A Phase I Pharmacokinetic and Pharmacodynamic Study of PX-12, a Novel Inhibitor of Thioredoxin-1, in Patients with Advanced Solid Tumors

Purpose: Thioredoxin-1 (Trx-1) is a cellular redox protein that promotes tumor growth, inhibits apoptosis, and up-regulates hypoxia-inducible factor-1α and vascular endothelial growth factor. Objectives of this study were to determine safety, tolerability, pharmacodynamics, and pharmacokinetics of PX-12, a small-molecule inhibitor of Trx-1.

Experimental Design: Thirty-eight patients with advanced solid tumors received PX-12 at doses of 9 to 300 mg/m², as a 1- or 3-h i.v. infusion on days 1 to 5, repeated every 3 weeks.

Results: At the 300 mg/m² dose level, one patient experienced a reversible episode of pneumonitis during the first cycle, and a second patient developed pneumonitis after the second cycle. Doses up to 226 mg/m² were well tolerated, and grade 3/4 events were uncommon (<3% of patients). The limiting factor on this dosing schedule was pungent odor caused by expired drug metabolite, 2-butanethiol. The best response was stable disease in seven patients (126-332 days). Whereas PX-12 was not detectable following the infusion, the C_max of its inactive metabolite, 2-mercaptoimidazole, increased linearly with dose. PX-12 treatment lowered plasma Trx-1 concentrations in a dose-dependent manner.

Conclusions: PX-12, the first Trx-1 inhibitor to enter clinical trials, was tolerated up to a dose of 226 mg/m² by a 3-h infusion. Based on pharmacodynamic and pharmacokinetic data, a trial of prolonged infusion schedule of PX-12 has been initiated.

Thioredoxin-1 (Trx-1) is a low-molecular weight (10-12 kDa) redox protein overexpressed in tumors and associated with growth stimulation, antiapoptosis, and angiogenesis (1–12). Trx-1–transfected cells are resistant to doxorubicin, vincristine, cisplatin, and cytosine arabinoside (8, 13, 14). Retrospective analyses in colorectal cancer and non–small cell lung cancer have shown that Trx-1 overexpression may be an independent prognostic factor (6, 15).

PX-12, 1-methylpropyl 2-imidazolyl disulfide, has shown both excellent in vitro and promising in vivo antitumor activity (16), irreversibly thio-alkylating the Cys⁷³ thioredoxin residue. PX-12 inhibition of Trx-1 results in subsequent inhibition of the hypoxia-induced increase in HIF-1α protein and VEGF secretion (17). PX-12 is rapidly metabolized, providing two inactive metabolites, volatile 2-butanethiol and 2-mercaptoimidazole (2-MI; Fig. 1). Rats showed the greatest sensitivity to PX-12; the severe toxic dose was 90 mg/m² daily × 5 with local phlebitis as the major toxicity (18).

Here, we report the first-in-human phase I trial of PX-12 in patients with advanced solid tumors. The starting dose of PX-12 was 9 mg/m² daily × 5 repeated every 3 weeks, which was 1/10 of the severe toxic dose in rats. The primary objectives were to determine the maximally tolerated dose and safety of PX-12. The secondary objectives were to assess the pharmacokinetics, pharmacodynamics, and efficacy of PX-12.

Patient selection. Eligible patients were over 18 years of age and had histologically confirmed, advanced cancer refractory to standard therapies. Other requirements are the following: Eastern Cooperative Oncology Group performance status 0 to 2 and no chemotherapy or radiation at least 4 weeks prior. Adequate organ function was defined by WBC >3,000/μL, platelets >100,000/μL, hemoglobin >9 g/dL, absolute neutrophil count >1500/μL, bilirubin <2.0 mg/dL, serum creatinine <2.0 mg/dL, and aspartate transaminase <2.0× upper limit of normal.

Exclusion criteria included the following: untreated brain metastasis, active infection, major surgery within 4 weeks, and any serious concomitant conditions that would place the patient at excessive or unacceptable risk of toxicity. Before entering the study, each patient gave written informed consent, according to institutional and federal guidelines. This study was conducted at the University of Pittsburgh Cancer Institute and the Arizona Cancer Center (Tucson, AZ).

Drug administration. PX-12 was supplied in sterile 2-mL vials containing 36.07 mg/mL PX-12 in polyethylene glycol. The appropriate amount of PX-12 was added to supplied diluent (0.013 N HCl) to prepare PX-12 solution (2.5 mg/mL) and made to final volume (250 mL) by adding 5% dextrose. Because of local phlebitis in animal studies, PX-12 was administered via central venous catheter over 1 h (16 patients) or 3 h (22 patients), with a prophylactic dose of warfarin (1 mg/d).

Dose escalation. The starting dose of PX-12 was 9 mg/m² over 1 h, daily × 5, repeated every 21 days. An accelerated dose titration schedule with three patients per cohort was used initially. In the absence of grade 2 toxicity, subsequent dose levels were escalated by 100%, up to a dose of 72 mg/m². The National Cancer Institute-Common Toxicity Criteria version 2.0 was used for toxicity monitoring. Dose-limiting toxicity was defined as grade 4 hematologic or drug-related grade 3, or greater, nonhematologic toxicity reported in the first cycle. The dose level at which more than one of six patients experienced a dose-limiting toxicity was considered the maximally tolerated dose. Intrapatient dose escalation was allowed but only if a higher dose level cohort had been completed without dose-limiting toxicities.

In the first cohort, one of three patients developed grade 2 dermatitis (groin rash). Subsequently, the next patient was accrued at 12 mg/m² (33% escalation), but further review determined that the rash was not drug related, and 100% dose escalation resumed to 18, 36, and 72 mg/m². At the 96 mg/m² cohort, due to intense cough and pungent odor of the expired breath, the protocol was modified to prolong the infusion to 3 h. On this schedule (3-h infusion), dosing resumed at 54 mg/m² and was escalated by 33% increments to 72, 96, 128, 170, 226, and 300 mg/m².

Study requirements and assessments. A history and physical examination were completed prestudy and before every cycle. A complete blood count, electrolytes, chemistries, and urine analysis were obtained prestudy and then weekly. Imaging studies for tumor measurement were obtained at baseline and after every two cycles. The Response Evaluation Criteria in Solid Tumors was used for tumor measurements (19).

Dose modifications. Patients were required to meet prestudy requirements before every cycle, and/or drug-related toxicity had to resolve to <grade 1 before resuming therapy. In the event of a dose-limiting toxicity, continuation of therapy was allowed if toxicity had improved to ≤grade 1 within 2 weeks, but with dose reduction to the previous lower dose level, or at 50% of the dose if the dose-limiting toxicity occurred at the dose level 1. Patients who could not resume therapy within 2 weeks were removed from the study.

Pharmacokinetic methods. During cycle 1, on days 1 and 5, 7 mL blood samples were collected into EDTA tube before PX-12 administration, at 15 min into the infusion, at the end of infusion, and at 2, 5, 10, 15, 30, 60, 90, 120, and 240 min following completion of infusion. On days 2, 3, and 4, blood samples were collected before dosing and at end of infusion (up to cycle 4). On weekly follow-up visits (days 8 and 15), additional samples were collected. Samples were centrifuged at room temperature for 5 min at 1,500 × g, and the resulting plasma was transferred to 15-mL polypropylene tubes and centrifuged for 5 min at 2,500 × g. The supernatant was aliquoted into prelabeled, 1.5-mL tubes and stored frozen at −70°C until assayed.

For the analysis of 2-MI, 200 μL plasma was mixed with 1 mL acetonitrile. The mixture was vortexed for 5 min and centrifuged at 15,000 × g at room temperature for 5 min, and the supernatant (1 mL) was transferred to a 1.5-mL polypropylene tube and dried using a centrifugal evaporator. The residues were reconstituted in 40 μL isopropanol/methanol (32:68, v/v), sonicated until dissolved, vortexed for 5 min, and, following the addition of 85 μL hexane, centrifuged at 15,000 × g for 5 min. The resulting supernatant was analyzed by high-performance liquid chromatography using a CyanoNova-Pack column (3.9 × 150 mm, 5 μm) with a cyanoguard column (3.9 × 20 mm, 5 μm; Waters, Milford, MA), column temperature of 30°C, and an isocratic mobile phase of isopropanol/methanol/hexane (4:8.5:87.5, v/v/v) at a flow rate of 1 mL/min. 2-MI was detected at 254 nm and quantitated versus standard curve (20). The 2-MI concentration versus time curves were analyzed by the noncompartment approach using WinNonLin (version 4.0.1; Pharsight Corp., Mountain View, CA). The following variables were obtained: area under the plasma concentration-time profile up to 240 min post–PX-12 infusion, maximum plasma concentration (C_max), and time to reach maximum plasma concentration (T_max).

Concentrations of 2-butanethiol in the expired air of two patients were measured, during, at the end of infusion, and 30 min following the end of infusion using collision rate-mass spectrometry (21). Breath samples were collected in 10 L Tedlar bags (SKC, Inc., Eighty Four, PA) and compared with normal control samples.

Pharmacodynamic methods. Plasma samples were analyzed for Trx-1 concentrations using an ELISA. Nunc Immunolon 2 HB microtiter plates (Cole-Parmer, Vernan Hills, IL) were coated with mouse anti-human Trx-1 antibody [clone 2B1; 100 μL of 1 μg/mL in 50 mmol/L sodium carbonate buffer (pH 9.6), Serotec, Raleigh, NC] at 4°C overnight followed by blocking with blocking buffer [BLB; 100 μL of 3% bovine serum albumin in PBS, 0.05% Tween 20] at 4°C overnight. Plates were used immediately or stored at −20°C for up to 7 days. Standards (100 μL of purified human Trx-1) or patient plasma samples [1:2 or 1:4 dilutions in 50 mmol/L sodium carbonate buffer (pH 9.6)] were added in triplicate, incubated for 2 h at room temperature, and washed with PBS-Tween 20. Chicken anti-human PX-12hTrx antibody (100 μL of a 1:500 dilution in BLB; ProlX Pharmaceuticals, Tucson, AZ) was added, incubated for 2 h at room temperature followed by addition of 100 μL of a 1:2,500 dilution of goat anti-chicken IgY–horseradish peroxidase antibody (AVES Labs, Tigard, OR), and incubated for 1 h at room temperature. After washing, 100 μL substrate (1-Step Ultra TMB-ELISA, Pierce, Rockford, IL) was added. Absorbance was read at 650 nm at 5, 15, and 30 min after the addition of substrate, and Trx-1levels were quantitated using standard curve from each plate.

Statistical analyses of pharmacodynamic data. Plasma Trx-1 analyses were limited only to nonhemolyzed blood samples. Therefore, number of samples at different time points ranged from 28 to 36. Geometric means were used to describe the ratio of Trx-1 measurements at subsequent time points compared with the initial measurement (cycle 1 pretreatment). The statistical significance was based on a one-sample t test of whether the logarithm of the ratio differed from zero (i.e., whether the ratio differed from 1). Linear regression was used to investigate the relationship of the logarithm of the ratio to dose level. Survival was estimated using Kaplan-Meier survival curves, and subgroup survival curves were compared using log-rank tests.

Patient characteristics. Patient characteristics are provided in Table 1. A total of 38 patients were treated at dose levels ranging from 9 to 300 mg/m² and received a total of 142 cycles (median, 2; range, 1-14; Table 2). Predominant tumor type was colorectal cancer (n = 18) followed by sarcoma (n = 5), lung (n = 4), and pancreas (n = 3). In two patients, the PX-12 was dose escalated, from 9 to 36 mg/m² and from 18 to 36 mg/m². Thirty-seven patients were evaluable for response.

Safety. At the dose of 300 mg/m², one patient developed pneumonitis on day 5 of cycle 1. This manifested as a syndrome of fever (102°F), cough, hypoxia, and bilateral interstitial infiltrates. The patient was hospitalized and treated with high-dose steroids and oxygen. The symptoms improved over 48 h, and the infiltrates were resolved. The second of three patients treated with PX-12 at 300 mg/m² developed shortness of breath after his second cycle, and computed tomography scan revealed bilateral pneumonitis. Due to the severity of the respiratory toxicity seen at this dose level, although one event occurred in the second cycle, further expansion of this dose cohort was aborted (Tables 2 and 3). Most patients (86.1%) experienced a mild cough (grade 1) that occurred within a few minutes of PX-12 infusion and lasted for 10 to 15 min. The dry cough, most noticeable on day 1 and subsiding on subsequent treatment days, increased in intensity at doses >96 mg/m²/h. Although PX-12 is odorless, a garlic-like, pungent odor developed minutes into the infusion of PX-12, with intensity of the odor increasing with doses >96 mg/m², so that air purifiers and good ventilation were required for administration. The odor is attributed to the expired PX-12 metabolite 2-butanethiol, a known irritant, which is a component of food flavorings (list of designated additives at The Japan Food Chemical Research Foundation⁵

), fragrances,⁶ and an odorant in natural gas (MSDS M36053 Captan 90). In this study, the highest concentration of 2-butanethiol in breath was determined to be 0.34 ppm. This level is well below toxic levels reported for related volatile organic compounds; in mice and rats, the LD₅₀ ranges from 4,000 to 16,000 ppm/4 h. It is however possible that the hypersensitivity or an idiosyncratic reaction to 2-buthanethiol was responsible for pneumonitis seen in two patients.

One patient developed a non-Q wave myocardial infarction after his second cycle of therapy (170 mg/m² dose), and one patient treated at the dose of 226 mg/m² expired due to disease progression 26 days after the last dose of PX-12.

More common low-grade events (grade 1 or 2) included fever, fatigue, nausea, and vomiting (Table 3). Routine antiemetics were not given initially, but at doses ≥96 mg/m², prophylactic therapy with prochlorperazine was recommended.

Efficacy. There were no objective responses based on Response Evaluation Criteria in Solid Tumors. Seven (18%) patients achieved durable stable disease, receiving at least six cycles of therapy (range, 6-14 cycles). These included two patients with sarcoma and one patient each with colorectal, appendiceal, and renal cancer. Of the seven patients with SD, three were treated between 9 and 36 mg/m², three received 72 or 96 mg/m², and one received 226 mg/m². A patient with appendiceal adenocarcinoma had a minor response, 18.3% by Response Evaluation Criteria in Solid Tumors, and remained on study for 14 cycles (322 days; Table 2).

Pharmacokinetics. PX-12 was not detectable in patient plasma at any dose level. However, the PX-12 metabolite, 2-MI, was detected in the sample collected 15 min into infusion. The detection limit of PX-12 from plasma is ∼100 μmol/L with the UV detection, whereas the detection limit for 2-MI is ∼5 μmol/L. Figure 2 shows the average plasma concentration versus time profiles of 2-MI obtained on day 1 at cycle 1 in patients treated with 3-h infusions of PX-12. 2-MI concentrations reached a maximum within 15 min after the PX-12 infusion was terminated but seemed to increase again 30 to 120 min after infusion. Following this, 2-MI levels declined as a function of time with a return to baseline at 24 h before subsequent dosing.

A 1-h infusion of PX-12 (day 1 at cycle 1) produced a C_max and area under the plasma concentration-time profile of 2-MI, which gradually increased as the PX-12 dose was increased from 9 to 72 mg/m². No further increase in 2-MI was seen between 72 and 96 mg/m² (Fig. 3). In contrast, with the 3-h infusion (day 1 at cycle 1), both C_max and area under the plasma concentration-time profile of 2-MI increased linearly as PX-12 dose was increased up to 300 mg/m² (Fig. 4). There was no accumulation of 2-MI over the 5 consecutive treatment days (data not shown).

Pharmacodynamic results.Table 4 shows the range and mean of Trx-1 concentrations (ng/mL) for all evaluable plasma samples at time points during cycle 1 on day 1 at the end of the infusion cycle 1 on day 5 and for the pretreatment sample at the beginning of cycle 2 (on day 22). Data available for patients receiving the same dose of PX-12 (72 or 96 mg/m²) over either 1 h (three patients per dose) or 3 h (two and three patients) indicated that the 3-h infusion lowered the Trx-1 levels by the end of the infusion, from 81.8 ng/mL at baseline to 54.6 ng/mL (group averages). Mean plasma Trx-1 concentration decreased from cycle 1, pretreatment at each of the five time points evaluated, end of the infusion on day 1, 60 min from end of infusion, C1 minimum (lowest Trx-1 concentration at the end of infusion, at 60 or 240 min) end of infusion day 5, and before dosing day 1 at cycle 2 (ratio, <1). Of the five time points evaluated, only the ratio of the cycle 2 pretreatment thioredoxin concentration compared with the cycle 1 pretreatment concentration was significantly related to dose level (P = 0.05), with the ratio tending to decrease with increasing dose level, although there was considerable scatter in the data points (data not shown). Exploratory analyses showed that a 25% or greater decrease in plasma Trx-1 levels during the first cycle (predose cycle 1 to predose cycle 2; each patient used as their own control) correlated with overall survival (P = 0.015) but not with the time to progression (P = 0.72; data not shown).

PX-12, the first Trx-1 inhibitor to enter phase I trials, was studied to identify the optimal schedule/dosing and safety, with pharmacokinetic and exploratory pharmacodynamic evaluations.

Dose-limiting respiratory toxicity clinically manifesting as pneumonitis was observed at a dose of 300 mg/m². Although reversible, further dose escalation was aborted. The cough and pungent odor were more intense when PX-12 was given over 1 h and increased with dose. For that reason, the infusion schedule was prolonged to 3 h, and carbon air purifiers were used with closed ventilation to reduce the odor at dose levels ≥96 mg/m².

These effects were presumably due to the metabolite 2-butanethiol and were seen in dogs at a similar dose (18). The other plausible explanations are that the thioredoxin has a specific target protein (perhaps one of the bioactive peptides or enzymes that metabolize them in the lung tissue) that may be overexpressed in airways or that there is a higher delivery/uptake of the PX-12 in the lung tissue leading to airway/lung specific toxicity. Treatment with PX-12 lowered serum Trx-1, and this effect lasted up to day 21. Those patients achieving >25% decrease in serum Trx-1 seemed to show prolonged survival; however, these analyses should be interpreted as exploratory, as the dose levels were not randomized, and survival may have been influenced by patient selection, variable patient characteristics, and the small sample size. This observed decrease in plasma Trx-1 concentrations with PX-12 treatment may be a useful biomarker, but further validation is warranted. An earlier preliminary study using plasma samples from mice and also from a small subset of patient showed that exploratory surface-enhanced laser/desorption/ionization (SELDI) time-of-flight mass spectrometry provides similar results, in terms of detecting change in plasma Trx-1 concentration pre– and post–PX-12 treatment (22). However, it was decided to use ELISA to analyze the full set of patient data from this study due to its robustness and reproducibility, which was difficult to achieve with SELDI technology, in our experience.

No objective tumor responses were seen, but seven (18%) patients achieved lasting stable disease, and one patient had a minor response.

Intact PX-12 was not detectable in any pharmacokinetic samples, possibly due to its rapid distribution, its lipophilic properties, and/or rapid thiol-disulfide exchange with reduced cysteine residues, including those on circulating Trx-1. In addition, the extraction recovery of PX-12 from plasma is only 2% to 5%, which is probably due to thiol-disulfide exchange with reduced thiols in plasma or irreversible binding to plasma components, making it difficult to detect. The pharmacokinetic data for 2-MI suggested that PX-12 infused over 1 h may have rapidly saturated the circulating reduced thiols at doses as low as 72 mg/m², which limited its subsequent formation. In contrast, PX-12 infused over 3 h did not seem to saturate thiols up to a total dose of 300 mg/m². 2-MI levels increased linearly with 3-h level infusion but not 1-h infusion. With the 3-h infusion regimen, plasma peak concentrations of 2-MI and total systemic exposure seem to increase proportionally as the dose is increased. This may account for the more pronounced decrease in circulating Trx-1 observed when PX-12 was delivered over 3 versus 1 h.

Although the PX-12 was well tolerated at doses up to 226 mg/m²/3 h, this dose/schedule may not be practical because of the need for negative pressure treatment area. In addition, disease stabilization was seen at doses of 9 to 36 mg/m², indicating that PX-12 may be active at doses lower then the maximally tolerated dose. Posttreatment decrease in circulating Trx-1 seemed more pronounced after 3-h infusion, suggesting that prolongation of the PX-12 infusion time (12-24 h) may provide additional therapeutic benefit and improve tolerability. A trial using a 24-h infusion of PX-12 every 14 days has been initiated.

We thank Achyt Bhattacharrya M.D. and Ryan Williams B.S. (Arizona Cancer Center) for their assistance with immunocytochemistry assays and Joe Grabowski Ph.D. and Mark Morris M.Sc. (University of Pittsburgh) for their assistance with the breath metabolite analyses.