the HDJ2 farnesyla- fects. Oncogene, 18: 7514–7526, 1999. tion data in this and other clinical trials, to date, is the lack of 4. Gibbs, J. B., and Oliff, A. The potential of farnesyltransferase inhibitors as cancer chemotherapeutics. Annu. Rev. Pharmacol. Toxicol., 37: robust information relating drug effects on farnesylation in 143–166, 1997. PBMCs to tumor tissues in humans. A final issue pertaining to 5. End, D. W. Farnesyl protein transferase inhibitors and other therapies assessing the farnesylation status of HDJ2 and other proteins targeting the Ras signal transduction pathway. Invest. New Drugs, 17: that might suffice as surrogates of relevant target effects is not 241–258, 1999. knowing the precise mechanism by which FTIs inhibit tumor 6. End, D. W., Smets, G., Todd, A. V., Applegate, T. L., Fuery, C. J., proliferation and/or induce cytotoxicity. This is shown by the Angibaud, P., Venet, M., Sanz, G., Poignet, H., Skrzat, S., Devine, A., Wouters, W., and Bowden, C. Characterization of the antitumor effects responsiveness of cancers with both wild-type ras and K-ras of the selective farnesyl protein transferase inhibitor tipifarnib in vivo mutations in preclinical studies and clear tumor regressions in and in vitro. Cancer Res., 61: 131–137, 2001. patients with advanced malignancies with an inherently low 7. Bos, J. L. ras oncogenes in human cancer. A review. Cancer Res., incidence of ras mutations including breast carcinoma, high- 49: 4682–4689, 1989. grade astrocytoma, and several types of hematopoietic malig- 8. Sebti, S. M., and Hamilton, A. D. New approaches to anticancer drug nancies (20, 21, 28, 29). Although the activity of the FTIs in design based on the inhibition of farnesyltransferase. Drug Discov Today, 3: 26–33, 1998. these situations may be accounted for by inhibition of Ras 9. Martinez-Lacaci, I., Kannan, S., De Santis, M., Bianco, C., Kim, N., farnesylatin and hyperactive signaling activity upstream of Ras, Wallace-Jones, B., Ebert, A. D., Wechselberger, C., and Salomon, D. S. there is a large body of preclinical evidence implicating RhoB, RAS transformation causes sustained activation of epidermal growth Akt2, centromere-associated protein-E, centromere-associated factor receptor and elevation of mitogen-activated protein kinase in human mammary epithelial cells. Int. J. Cancer, 88: 44–52, 2000. protein-F, and other farnesylated proteins as the principal target 10. Liu, A., Cerniglia, G. J., Bernhard, E. J., and Prendergast, G. C. of the FTIs (10–13). The possibility that the principal target of RhoB is required to mediate apoptosis in neoplastically transformed the FTIs is tumor dependent presents additional challenges to cells after DNA damage. Proc. Natl. Acad. Sci. USA, 98: 6192–6197, identifying and developing feasible and reliable assays that 2001. reflect the effects of the FTIs on tumor growth. 11. Cox, A. D. Farnesyltransferase inhibitors: potential role in the treatment of cancer. Drugs, 61: 723–732, 2001. The present study demonstrated the administration of the 2 12. Ashar, H. R., James, L., Gray, K., Carr, D., Black, S., Armstrong, combination of 200 mg of tipifarnib BID plus 1000 mg/m L., Bishop, W. R., and Kirschmeier, P. Farnesyl transferase inhibitors gemcitabine i.v. weekly for 3 weeks every 28 days is feasible, block the farnesylation of CENP-E and CENP-F and alter the associa- and these doses are recommended for further disease-directed tion of CENP-E with the microtubules. J. Biol. Chem., 275: 30451– studies of the regimen. At these doses, major drug-drug inter- 30457, 2000. actions were not evident, and farnesylation of HDJ2 in PBMCs 13. Crespo, N. C., Ohkanda, J., Yen, T. J., Hamilton, A. D., and Sebti, S. M. The farnesyltransferase inhibitor, FTI-2153, blocks bipolar spin- was inhibited in most patients in whom serial assessments were dle formation and chromosome alignment and causes prometaphase performed. Nevertheless, treatment with the tipifarnib/gemcit- accumulation during mitosis of human lung cancer cells. J. Biol. Chem., abine regimen at these doses should be limited to patients with 276: 16161–16167, 2001. good performance status, organ function, and a history of no 14. Janssen Research Foundation. Clinical Investigator’s Brochure. Ti- more than minimal myelotoxic therapy similar to those in the tusville, NJ: Janssen Research Foundation, 2001. present study until patients who are potentially at higher risk for 15. Venet, M., Angibaud, P., Sanz, G., Poignet, H., End, D., and Bowden., D. Synthesis and in vitro structure-activity relationships of toxicity are studied. The tolerance of this regimen has been imidazolyl-2-quinolinones as farnesyl protein transferase inhibitors. further substantiated in a Phase III study, in which 688 patients Proc. Am. Assoc. Cancer Res., 39: 2171, 1998. with advanced pancreatic cancer were randomized to treatment 16. End, D., Skrzat, S., Devine, A., Angibaud, P., Venet, M., Sanz, G., with either the identical tipifarnib/gemcitabine regimen or gem- and Dowden, C. R115777, a novel imidazole farnesyl protein transfer- citabine on an identical dose schedule plus placebo (22). The ase inhibitor (FTI): biochemical and cellular effects in H-ras and K-ras dominant systems. Proc. Am. Assoc. Cancer Res., 39: 270, 1998. survival of patients treated in the experimental arm was similar 17. Zujewski, J., Horak, I. D., Bol, C. J., Woestenborghs, R., Bowden, to that of patients treated with gemcitabine alone. The results of C., End, D. W., Piotrovsky, V. K., Chiao, J., Belly, R. T., Todd, A., this Phase III study alone should not preclude further evaluations of Kopp, W. C., Kohler, D. R., Chow, C., Noone, M., Hakim, F. T., Larkin, the tipifarnib/gemcitabine regimen in other malignancies, such as G., Gress, R. E., Nussenblatt, R. B., Kremer, A. B., and Cowan, K. H. metastatic breast cancer, in which the single-agents tipifarnib and Phase I and pharmacokinetic study of farnesyl protein transferase inhibitor tipifarnib in advancer cancer. J. Clin. Oncol., 18: 927–941, 2000. gemcitabine each have reported activity (20, 30, 31). It is of interest 18. Hudes, G., Schol, J., Baab, J., Rogatko, A., Bol, C., Horak, I., that the development of tipifarnib continues in hematological ma- Langer, D., Goldstein, L. J., Szarka, C., Meropol, N. J., and Weiner, L. lignancies, gliomas, and breast cancer (28, 29). Phase I clinical and pharmacokinetic trial of the farnesyltransferase

inhibitor tipifarnib on a 21-day schedule. Proc. Am. Soc. Clin. Oncol., 25. Kanazawa, M., Terada, K., Kato, S., and Mori, M. HSDJ, a human 18: 156a, 1999. homolog of DnaJ, is farnesylated and is involved in protein import into 19. Crul, M., de Klerk, G. J., Swart, M., van’t Veer, L. J., de Jong, D., mitochondria. J. Biochem., 121: 890–895, 1997. Boerrigter, L., Palmer, P. A., Bol, C. J., Tan, H., de Gast, G. C., Beijnen, 26. Gibaldi, M., and Perrier, D. Pharmacokinetics, Ed. 2., Appendix D. J. H., and Schellens, J. H. Phase I clinical and pharmacologic study of Estimation of Areas, pp. 445–449. New York: Marcel Dekker, 1982. chronic oral administration of the farnesyl protein transferase inhibitor 27. Lobell, R. B., Liu, D., Buser, C. A., Davide, J. P., DePuy, E., tipifarnib in advanced cancer. J. Clin. Oncol., 20: 2726–2735, 2002. Hamilton, K., Koblan, K. S., Lee, Y., Mosser, S., Motzel, S. L., 20. Johnston, S., Ellis, P., Houston, S., Hickish, T., Howes, A., Palmer, Abbruzzese, J. L., Fuchs, C. S., Rowinsky, E. K., Rubin, E. H., Sharma, P., and Horak, I. A phase II study of the farnesyl transferase inhibitor S., Deutsch, P. J., Mazina, K. E., Morrison, B. W., Wildonger, L., Yao, R115777 in patients with advanced breast cancer. Proc. Am. Soc. Clin. S. L., and Kohn, N. E. Preclinical and clinical pharmacodynamic as- Oncol., 19: 83a, 2000. sessment of L-778, 123, a dual inhibitor of farnesyl:protein transferase 21. Karp, J. E., Lancet, J. E., Kaufmann, S. H., End, D. W., Wright, and geranylgeranyl:protein transferase type-1. Mol. Cancer Ther., 1: J. J., Bol, K., Horak, I., Tidwell, M. L., Liesveld, J., Kottke, T. J., Ange, 747–758, 2002. D., Buddharaju, L., Gojo, I., Highsmith, W. E., Belly, R. T., Hohl, R. J., 28. Gotlib, J., Dugan, K., Katamneni, U., Sridhar, K., Wright, J., Rybak, M. E., Thibault, A., and Rosenblatt, J. Clinical and biologic Thibault, A., Ryback, M. E., and Greenberg, P. L. Phase I/II study of activity of the farnesyltransferase inhibitor tipifarnib in adults with farnesyltransferase inhibitor R115777 (Zarnestra) in patients with my- refractory and relapsed acute leukemias: a phase 1 clinical-laboratory eloproliferative disorders (MPDs); preliminary results. Proc. Am. Soc. correlative trial. Blood, 97: 3361–3369, 2001. Clin. Oncol., 21: 4a, 2002. 22. Van Cutsem, E., Karasek, P., Oettle, H., Vervenne, W. L., Szaw- 29. Cloughesy, T. F., Kuhn, J., Wen, P., Chang, S. M., Schiff, D., lowski, A., Schoffski, P., Post, S., Neumann, H., Safran, H., Humblet, Greenberg, H., Junck, L., Robins, I., DeAngelis, L. M., Yung, A., Y., van de Velde, H., Ma, Y., and Von Hoff, D. Phase III trial comparing Groves, M., Fink, K., Abrey, L. E., Lieberman, F., Mehta, M. P., Raizer, gemcitabine ϩ R115777 (Zarnestra) versus gemcitabine ϩ placebo in J. J., Hess, K., and Prados, M. Phase II trial of R115777 (Zarnestra) in advanced pancreatic cancer (PC). Proc. Am. Soc. Clin. Oncol., 21: 130a, patients with recurrent gliomas not taking enzyme inducing antiepileptic 2002. drugs (EIAED):a North American Brain Tumor Consortium (NABTC) 23. Venook, A. P., Egorin, M. J., Rosner, G. L., Hollis, D., Mani, S., report. Proc. Am. Soc. Clin. Oncol., 21: 80a, 2002. Hawkins, M., Byrd, J., Hohl, R., Budman, D., Meropol, N. J., and 30. Blackstein, M., Vogel, C. L., Ambinder, R., Cowan, J., Inglesias, J., Ratain, M. J. Phase I and pharmacokinetic trial of gemcitabine in and Melemed, A. Gemcitabine as first-line therapy in patients with patients with hepatic or renal dysfunction: Cancer and Leukemia Group metastatic breast cancer: a phase II trial. Oncology, 62: 2–8, 2002. B 9565. J. Clin. Oncol., 18: 2780–2787, 2000. 31. Spielmann, M., Llombart-Cussac, A., Kalla, S., Espie, M., Namer, 24. Kalckar, H. M. Differential spectrophotometry of purine com- M., Ferrero, J. M., Dieras, V., Fumoleau, P., Cuvier, C., Perrocheau, G., pounds by means of specific enzymes. J. Biol. Chem., 167: 461–471, and et al. Single-agent gemcitabine is active in previously treated 1947. metastatic breast cancer. Oncology, 60: 303–307, 2001.

Downloaded from clincancerres.aacrjournals.org on September 26, 2021. © 2003 American Association for Cancer Research. A Phase I, Pharmacokinetic, and Biological Study of the Farnesyltransferase Inhibitor Tipifarnib in Combination with Gemcitabine in Patients with Advanced Malignancies

Amita Patnaik, S. Gail Eckhardt, Elzbieta Izbicka, et al.

Clin Cancer Res 2003;9:4761-4771.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/9/13/4761

Cited articles This article cites 27 articles, 12 of which you can access for free at: http://clincancerres.aacrjournals.org/content/9/13/4761.full#ref-list-1

Citing articles This article has been cited by 3 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/9/13/4761.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/9/13/4761. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.