Phase I and Pharmacologic Study of 17-(Allylamino)-17-Demethoxygeldanamycin in Adult Patients With Solid Tumors
http://www.100md.com
《临床肿瘤学》
the Center for Cancer Research, Division of Cancer Prevention, and Cancer Therapy Evaluation Program, Division of Cancer Treatment and Diagnosis, National Cancer Institute, Bethesda, MD
ABSTRACT
PURPOSE: To determine the clinical toxicities of 17-(allylamino)-17-demethoxygeldanamycin (17-AAG) given as a 1-hour infusion daily for 5 days every 3 weeks.
PATIENTS AND METHODS: Nineteen patients received 17-AAG over six dose levels (10 to 56 mg/m2) using an accelerated titration scheme. Drug levels of 17-AAG were determined by high-performance liquid chromatography. Biologic effects of 17-AAG were monitored by changes in the content of target proteins by immunoblot analysis of lysates prepared from peripheral-blood mononuclear cells.
RESULTS: Toxicity was acceptable at doses up to 28 mg/m2. The cohort was expanded to three patients at 40 mg/m2 because a second occurrence of grade 2 hepatic transaminitis occurred. Two of six assessable patients who received 56 mg/m2 had reversible, grade 3 hepatic transaminitis. Five additional patients were enrolled at 40 mg/m2; none had dose-limiting toxicity. The maximum plasma concentrations (Cmax) of 17-AAG at 40 and 56 mg/m2 were 1,724 and 2,046 ng/mL, respectively; the average plasma exposures (AUC) were 2,809 and 6,708 hours·ng/mL, respectively. Less than 3% of the daily dose was excreted into the urine. Clearance did not correlate with body-surface area. Possible biologic activity was suggested by apparent increased protein content of either glucose-related 78 kd protein or heat shock protein 70 with 14 mg/m2 and decreased protein content of either Lck or Raf1 with 28 mg/m2 of 17-AAG.
CONCLUSION: 17-AAG 40 mg/m2 (median dose, 70 mg) was well tolerated when given daily for 5 days every 3 weeks.
INTRODUCTION
17-(Allylamino)-17-demethoxygeldanamycin (NSC 330507; 17-AAG) is a derivative of geldanamycin, a benzoquinoid ansamycin compound produced by yeast that has growth inhibitory activity against tumor cells.1,2 The target of this derivative is heat shock protein 90 (Hsp90), a highly conserved protein found in relative abundance in mammalian cells.3–5 Hsp90 interacts with a number of key signaling proteins and is required for their normal stability, function, and intracellular localization. Hsp90 promotes the correct conformational folding of a subset of cell signaling proteins into their functional forms and may enhance the proper refolding of proteins destabilized by various environmental stresses. Under nonstress conditions, Hsp90 is predominantly cytoplasmic; after acute stress, it rapidly accumulates in nuclei. A large number of client proteins functionally depend on Hsp90 and include ligand-activated transcription factors (estrogen, progesterone, glucocorticoid, and androgen receptors) and certain transmembrane receptors (platelet-derived growth factor receptor and epidermal growth factor receptors Her1 and Her2).6–11 Inhibition of Hsp90 also interferes with multiple intracellular signal transduction pathways by destabilizing members of the Src family, Bcr-Abl, Akt, cyclin-dependent-kinases Cdk4 and Cdk6, Raf-1, mutated p53, Wee1, and certain basic helix-loop-helix transcription factors, including hypoxia-inducible factor 1.3–5,12–19
High-affinity binding of geldanamycin to the adenosine triphosphate-dependent conformational switch region of Hsp90 alters the chaperone's association with its client proteins, leading to rapid degradation of client proteins by the ubiquitin-dependent proteosome pathway.3–5,20–22 Cancer cells frequently overexpress Hsp90 and are very sensitive to its pharmacologic disruption, possibly because many survival and growth-regulating proteins are dependent on normal Hsp90 chaperone function.3–5
Preclinical activity of geldanamycin has been shown in cancer xenografts, but dose-dependent lethal liver toxicity in vivo precluded its development.23 A search for less toxic analogs led to the development of 17-AAG.24 Incubation of 17-AAG with hepatic microsomes generated several metabolites, primarily 17-amino-17-demethoxygeldanamycin (NSC255109; 17-AG), resulting from the oxidative loss of the 17-allyl group from the ansamycin ring.25 17-AG is biologically active and binds to Hsp90 with equipotency as 17-AAG. Murine studies confirmed that 17-AG is a major metabolite of 17-AAG.26 Murine studies suggested 17-AAG clearance was saturable.26 In vitro, cytochrome P450 isoform CYP3A4 was responsible for this metabolic conversion, suggesting a potential for drug-drug interactions impacting on the pharmacokinetics of 17-AAG with agents that alter CYP3A4 activity (grapefruit juice, ketoconazole, fluconazole, itraconazole, cyclosporine, erythromycin, clarithromycin, cimetidine, terfenadine, astemizole, indinavir, and nelfinavir).25
The most recently developed formulation of 17-AAG is a dimethylsulfoxide/phospholipid vehicle. Gastrointestinal, hepatic, and cholecystic toxicities were dose-limiting in beagle dogs that received 17-AAG on a daily-for-5-days schedule. The recommended starting dose of 17-AAG was 10 mg/m2/d given over 1 hour intravenously daily for 5 days every 3 weeks (one 10th the maximum-tolerated dose [MTD] in dogs). Binding of geldanamycin analogs to Hsp90 causes dissociation from the transcription factor Hsf1; Hsf1 trimerizes and is converted to its active form, which induces transcription of several heat shock proteins, including Hsp90 and Hsp72. Glucose-related 78 kd protein (Grp78) is constitutively expressed in endoplasmic reticulum lumens but is rapidly elevated in response to Hsp90, as is its cytoplasmic homolog Hsp70. The serine/threonine kinase Raf-1 is a component of the Ras-MAP kinase signaling pathway involved in proliferation, differentiation, and apoptosis. Interaction with Hsp90 is required for Raf-1 stability, function, and proper intracellular localization. Lck, a member of the Src family tyrosine kinases, is expressed at high levels in resting lymphocytes. These latter two proteins are sensitive to destabilization/degradation after 17-AAG. These four biomarkers were selected for monitoring in this trial. Preliminary results of this trial were presented in abstract form.27,28 The final results are reported herein.
PATIENTS AND METHODS
Inclusion criteria included adult patients with solid tumors whose disease had progressed on standard therapy, a performance status of 0 to 2, recovery from prior therapies, absolute granulocyte and platelet counts 2,000/μL and 100,000/μL, normal serum bilirubin and creatinine, and AST two times the upper limits of normal. Exclusion criteria included concomitant treatment with drugs that interfere with hepatic CYP3A4 metabolism, serious concurrent illness that would jeopardize the ability to receive the protocol treatment, and history of serious allergy to egg products. This study had the approval of the local institutional review boards. All patients gave written informed consent.
17-AAG was provided by the Cancer Therapy Evaluation Program, Division of Cancer Treatment and Diagnosis, National Cancer Institute (NCI), Bethesda, MD, as a sterile vial containing 50 mg of 17-AAG in 2.0 mL of dimethylsulfoxide. The diluent (egg phospholipid [EPL] diluent; NSC 704057) was supplied in a 50 mL vial containing 48 mL of 2% egg phospholipids and 5% dextrose in water. After thawing, 2 mL of the concentrate was added to 48 mL of EPL diluent. The final solution was dispensed in a glass bottle and infused intravenously over 1 hour.
An accelerated titration phase I clinical trial design as outlined by Simon et al (design 2b) was used with 40% dose increments.29 One patient was entered per dose level until either one exhibited cycle 1 dose-limiting toxicity (DLT) or two patients (at any dose level) experienced grade 2 drug-related toxicity during their first cycle. The accelerated dose titration scheme was then stopped; additional patients were added at the dose that triggered the change. The study design then called for three patients to be treated at each subsequent level. If one of three patients experienced cycle 1 DLT, the cohort was expanded to six patients; if no other patients experienced DLT, dose escalation proceeded. If two or more patients at a given level experienced DLT, the MTD was exceeded, and three more patients were to be treated at the lower dose. Patients were re-treated on day 22 provided that all drug-related toxicities had recovered to at least grade 0 or 1 (except for anemia and alopecia), the absolute granulocytes were 2,000/μL, and the platelets were 100,000/μL.
Toxicity was graded according to NCI Common Toxicity Criteria version 2.0 (http://ctep.info.nih.gov). DLT was defined as any grade 3 nonhematologic toxicity, grade 4 hematologic toxicity, or the inability to resume treatment by day 36 (longer than a 2-week delay) because of drug-related toxicity (except anemia and alopecia of any grade, grade 3 nausea or vomiting with suboptimal antiemetics, and grade 3 fever in the absence of neutropenia and infection).
At the start of each cycle, a history and physical examination, CBC count with differential, AST, ALT, total bilirubin, gamma glutamyltransferase, alkaline phosphatase, serum electrolytes, blood urea nitrogen, creatinine, and urinalysis were done. A CBC count with differential, AST/ALT, and total bilirubin were obtained twice weekly. Tumor restaging was done every two cycles.
Clinical Pharmacokinetics
Planned blood samples were to be obtained from a peripheral vein on days 1 and 5 at the following times, with time zero indicating the start of the 60-minute infusion: preinfusion; 20, 40, and 55 minutes, then at 75 and 90 minutes and 2, 2.5, 3.5, 4.5, 6.5, 8.5, 12.5, and 24 hours. Heparinized blood was collected and stored on ice until delivered to the laboratory, then centrifuged at 800 x g for 15 minutes at 4°C; the plasma was isolated and stored at –70°C until analysis. Urine collections (24-hour) were performed for the first 3 days of cycle one. An aliquot was frozen and stored at –70°C.
17-AG was provided by the Drug Synthesis and Chemistry Branch, Division of Cancer Treatment and Diagnosis, NCI. All chemicals were purchased from Sigma Chemicals (St Louis, MO) unless otherwise specified. High-performance liquid chromatography reagent-grade solvents were purchased from Fisher Scientific (Fair Lawn, NJ). Plasma (0.5 mL) was extracted with 10 mL of ethyl acetate, evaporated to dryness for 40 minutes at 40°C, and resuspended in 220 μL of solvent A (see below). The Waters (Milford, MA) high-performance liquid chromatography system consisted of a 600E pump and controller, a 717-Plus autosampler (0.1 mL injected), and a 996 photodiode array detector. 17-AAG and 17-AG were resolved with a Phenomenex (Torrance, CA) C18 Kingsorb column (3 μ, 150 x 4.6 mm) equipped with a C18 Security guard cartridge (Phenomenex, Torrance, CA). A gradient elution method was used (1 mL/min): solvents A and B were 25 mmol/L of ammonium acetate, pH 4.0, in acetonitrile (A, 75%, 25%; B, 50%/50% volume to volume ratio). For the initial 10 minutes, the composition of the mobile phase was 85% A, 15% B, followed by a linear change to 100% B over 10 minutes, then kept at 100% B for another 20 minutes. The column was then equilibrated at initial conditions for 10 minutes. 17-AG And 17-AAG were monitored at 328.5 nm and 333 nm. The retention times of 17-AG and 17-AAG were 25.1 and 36.3 minutes, respectively. An external calibration curve was used. The data were analyzed by noncompartmental methods with WINNonlin Pro version 4.0 (Pharsight Corporation, Mountain View, CA). The area under the plasma concentration-time curve was determined by a linear trapezoidal model.
A volume of 1 mL of urine was combined with 10 μL of 100 μmol/L of geldanamycin (as an internal standard) in a 1.5-mL tube and vortex-mixed for 15 seconds. The urine samples were subjected to solid-phase extraction followed by analysis using isocratic reversed-phase high-performance liquid chromatography assay with photodiode array as previously described.30 The retention times for 17-AG and 17-AAG were 4.9 and 14.6 minutes, respectively. The linear range of both assays was 12.5 to 2,500 nmol/L.
Molecular Marker Studies
Peripheral blood was drawn into a 10-mL heparinized tube daily or every other day for the initial 5 days of cycle 1 before the 17-AAG dose. Mononuclear cells were isolated by Ficoll-Hypaque density centrifugation; intact cell pellets were frozen and stored at –70°C. The pellets were thawed in the presence of NP-40-containing lysis buffer (with protease inhibitors), and the protein concentration was determined; equal amounts of total protein (20 to 50 μg) were separated by sodium dodecyl sulfate-polyacrylamide electrophoresis and electro-transferred to nitrocellulose membranes, which were blotted with appropriate antibodies. Signals were detected by chemiluminescence as previously described.31,32
RESULTS
Patients
Nineteen patients were entered at six dose levels of 17-AAG ranging from 10 to 56 mg/m2. The majority of patients had only minor cancer-related symptoms, had tumors that arose in the gastrointestinal tract, and had been heavily pretreated (Table 1). A total of 36 cycles were given (median number, two cycles; range, one to four cycles). The median time to treatment failure was 42 days. Three patients (16%) had stable disease at the initial restaging before cycle 3.
Clinical Toxicity
One patient was entered at each of the first three levels (10, 14, and 20 mg/m2) without incident (Table 2). The patient entered at 28 mg/m2 had a grade 2 elevation in the alkaline phosphatase during cycle 1. Dose escalation proceeded to 40 mg/m2; the first patient experienced grade 2 AST elevation. Therefore, two additional patients were entered; neither experienced worse than grade 1 nonhematologic toxicity. The dose was escalated to 56 mg/m2; the first patient entered at this dose level experienced grade 3 alkaline phosphatase elevation during cycle 1, and the dose level was expanded. One patient received only two doses of drug because of a sudden increase in the liver function test results. Laboratory studies to determine eligibility done 5 days before starting drug revealed an AST of 63 U/L and a bilirubin of 0.7 mg/dL. Repeat liver function tests obtained on day 1 of therapy revealed the bilirubin and AST had increased to 1.8 mg/dL and 111 U/L. The elevated liver function tests were confirmed, and 17-AAG was discontinued. Restaging computed tomography studies confirmed massive progression in liver disease compared with the baseline computed tomography scan obtained 2 weeks before. Before discharge home to hospice care, the bilirubin and AST were 3.7 mg/dL and 847 U/L, respectively. Because the elevation in the liver function tests occurred before dosing and was due to disease progression, 17-AAG was not thought to be causal. Two of five more patients had transient grade 3 elevations in AST, ALT, and/or alkaline phosphatase in cycle 1, which exceeded the MTD. Additional patients were enrolled at 40 mg/m2, the recommended dose for future studies. Nonhematologic toxicity per dose level for each assessable cycle is shown in Table 3. Reversible elevations in the AST, ALT, alkaline phosphatase, bilirubin, and gamma glutamyltransferase were the most noticeable toxicities. The median onset of AST elevation was day 5 (range, 4 to 8 days), and the value remained elevated for a median of 7 days (range, 4 to 11 days). Figure 1 shows a representative profile for a patient who had dose-limiting hepatitis. Mild to moderate nausea and vomiting complicated 11% and 36% of cycles at 40 and 56 mg/m2, respectively, but patients responded well to antiemetic therapy. Most patients complained of mild to moderate fatigue at all dose levels. Low grade fever (38.0 to 39.0°C) was documented in 17%, 22%, and 55% of cycles at 28, 40, and 56 mg/m2, respectively. Grade 2 granulocyte toxicity was observed in only one cycle (40 mg/m2); the worst platelet toxicity was grade 1 (56 mg/m2).
Pharmacokinetic Data
Pharmacokinetic sampling was performed after the first dose of 17-AAG in all patients; the data for 40 and 56 mg/m2 are shown in Figure 2. 17-AG plasma levels were determined at the two highest dose levels (Fig 3). The apparent increase in 17-AG plasma levels at 24 hours is likely explained by the drop-out of patients whose plasma levels were not measurable at 24 hours. The pharmacokinetic data are summarized in Tables 4 and 5. The maximum plasma concentration (Cmax) at the two highest doses averaged 1,724 and 2,046 ng/mL (2.9 and 3.5 nmol/mL), respectively. No correlation was evident between body-surface area and either 17-AAG clearance (Pearson correlation coefficient, r = 0.347; P = .145) or volume of distribution (r = 0.321; P = .181). When the data for all doses were combined, the terminal half-life of 17-AAG averaged 4.8 hours, and the clearance averaged 180 L/h. Because the metabolic conversion to 17-AG is not known, the clearance and volume of distribution are presented as divided by the fraction metabolized to 17-AG. A disproportionately large increase in the area under the curve (AUC) of 17-AAG (2.4-fold) and 17-AG (5.1-fold) was seen from 40 to 56 mg/m2, suggesting possible nonlinear pharmacokinetics. However, there was no correlation between dose and clearance (r = –0.214; P = .38), and no significant differences were seen when the clearances where compared according to whether the patient received more or less than the median dose. Because pharmacokinetic data on more than one patient are only available at the two highest doses, the issue of nonlinear clearance cannot be resolved.
Renal excretion of both 17-AAG and 17-AG seemed to be minor; the daily excretion of both active compounds averaged less than 3% of the total daily dose of 17-AAG. There was no correlation between the AUC of 17-AAG or 17-AG and the absence or presence of AST elevations: normal AST values for cycle 1 (n = 10) versus grade 1 to 3 AST elevations (n = 8): 17-AAG, median 3,598 versus 2,567 hours·ng/mL; 17-AG, 1,725 versus 724 ng·ng/mL.
Repeat sampling was obtained after the fifth dose in 14 patients. The day 5 dose was not given in four patients (held for elevated liver function tests, three patients; onset of disease-related melena, one patient), and blood samples were not drawn in another because of poor venous access. In the sample drawn before the day 5 dose, 17-AAG was detected in only three patients, whereas 17-AG was detected in all samples (average of 108 ng/mL at 40 to 56 mg/m2). The median Cmax (25th and 75th percentiles) of 17-AG was significantly higher after the fifth daily dose of 17-AAG: 211 ng/mL (range, 150 to 370 ng/mL) versus 128 ng/mL (range, 81 to 154 ng/mL; P < .001, Wilcoxon rank sum test), as was the 17-AG AUC: 1,653 ng·hr/mL (range, 1,086 to 4,502 ng·hr/mL) versus 720 ng·hr/mL (range, 578 to 823 ng·hr/mL; P < .001). The AUC of 17-AAG tended to be higher on day 5: 3,119 ng·hr/mL (range, 1,642 to 4,763 ng·hr/mL) versus 2,397 ng·hr/mL (range, 1,396 to 2, 646 ng·hr/mL; P = .035).
Molecular Marker Studies
Mononuclear cells from peripheral blood were used as a surrogate tissue to examine possible biologic effects of 17-AAG. Protein levels were qualitatively determined by Western blot analysis. An apparent increase in Grp78 was seen by day 2 or 3 with all but the lowest dose of 17-AAG (Fig 4). There was insufficient protein to run all four markers, and therefore, the data are presented as an increase in either Grp78 or Hsp70 or a decrease in either Lck or Raf-1. Apparent increases in Grp78 or Hsp70 were seen with doses 14 mg/m2, whereas decreases in Lck or Raf-1 were seen in some patients at does 28 mg/m2 (Table 6). Overall, 72% of samples showed an increase in either Grp78 or Hsp70; 45% showed a decrease in either Lck or Raf1.
DISCUSSION
The initial clinical development of 17-AAG involved two schedules: a 1-hour infusion daily for 5 days every 3 weeks or weekly. In our clinical trial, reversible, grade 3 hepatic toxicity manifested by elevations in various liver enzymes was dose-limiting at 56 mg/m2; 40 mg/m2 daily for 5 days is the recommended dose. The transaminases had returned to baseline before the next cycle of 17-AAG. Patients who experienced transaminase elevation during cycle 1 invariably were noted to have transaminitis during cycle 2. Because very few patients received more than two cycles, the potential for cumulative hepatotoxicity is unknown. Another trial using the same schedule reported dose-limiting diarrhea, thrombocytopenia, and reversible hepatotoxicity at 80 mg/m2.33 In contrast, clinical trials using the weekly schedule have reported dose-limiting liver toxicity (bilirubin, AST), fatigue, nausea/vomiting, and anemia with doses 431 mg/m2.34,35 Because the mechanism of 17-AAG-associated hepatotoxicity has not been elucidated, the schedule-dependent hepatic toxicity argues that sequential daily treatment may not be optimal; intermittent exposure to 17-AAG may be better tolerated.
Pharmacokinetic data indicated considerable interpatient variability. Because patients were not allowed to take drugs or foods that are known to affect CYP3A4 activity, interpatient variability may in part be due to polymorphisms in CYP3A4 and CYP3A5 genes that could affect 17-AAG metabolism.25,36,37 Although the Cmax of 17-AG was lower than that of 17-AAG, the AUC was 25% to 50% of the 17-AAG AUC because of its longer half-life. There was a striking difference in the pharmacokinetic parameters of 17-AG within individual patients when the day 1 and day 5 results were compared, which may in part be due to the continued presence of 17-AG 24 hours later; however, the higher AUC suggests that daily dosing may interfere with clearance of this metabolite. The AUC of 17-AAG was also higher, but there was considerable overlap in the day 1 and day 5 values. The increase in 17-AG AUC with daily dosing provides another reason to favor intermittent dosing. Because there was no correlation between body-surface area and clearance of drug, dosing unrelated to body-surface area may be appropriate in future studies.
We did not measure protein binding in this study, but 17-AAG is known to be highly protein bound in murine plasma. Although 17-AAG exerts biologic activity in cell culture experiments at nanomolar concentrations, this largely represents free drug. The binding affinity of 17-AAG for Hsp90 is in the low micromolar concentration range. Geldanamycin and 17-AAG accumulate in cells in vitro, producing higher intracellular concentrations than expected.38 The specific method of membrane transport has not yet been elucidated.
The results of the biomarker studies suggest that biologic effects were seen as early as 24 to 48 hours from the start of a daily schedule of 17-AAG compared with each patient's pretherapy samples. 17-AG is equipotent with 17-AAG; their combined exposure likely contributes to the apparent molecular effects. It is possible that other factors (eg, other drugs, stress) may have contributed to changes in the protein content, but 17-AAG was the only consistent treatment received. If these observations are real, the failure to identify down-regulation of Lck and Raf1 at lower doses may perhaps be due to differences in signal pathway characteristics. Because we do not have control data for the daily biomarker studies in subjects who have not received 17-AAG, we cannot exclude the possibility that these findings represent random fluctuations.
Based on the collective experience seen with the initial phase I trials, intermittent dosing schedules of 17-AAG are being explored. Ongoing efforts to develop analogs may offer a more convenient formulation and less metabolism to potentially toxic metabolites.39 Incorporation of pharmacologic and biomarker studies may provide important insights and help select the optimal regimen, but attention must be paid toward appropriate validation of such biomarker studies.
Authors' Disclosures of Potential Conflicts of Interest
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Employment: J. Michael Hamilton, GlaxoSmithKline.
NOTES
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
DeBoer C, Meulman PA, Wnuk RJ, et al: Geldanamycin, a new antibiotic. J Antibiot (Tokyo) 23:442-447, 1970
Schulte TW, Neckers LM: The benzoquinone ansamycin 17-allylamino-17-demethoxygeldanamycin binds to HSP90 and shares important biologic activities with geldanamycin. Cancer Chemother Pharmacol 42:273-279, 1998
Ochel HJ, Eichhorn K, Gademann G: Geldanamycin: The prototype of a class of antitumor drugs targeting the heat shock protein 90 family of molecular chaperones. Cell Stress Chaperones 6:105-112, 2001
Neckers L: Hsp90 inhibitors as novel cancer chemotherapeutic agents. Trends Mol Med 8:S55-61, 2002 (suppl)
Blagosklonny MV: Hsp-90-associated oncoproteins: Multiple targets of geldanamycin and its analogs. Leukemia 16:455-462, 2002
Fliss AE, Benzeno S, Rao J, et al: Control of estrogen receptor ligand binding by Hsp90. J Steroid Biochem Mol Biol 72:223-230, 2000
Smith DF, Whitesell L, Nair SC, et al: Progesterone receptor structure and function altered by geldanamycin, an hsp90-binding agent. Mol Cell Biol 15:6804-6812, 1995
Vanaja DK, Mitchell SH, Toft DO, et al: Effect of geldanamycin on androgen receptor function and stability. Cell Stress Chaperones 7:55-64, 2002
Supino-Rosin L, Yoshimura A, Yarden Y, et al: Intracellular retention and degradation of the epidermal growth factor receptor, two distinct processes mediated by benzoquinone ansamycins. J Biol Chem 275:21850-21855, 2000
Munster PN, Marchion DC, Basso AD, et al: Degradation of HER2 by ansamycins induces growth arrest and apoptosis in cells with HER2 overexpression via a HER3, phosphatidylinositol 3'-kinase-AKT-dependent pathway. Cancer Res 62:3132-3137, 2002
Bucci M, Roviezzo F, Cicala C, et al: Geldanamycin, an inhibitor of heat shock protein 90 (Hsp90) mediated signal transduction has anti-inflammatory effects and interacts with glucocorticoid receptor in vivo. Br J Pharmacol 131:13-16, 2000
Whitesell L, Mimnaugh EG, De Costa B, et al: Inhibition of heat shock protein HSP90-pp60v-src heteroprotein complex formation by benzoquinone ansamycins: Essential role for stress proteins in oncogenic transformation. Proc Natl Acad Sci U S A 91:8324-8328, 1994
An WG, Schulte TW, Neckers LM: The heat shock protein 90 antagonist geldanamycin alters chaperone association with p210bcr-abl and v-src proteins before their degradation by the proteasome. Cell Growth Differ 11:355-360, 2000
Bijlmakers MJ, Marsh M: Hsp90 is essential for the synthesis and subsequent membrane association, but not the maintenance, of the Src-kinase p56(lck). Mol Biol Cell 11:1585-1595, 2000
Nimmanapalli R, O'Bryan E, Bhalla K: Geldanamycin and its analogue 17-allylamino-17-demethoxygeldanamycin lowers Bcr-Abl levels and induces apoptosis and differentiation of Bcr-Abl-positive human leukemic blasts. Cancer Res 61:1799-1804, 2001
Schulte TW, Blagosklonny MV, Ingui C, et al: Disruption of the Raf-1-Hsp90 molecular complex results in destabilization of Raf-1 and loss of Raf-1-Ras association. J Biol Chem 270:24585-24588, 1995
Whitesell L, Sutphin P, An WG, et al: Geldanamycin-stimulated destabilization of mutated p53 is mediated by the proteasome in vivo. Oncogene 14:2809-2816, 1997
Whitesell L, Sutphin PD, Pulcini EJ, et al: The physical association of multiple molecular chaperone proteins with mutant p53 is altered by geldanamycin, an hsp90-binding agent. Mol Cell Biol 18:1517-1524, 1998
Mabjeesh NJ, Post DE, Willard MT, et al: Geldanamycin induces degradation of hypoxia-inducible factor 1alpha protein via the proteosome pathway in prostate cancer cells. Cancer Res 62:2478-2482, 2002
Prodromou C, Roe SM, O'Brien R, et al: Identification and structural characterization of the ATP/ADP-binding site in the Hsp90 molecular chaperone. Cell 90:65-75, 1997
Grenert JP, Sullivan WP, Fadden P, et al: The amino-terminal domain of heat shock protein 90 (hsp90) that binds geldanamycin is an ATP/ADP switch domain that regulates hsp90 conformation. J Biol Chem 272:23843-23850, 1997
Stebbins CE, Russo AA, Schneider C, et al: Crystal structure of an Hsp90-geldanamycin complex: Targeting of a protein chaperone by an antitumor agent. Cell 89:239-250, 1997
Supko JG, Hickman RL, Grever MR, et al: Preclinical pharmacologic evaluation of geldanamycin as an antitumor agent. Cancer Chemother Pharmacol 36:305-315, 1995
Page J, Heath J, Fulton R, et al: Comparison of geldanamycin (NSC-122750) and 17-allylaminogeldanamycin (NSC-330507D) toxicity in rats. Proc Am Assoc Cancer Res 38:308, 1997 (abstr 2067)
Egorin MJ, Rosen DM, Wolff JH, et al: Metabolism of 17-(allylamino)-17- demethoxygeldanamycin (NSC 330507) by murine and human hepatic preparations. Cancer Res 58:2385-2396, 1998
Egorin MJ, Zuhowski EG, Rosen DM, et al: Plasma pharmacokinetics and tissue distribution of 17-(allylamino)-17-demethoxygeldanamycin (NSC 330507) in CD2F1 mice1. Cancer Chemother Pharmacol 47:291-302, 2001
Wilson RH, Takimoto CH, Agnew EB, et al: Phase I pharmacologic study of 17-(allylamino)-17-demethoxygeldanamycin in adult patients with advanced solid tumors. Proc Am Soc Clin Oncol 20:82a, 2001 (abstr 325)
Agnew EB, Wilson RH, Morrison G, et al: Clinical pharmacokinetics of 17-(allylamino)-17-demethoxygeldanaymcin and the active metabolite 17-(amino)-demethoxygeldanamycin given as a one-hour infusion daily for 5 days. Proc Am Assoc Cancer Res 43:272, 2002 (abstr 1349)
Simon R, Freidlin B, Rubinstein L, et al: Accelerated titration designs for phase I clinical trials in oncology. J Natl Cancer Inst 89:1138-1147, 1997
Agnew EB, Wilson RH, Grem JL, et al: Measurement of the novel antitumor agent 17-(allylamino)-17-demethoxygeldanamycin in human plasma by high-performance liquid chromatography. J Chromatogr B Biomed Sci Appl 755:237-243, 2001
Marcu MG, Doyle M, Bertolotti A, et al: Heat shock protein 90 modulates the unfolded protein response by stabilizing IRE1alpha. Mol Cell Biol 22:8506-8513, 2002
Marcu M, Schulte TW, Neckers LM: Novobiocin and other coumarin antibiotics bind to heat shock protein 90 and result in the degradation of hsp90-dependent signaling proteins p185erbB2, p60v-src, raf-1 and mutated p53. J Natl Cancer Inst 92:242-248, 2000
Munster PN, Tong W, Schwartz L, et al: Phase I trial of 17-(allylamino)-17-demethoxygeldanamycin in patients with advanced solid malignancies. Proc Am Soc Clin Oncol 20:82a, 2001 (abstr 327)
Goetz M, Toft D, Reid J, et al: A phase I trial of 17-allyl-amino-geldanamycin in patients with advanced cancer. Eur J Cancer 38:54, 2002 (suppl 7, abstr 170)
Banerji U, O'Donnell A, Scurr M, et al: Phase I trial of the heat shock protein 90 (hsp90) inhibitor 17-allylamino 17-demethoxygeldanamycin: Pharmacokinetic profile and pharmacodynamic endpoints. Proc Am Soc Clin Oncol 20:82a, 2001 (abstr 326)
Nagasubramanian R, Innocenti F, Ratain MJ: Pharmacogenetics in cancer treatment. Annu Rev Med 54:427-452, 2003
van Schaik RHN: Implications of cytochrome P450 genetic polymorphisms on the toxicity of antitumor agents. Ther Drug Monit 26:236-240, 2004
Chiosis G, Huerzo H, Rosen N, et al: 17AAG: Low target binding affinity and potent cell activity-finding an explanation. Mol Cancer Ther 2:123-129, 2003
Egorin MJ, Lagattuta TF, Hamburger DR, et al: Pharmacokinetics, tissue distribution, and metabolism of 17-(dimethylaminoethylamino)-17-demethoxygeldanamycin (NSC 707545) in CD2F1 mice and Fischer 344 rats. Cancer Chemother Pharmacol 49:7-19, 2002(Jean L. Grem, Geraldine M)
ABSTRACT
PURPOSE: To determine the clinical toxicities of 17-(allylamino)-17-demethoxygeldanamycin (17-AAG) given as a 1-hour infusion daily for 5 days every 3 weeks.
PATIENTS AND METHODS: Nineteen patients received 17-AAG over six dose levels (10 to 56 mg/m2) using an accelerated titration scheme. Drug levels of 17-AAG were determined by high-performance liquid chromatography. Biologic effects of 17-AAG were monitored by changes in the content of target proteins by immunoblot analysis of lysates prepared from peripheral-blood mononuclear cells.
RESULTS: Toxicity was acceptable at doses up to 28 mg/m2. The cohort was expanded to three patients at 40 mg/m2 because a second occurrence of grade 2 hepatic transaminitis occurred. Two of six assessable patients who received 56 mg/m2 had reversible, grade 3 hepatic transaminitis. Five additional patients were enrolled at 40 mg/m2; none had dose-limiting toxicity. The maximum plasma concentrations (Cmax) of 17-AAG at 40 and 56 mg/m2 were 1,724 and 2,046 ng/mL, respectively; the average plasma exposures (AUC) were 2,809 and 6,708 hours·ng/mL, respectively. Less than 3% of the daily dose was excreted into the urine. Clearance did not correlate with body-surface area. Possible biologic activity was suggested by apparent increased protein content of either glucose-related 78 kd protein or heat shock protein 70 with 14 mg/m2 and decreased protein content of either Lck or Raf1 with 28 mg/m2 of 17-AAG.
CONCLUSION: 17-AAG 40 mg/m2 (median dose, 70 mg) was well tolerated when given daily for 5 days every 3 weeks.
INTRODUCTION
17-(Allylamino)-17-demethoxygeldanamycin (NSC 330507; 17-AAG) is a derivative of geldanamycin, a benzoquinoid ansamycin compound produced by yeast that has growth inhibitory activity against tumor cells.1,2 The target of this derivative is heat shock protein 90 (Hsp90), a highly conserved protein found in relative abundance in mammalian cells.3–5 Hsp90 interacts with a number of key signaling proteins and is required for their normal stability, function, and intracellular localization. Hsp90 promotes the correct conformational folding of a subset of cell signaling proteins into their functional forms and may enhance the proper refolding of proteins destabilized by various environmental stresses. Under nonstress conditions, Hsp90 is predominantly cytoplasmic; after acute stress, it rapidly accumulates in nuclei. A large number of client proteins functionally depend on Hsp90 and include ligand-activated transcription factors (estrogen, progesterone, glucocorticoid, and androgen receptors) and certain transmembrane receptors (platelet-derived growth factor receptor and epidermal growth factor receptors Her1 and Her2).6–11 Inhibition of Hsp90 also interferes with multiple intracellular signal transduction pathways by destabilizing members of the Src family, Bcr-Abl, Akt, cyclin-dependent-kinases Cdk4 and Cdk6, Raf-1, mutated p53, Wee1, and certain basic helix-loop-helix transcription factors, including hypoxia-inducible factor 1.3–5,12–19
High-affinity binding of geldanamycin to the adenosine triphosphate-dependent conformational switch region of Hsp90 alters the chaperone's association with its client proteins, leading to rapid degradation of client proteins by the ubiquitin-dependent proteosome pathway.3–5,20–22 Cancer cells frequently overexpress Hsp90 and are very sensitive to its pharmacologic disruption, possibly because many survival and growth-regulating proteins are dependent on normal Hsp90 chaperone function.3–5
Preclinical activity of geldanamycin has been shown in cancer xenografts, but dose-dependent lethal liver toxicity in vivo precluded its development.23 A search for less toxic analogs led to the development of 17-AAG.24 Incubation of 17-AAG with hepatic microsomes generated several metabolites, primarily 17-amino-17-demethoxygeldanamycin (NSC255109; 17-AG), resulting from the oxidative loss of the 17-allyl group from the ansamycin ring.25 17-AG is biologically active and binds to Hsp90 with equipotency as 17-AAG. Murine studies confirmed that 17-AG is a major metabolite of 17-AAG.26 Murine studies suggested 17-AAG clearance was saturable.26 In vitro, cytochrome P450 isoform CYP3A4 was responsible for this metabolic conversion, suggesting a potential for drug-drug interactions impacting on the pharmacokinetics of 17-AAG with agents that alter CYP3A4 activity (grapefruit juice, ketoconazole, fluconazole, itraconazole, cyclosporine, erythromycin, clarithromycin, cimetidine, terfenadine, astemizole, indinavir, and nelfinavir).25
The most recently developed formulation of 17-AAG is a dimethylsulfoxide/phospholipid vehicle. Gastrointestinal, hepatic, and cholecystic toxicities were dose-limiting in beagle dogs that received 17-AAG on a daily-for-5-days schedule. The recommended starting dose of 17-AAG was 10 mg/m2/d given over 1 hour intravenously daily for 5 days every 3 weeks (one 10th the maximum-tolerated dose [MTD] in dogs). Binding of geldanamycin analogs to Hsp90 causes dissociation from the transcription factor Hsf1; Hsf1 trimerizes and is converted to its active form, which induces transcription of several heat shock proteins, including Hsp90 and Hsp72. Glucose-related 78 kd protein (Grp78) is constitutively expressed in endoplasmic reticulum lumens but is rapidly elevated in response to Hsp90, as is its cytoplasmic homolog Hsp70. The serine/threonine kinase Raf-1 is a component of the Ras-MAP kinase signaling pathway involved in proliferation, differentiation, and apoptosis. Interaction with Hsp90 is required for Raf-1 stability, function, and proper intracellular localization. Lck, a member of the Src family tyrosine kinases, is expressed at high levels in resting lymphocytes. These latter two proteins are sensitive to destabilization/degradation after 17-AAG. These four biomarkers were selected for monitoring in this trial. Preliminary results of this trial were presented in abstract form.27,28 The final results are reported herein.
PATIENTS AND METHODS
Inclusion criteria included adult patients with solid tumors whose disease had progressed on standard therapy, a performance status of 0 to 2, recovery from prior therapies, absolute granulocyte and platelet counts 2,000/μL and 100,000/μL, normal serum bilirubin and creatinine, and AST two times the upper limits of normal. Exclusion criteria included concomitant treatment with drugs that interfere with hepatic CYP3A4 metabolism, serious concurrent illness that would jeopardize the ability to receive the protocol treatment, and history of serious allergy to egg products. This study had the approval of the local institutional review boards. All patients gave written informed consent.
17-AAG was provided by the Cancer Therapy Evaluation Program, Division of Cancer Treatment and Diagnosis, National Cancer Institute (NCI), Bethesda, MD, as a sterile vial containing 50 mg of 17-AAG in 2.0 mL of dimethylsulfoxide. The diluent (egg phospholipid [EPL] diluent; NSC 704057) was supplied in a 50 mL vial containing 48 mL of 2% egg phospholipids and 5% dextrose in water. After thawing, 2 mL of the concentrate was added to 48 mL of EPL diluent. The final solution was dispensed in a glass bottle and infused intravenously over 1 hour.
An accelerated titration phase I clinical trial design as outlined by Simon et al (design 2b) was used with 40% dose increments.29 One patient was entered per dose level until either one exhibited cycle 1 dose-limiting toxicity (DLT) or two patients (at any dose level) experienced grade 2 drug-related toxicity during their first cycle. The accelerated dose titration scheme was then stopped; additional patients were added at the dose that triggered the change. The study design then called for three patients to be treated at each subsequent level. If one of three patients experienced cycle 1 DLT, the cohort was expanded to six patients; if no other patients experienced DLT, dose escalation proceeded. If two or more patients at a given level experienced DLT, the MTD was exceeded, and three more patients were to be treated at the lower dose. Patients were re-treated on day 22 provided that all drug-related toxicities had recovered to at least grade 0 or 1 (except for anemia and alopecia), the absolute granulocytes were 2,000/μL, and the platelets were 100,000/μL.
Toxicity was graded according to NCI Common Toxicity Criteria version 2.0 (http://ctep.info.nih.gov). DLT was defined as any grade 3 nonhematologic toxicity, grade 4 hematologic toxicity, or the inability to resume treatment by day 36 (longer than a 2-week delay) because of drug-related toxicity (except anemia and alopecia of any grade, grade 3 nausea or vomiting with suboptimal antiemetics, and grade 3 fever in the absence of neutropenia and infection).
At the start of each cycle, a history and physical examination, CBC count with differential, AST, ALT, total bilirubin, gamma glutamyltransferase, alkaline phosphatase, serum electrolytes, blood urea nitrogen, creatinine, and urinalysis were done. A CBC count with differential, AST/ALT, and total bilirubin were obtained twice weekly. Tumor restaging was done every two cycles.
Clinical Pharmacokinetics
Planned blood samples were to be obtained from a peripheral vein on days 1 and 5 at the following times, with time zero indicating the start of the 60-minute infusion: preinfusion; 20, 40, and 55 minutes, then at 75 and 90 minutes and 2, 2.5, 3.5, 4.5, 6.5, 8.5, 12.5, and 24 hours. Heparinized blood was collected and stored on ice until delivered to the laboratory, then centrifuged at 800 x g for 15 minutes at 4°C; the plasma was isolated and stored at –70°C until analysis. Urine collections (24-hour) were performed for the first 3 days of cycle one. An aliquot was frozen and stored at –70°C.
17-AG was provided by the Drug Synthesis and Chemistry Branch, Division of Cancer Treatment and Diagnosis, NCI. All chemicals were purchased from Sigma Chemicals (St Louis, MO) unless otherwise specified. High-performance liquid chromatography reagent-grade solvents were purchased from Fisher Scientific (Fair Lawn, NJ). Plasma (0.5 mL) was extracted with 10 mL of ethyl acetate, evaporated to dryness for 40 minutes at 40°C, and resuspended in 220 μL of solvent A (see below). The Waters (Milford, MA) high-performance liquid chromatography system consisted of a 600E pump and controller, a 717-Plus autosampler (0.1 mL injected), and a 996 photodiode array detector. 17-AAG and 17-AG were resolved with a Phenomenex (Torrance, CA) C18 Kingsorb column (3 μ, 150 x 4.6 mm) equipped with a C18 Security guard cartridge (Phenomenex, Torrance, CA). A gradient elution method was used (1 mL/min): solvents A and B were 25 mmol/L of ammonium acetate, pH 4.0, in acetonitrile (A, 75%, 25%; B, 50%/50% volume to volume ratio). For the initial 10 minutes, the composition of the mobile phase was 85% A, 15% B, followed by a linear change to 100% B over 10 minutes, then kept at 100% B for another 20 minutes. The column was then equilibrated at initial conditions for 10 minutes. 17-AG And 17-AAG were monitored at 328.5 nm and 333 nm. The retention times of 17-AG and 17-AAG were 25.1 and 36.3 minutes, respectively. An external calibration curve was used. The data were analyzed by noncompartmental methods with WINNonlin Pro version 4.0 (Pharsight Corporation, Mountain View, CA). The area under the plasma concentration-time curve was determined by a linear trapezoidal model.
A volume of 1 mL of urine was combined with 10 μL of 100 μmol/L of geldanamycin (as an internal standard) in a 1.5-mL tube and vortex-mixed for 15 seconds. The urine samples were subjected to solid-phase extraction followed by analysis using isocratic reversed-phase high-performance liquid chromatography assay with photodiode array as previously described.30 The retention times for 17-AG and 17-AAG were 4.9 and 14.6 minutes, respectively. The linear range of both assays was 12.5 to 2,500 nmol/L.
Molecular Marker Studies
Peripheral blood was drawn into a 10-mL heparinized tube daily or every other day for the initial 5 days of cycle 1 before the 17-AAG dose. Mononuclear cells were isolated by Ficoll-Hypaque density centrifugation; intact cell pellets were frozen and stored at –70°C. The pellets were thawed in the presence of NP-40-containing lysis buffer (with protease inhibitors), and the protein concentration was determined; equal amounts of total protein (20 to 50 μg) were separated by sodium dodecyl sulfate-polyacrylamide electrophoresis and electro-transferred to nitrocellulose membranes, which were blotted with appropriate antibodies. Signals were detected by chemiluminescence as previously described.31,32
RESULTS
Patients
Nineteen patients were entered at six dose levels of 17-AAG ranging from 10 to 56 mg/m2. The majority of patients had only minor cancer-related symptoms, had tumors that arose in the gastrointestinal tract, and had been heavily pretreated (Table 1). A total of 36 cycles were given (median number, two cycles; range, one to four cycles). The median time to treatment failure was 42 days. Three patients (16%) had stable disease at the initial restaging before cycle 3.
Clinical Toxicity
One patient was entered at each of the first three levels (10, 14, and 20 mg/m2) without incident (Table 2). The patient entered at 28 mg/m2 had a grade 2 elevation in the alkaline phosphatase during cycle 1. Dose escalation proceeded to 40 mg/m2; the first patient experienced grade 2 AST elevation. Therefore, two additional patients were entered; neither experienced worse than grade 1 nonhematologic toxicity. The dose was escalated to 56 mg/m2; the first patient entered at this dose level experienced grade 3 alkaline phosphatase elevation during cycle 1, and the dose level was expanded. One patient received only two doses of drug because of a sudden increase in the liver function test results. Laboratory studies to determine eligibility done 5 days before starting drug revealed an AST of 63 U/L and a bilirubin of 0.7 mg/dL. Repeat liver function tests obtained on day 1 of therapy revealed the bilirubin and AST had increased to 1.8 mg/dL and 111 U/L. The elevated liver function tests were confirmed, and 17-AAG was discontinued. Restaging computed tomography studies confirmed massive progression in liver disease compared with the baseline computed tomography scan obtained 2 weeks before. Before discharge home to hospice care, the bilirubin and AST were 3.7 mg/dL and 847 U/L, respectively. Because the elevation in the liver function tests occurred before dosing and was due to disease progression, 17-AAG was not thought to be causal. Two of five more patients had transient grade 3 elevations in AST, ALT, and/or alkaline phosphatase in cycle 1, which exceeded the MTD. Additional patients were enrolled at 40 mg/m2, the recommended dose for future studies. Nonhematologic toxicity per dose level for each assessable cycle is shown in Table 3. Reversible elevations in the AST, ALT, alkaline phosphatase, bilirubin, and gamma glutamyltransferase were the most noticeable toxicities. The median onset of AST elevation was day 5 (range, 4 to 8 days), and the value remained elevated for a median of 7 days (range, 4 to 11 days). Figure 1 shows a representative profile for a patient who had dose-limiting hepatitis. Mild to moderate nausea and vomiting complicated 11% and 36% of cycles at 40 and 56 mg/m2, respectively, but patients responded well to antiemetic therapy. Most patients complained of mild to moderate fatigue at all dose levels. Low grade fever (38.0 to 39.0°C) was documented in 17%, 22%, and 55% of cycles at 28, 40, and 56 mg/m2, respectively. Grade 2 granulocyte toxicity was observed in only one cycle (40 mg/m2); the worst platelet toxicity was grade 1 (56 mg/m2).
Pharmacokinetic Data
Pharmacokinetic sampling was performed after the first dose of 17-AAG in all patients; the data for 40 and 56 mg/m2 are shown in Figure 2. 17-AG plasma levels were determined at the two highest dose levels (Fig 3). The apparent increase in 17-AG plasma levels at 24 hours is likely explained by the drop-out of patients whose plasma levels were not measurable at 24 hours. The pharmacokinetic data are summarized in Tables 4 and 5. The maximum plasma concentration (Cmax) at the two highest doses averaged 1,724 and 2,046 ng/mL (2.9 and 3.5 nmol/mL), respectively. No correlation was evident between body-surface area and either 17-AAG clearance (Pearson correlation coefficient, r = 0.347; P = .145) or volume of distribution (r = 0.321; P = .181). When the data for all doses were combined, the terminal half-life of 17-AAG averaged 4.8 hours, and the clearance averaged 180 L/h. Because the metabolic conversion to 17-AG is not known, the clearance and volume of distribution are presented as divided by the fraction metabolized to 17-AG. A disproportionately large increase in the area under the curve (AUC) of 17-AAG (2.4-fold) and 17-AG (5.1-fold) was seen from 40 to 56 mg/m2, suggesting possible nonlinear pharmacokinetics. However, there was no correlation between dose and clearance (r = –0.214; P = .38), and no significant differences were seen when the clearances where compared according to whether the patient received more or less than the median dose. Because pharmacokinetic data on more than one patient are only available at the two highest doses, the issue of nonlinear clearance cannot be resolved.
Renal excretion of both 17-AAG and 17-AG seemed to be minor; the daily excretion of both active compounds averaged less than 3% of the total daily dose of 17-AAG. There was no correlation between the AUC of 17-AAG or 17-AG and the absence or presence of AST elevations: normal AST values for cycle 1 (n = 10) versus grade 1 to 3 AST elevations (n = 8): 17-AAG, median 3,598 versus 2,567 hours·ng/mL; 17-AG, 1,725 versus 724 ng·ng/mL.
Repeat sampling was obtained after the fifth dose in 14 patients. The day 5 dose was not given in four patients (held for elevated liver function tests, three patients; onset of disease-related melena, one patient), and blood samples were not drawn in another because of poor venous access. In the sample drawn before the day 5 dose, 17-AAG was detected in only three patients, whereas 17-AG was detected in all samples (average of 108 ng/mL at 40 to 56 mg/m2). The median Cmax (25th and 75th percentiles) of 17-AG was significantly higher after the fifth daily dose of 17-AAG: 211 ng/mL (range, 150 to 370 ng/mL) versus 128 ng/mL (range, 81 to 154 ng/mL; P < .001, Wilcoxon rank sum test), as was the 17-AG AUC: 1,653 ng·hr/mL (range, 1,086 to 4,502 ng·hr/mL) versus 720 ng·hr/mL (range, 578 to 823 ng·hr/mL; P < .001). The AUC of 17-AAG tended to be higher on day 5: 3,119 ng·hr/mL (range, 1,642 to 4,763 ng·hr/mL) versus 2,397 ng·hr/mL (range, 1,396 to 2, 646 ng·hr/mL; P = .035).
Molecular Marker Studies
Mononuclear cells from peripheral blood were used as a surrogate tissue to examine possible biologic effects of 17-AAG. Protein levels were qualitatively determined by Western blot analysis. An apparent increase in Grp78 was seen by day 2 or 3 with all but the lowest dose of 17-AAG (Fig 4). There was insufficient protein to run all four markers, and therefore, the data are presented as an increase in either Grp78 or Hsp70 or a decrease in either Lck or Raf-1. Apparent increases in Grp78 or Hsp70 were seen with doses 14 mg/m2, whereas decreases in Lck or Raf-1 were seen in some patients at does 28 mg/m2 (Table 6). Overall, 72% of samples showed an increase in either Grp78 or Hsp70; 45% showed a decrease in either Lck or Raf1.
DISCUSSION
The initial clinical development of 17-AAG involved two schedules: a 1-hour infusion daily for 5 days every 3 weeks or weekly. In our clinical trial, reversible, grade 3 hepatic toxicity manifested by elevations in various liver enzymes was dose-limiting at 56 mg/m2; 40 mg/m2 daily for 5 days is the recommended dose. The transaminases had returned to baseline before the next cycle of 17-AAG. Patients who experienced transaminase elevation during cycle 1 invariably were noted to have transaminitis during cycle 2. Because very few patients received more than two cycles, the potential for cumulative hepatotoxicity is unknown. Another trial using the same schedule reported dose-limiting diarrhea, thrombocytopenia, and reversible hepatotoxicity at 80 mg/m2.33 In contrast, clinical trials using the weekly schedule have reported dose-limiting liver toxicity (bilirubin, AST), fatigue, nausea/vomiting, and anemia with doses 431 mg/m2.34,35 Because the mechanism of 17-AAG-associated hepatotoxicity has not been elucidated, the schedule-dependent hepatic toxicity argues that sequential daily treatment may not be optimal; intermittent exposure to 17-AAG may be better tolerated.
Pharmacokinetic data indicated considerable interpatient variability. Because patients were not allowed to take drugs or foods that are known to affect CYP3A4 activity, interpatient variability may in part be due to polymorphisms in CYP3A4 and CYP3A5 genes that could affect 17-AAG metabolism.25,36,37 Although the Cmax of 17-AG was lower than that of 17-AAG, the AUC was 25% to 50% of the 17-AAG AUC because of its longer half-life. There was a striking difference in the pharmacokinetic parameters of 17-AG within individual patients when the day 1 and day 5 results were compared, which may in part be due to the continued presence of 17-AG 24 hours later; however, the higher AUC suggests that daily dosing may interfere with clearance of this metabolite. The AUC of 17-AAG was also higher, but there was considerable overlap in the day 1 and day 5 values. The increase in 17-AG AUC with daily dosing provides another reason to favor intermittent dosing. Because there was no correlation between body-surface area and clearance of drug, dosing unrelated to body-surface area may be appropriate in future studies.
We did not measure protein binding in this study, but 17-AAG is known to be highly protein bound in murine plasma. Although 17-AAG exerts biologic activity in cell culture experiments at nanomolar concentrations, this largely represents free drug. The binding affinity of 17-AAG for Hsp90 is in the low micromolar concentration range. Geldanamycin and 17-AAG accumulate in cells in vitro, producing higher intracellular concentrations than expected.38 The specific method of membrane transport has not yet been elucidated.
The results of the biomarker studies suggest that biologic effects were seen as early as 24 to 48 hours from the start of a daily schedule of 17-AAG compared with each patient's pretherapy samples. 17-AG is equipotent with 17-AAG; their combined exposure likely contributes to the apparent molecular effects. It is possible that other factors (eg, other drugs, stress) may have contributed to changes in the protein content, but 17-AAG was the only consistent treatment received. If these observations are real, the failure to identify down-regulation of Lck and Raf1 at lower doses may perhaps be due to differences in signal pathway characteristics. Because we do not have control data for the daily biomarker studies in subjects who have not received 17-AAG, we cannot exclude the possibility that these findings represent random fluctuations.
Based on the collective experience seen with the initial phase I trials, intermittent dosing schedules of 17-AAG are being explored. Ongoing efforts to develop analogs may offer a more convenient formulation and less metabolism to potentially toxic metabolites.39 Incorporation of pharmacologic and biomarker studies may provide important insights and help select the optimal regimen, but attention must be paid toward appropriate validation of such biomarker studies.
Authors' Disclosures of Potential Conflicts of Interest
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Employment: J. Michael Hamilton, GlaxoSmithKline.
NOTES
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
DeBoer C, Meulman PA, Wnuk RJ, et al: Geldanamycin, a new antibiotic. J Antibiot (Tokyo) 23:442-447, 1970
Schulte TW, Neckers LM: The benzoquinone ansamycin 17-allylamino-17-demethoxygeldanamycin binds to HSP90 and shares important biologic activities with geldanamycin. Cancer Chemother Pharmacol 42:273-279, 1998
Ochel HJ, Eichhorn K, Gademann G: Geldanamycin: The prototype of a class of antitumor drugs targeting the heat shock protein 90 family of molecular chaperones. Cell Stress Chaperones 6:105-112, 2001
Neckers L: Hsp90 inhibitors as novel cancer chemotherapeutic agents. Trends Mol Med 8:S55-61, 2002 (suppl)
Blagosklonny MV: Hsp-90-associated oncoproteins: Multiple targets of geldanamycin and its analogs. Leukemia 16:455-462, 2002
Fliss AE, Benzeno S, Rao J, et al: Control of estrogen receptor ligand binding by Hsp90. J Steroid Biochem Mol Biol 72:223-230, 2000
Smith DF, Whitesell L, Nair SC, et al: Progesterone receptor structure and function altered by geldanamycin, an hsp90-binding agent. Mol Cell Biol 15:6804-6812, 1995
Vanaja DK, Mitchell SH, Toft DO, et al: Effect of geldanamycin on androgen receptor function and stability. Cell Stress Chaperones 7:55-64, 2002
Supino-Rosin L, Yoshimura A, Yarden Y, et al: Intracellular retention and degradation of the epidermal growth factor receptor, two distinct processes mediated by benzoquinone ansamycins. J Biol Chem 275:21850-21855, 2000
Munster PN, Marchion DC, Basso AD, et al: Degradation of HER2 by ansamycins induces growth arrest and apoptosis in cells with HER2 overexpression via a HER3, phosphatidylinositol 3'-kinase-AKT-dependent pathway. Cancer Res 62:3132-3137, 2002
Bucci M, Roviezzo F, Cicala C, et al: Geldanamycin, an inhibitor of heat shock protein 90 (Hsp90) mediated signal transduction has anti-inflammatory effects and interacts with glucocorticoid receptor in vivo. Br J Pharmacol 131:13-16, 2000
Whitesell L, Mimnaugh EG, De Costa B, et al: Inhibition of heat shock protein HSP90-pp60v-src heteroprotein complex formation by benzoquinone ansamycins: Essential role for stress proteins in oncogenic transformation. Proc Natl Acad Sci U S A 91:8324-8328, 1994
An WG, Schulte TW, Neckers LM: The heat shock protein 90 antagonist geldanamycin alters chaperone association with p210bcr-abl and v-src proteins before their degradation by the proteasome. Cell Growth Differ 11:355-360, 2000
Bijlmakers MJ, Marsh M: Hsp90 is essential for the synthesis and subsequent membrane association, but not the maintenance, of the Src-kinase p56(lck). Mol Biol Cell 11:1585-1595, 2000
Nimmanapalli R, O'Bryan E, Bhalla K: Geldanamycin and its analogue 17-allylamino-17-demethoxygeldanamycin lowers Bcr-Abl levels and induces apoptosis and differentiation of Bcr-Abl-positive human leukemic blasts. Cancer Res 61:1799-1804, 2001
Schulte TW, Blagosklonny MV, Ingui C, et al: Disruption of the Raf-1-Hsp90 molecular complex results in destabilization of Raf-1 and loss of Raf-1-Ras association. J Biol Chem 270:24585-24588, 1995
Whitesell L, Sutphin P, An WG, et al: Geldanamycin-stimulated destabilization of mutated p53 is mediated by the proteasome in vivo. Oncogene 14:2809-2816, 1997
Whitesell L, Sutphin PD, Pulcini EJ, et al: The physical association of multiple molecular chaperone proteins with mutant p53 is altered by geldanamycin, an hsp90-binding agent. Mol Cell Biol 18:1517-1524, 1998
Mabjeesh NJ, Post DE, Willard MT, et al: Geldanamycin induces degradation of hypoxia-inducible factor 1alpha protein via the proteosome pathway in prostate cancer cells. Cancer Res 62:2478-2482, 2002
Prodromou C, Roe SM, O'Brien R, et al: Identification and structural characterization of the ATP/ADP-binding site in the Hsp90 molecular chaperone. Cell 90:65-75, 1997
Grenert JP, Sullivan WP, Fadden P, et al: The amino-terminal domain of heat shock protein 90 (hsp90) that binds geldanamycin is an ATP/ADP switch domain that regulates hsp90 conformation. J Biol Chem 272:23843-23850, 1997
Stebbins CE, Russo AA, Schneider C, et al: Crystal structure of an Hsp90-geldanamycin complex: Targeting of a protein chaperone by an antitumor agent. Cell 89:239-250, 1997
Supko JG, Hickman RL, Grever MR, et al: Preclinical pharmacologic evaluation of geldanamycin as an antitumor agent. Cancer Chemother Pharmacol 36:305-315, 1995
Page J, Heath J, Fulton R, et al: Comparison of geldanamycin (NSC-122750) and 17-allylaminogeldanamycin (NSC-330507D) toxicity in rats. Proc Am Assoc Cancer Res 38:308, 1997 (abstr 2067)
Egorin MJ, Rosen DM, Wolff JH, et al: Metabolism of 17-(allylamino)-17- demethoxygeldanamycin (NSC 330507) by murine and human hepatic preparations. Cancer Res 58:2385-2396, 1998
Egorin MJ, Zuhowski EG, Rosen DM, et al: Plasma pharmacokinetics and tissue distribution of 17-(allylamino)-17-demethoxygeldanamycin (NSC 330507) in CD2F1 mice1. Cancer Chemother Pharmacol 47:291-302, 2001
Wilson RH, Takimoto CH, Agnew EB, et al: Phase I pharmacologic study of 17-(allylamino)-17-demethoxygeldanamycin in adult patients with advanced solid tumors. Proc Am Soc Clin Oncol 20:82a, 2001 (abstr 325)
Agnew EB, Wilson RH, Morrison G, et al: Clinical pharmacokinetics of 17-(allylamino)-17-demethoxygeldanaymcin and the active metabolite 17-(amino)-demethoxygeldanamycin given as a one-hour infusion daily for 5 days. Proc Am Assoc Cancer Res 43:272, 2002 (abstr 1349)
Simon R, Freidlin B, Rubinstein L, et al: Accelerated titration designs for phase I clinical trials in oncology. J Natl Cancer Inst 89:1138-1147, 1997
Agnew EB, Wilson RH, Grem JL, et al: Measurement of the novel antitumor agent 17-(allylamino)-17-demethoxygeldanamycin in human plasma by high-performance liquid chromatography. J Chromatogr B Biomed Sci Appl 755:237-243, 2001
Marcu MG, Doyle M, Bertolotti A, et al: Heat shock protein 90 modulates the unfolded protein response by stabilizing IRE1alpha. Mol Cell Biol 22:8506-8513, 2002
Marcu M, Schulte TW, Neckers LM: Novobiocin and other coumarin antibiotics bind to heat shock protein 90 and result in the degradation of hsp90-dependent signaling proteins p185erbB2, p60v-src, raf-1 and mutated p53. J Natl Cancer Inst 92:242-248, 2000
Munster PN, Tong W, Schwartz L, et al: Phase I trial of 17-(allylamino)-17-demethoxygeldanamycin in patients with advanced solid malignancies. Proc Am Soc Clin Oncol 20:82a, 2001 (abstr 327)
Goetz M, Toft D, Reid J, et al: A phase I trial of 17-allyl-amino-geldanamycin in patients with advanced cancer. Eur J Cancer 38:54, 2002 (suppl 7, abstr 170)
Banerji U, O'Donnell A, Scurr M, et al: Phase I trial of the heat shock protein 90 (hsp90) inhibitor 17-allylamino 17-demethoxygeldanamycin: Pharmacokinetic profile and pharmacodynamic endpoints. Proc Am Soc Clin Oncol 20:82a, 2001 (abstr 326)
Nagasubramanian R, Innocenti F, Ratain MJ: Pharmacogenetics in cancer treatment. Annu Rev Med 54:427-452, 2003
van Schaik RHN: Implications of cytochrome P450 genetic polymorphisms on the toxicity of antitumor agents. Ther Drug Monit 26:236-240, 2004
Chiosis G, Huerzo H, Rosen N, et al: 17AAG: Low target binding affinity and potent cell activity-finding an explanation. Mol Cancer Ther 2:123-129, 2003
Egorin MJ, Lagattuta TF, Hamburger DR, et al: Pharmacokinetics, tissue distribution, and metabolism of 17-(dimethylaminoethylamino)-17-demethoxygeldanamycin (NSC 707545) in CD2F1 mice and Fischer 344 rats. Cancer Chemother Pharmacol 49:7-19, 2002(Jean L. Grem, Geraldine M)