Clinical Cancer Research Vol. 8, 706-711, March 2002
© 2002 American Association for Cancer Research
Phase I Clinical and Pharmacokinetic Study of Oral Penclomedine (NSC 338720) in Adults with Advanced Solid Malignancy
Glenn Liu,
Jordan Berlin,
Kendra D. Tutsch,
Lynn Van Ummersen,
Amy Dresen,
Rebecca Marnocha,
Rhoda Arzomanian,
Dona Alberti,
Chris Feierabend,
Kimberly Binger and
George Wilding1
University of Wisconsin Comprehensive Cancer Center, Madison, Wisconsin 53792
 |
ABSTRACT
|
|---|
Penclomedine is a synthetic 
-picoline derivative that has shown antitumor activity both in preclinical development and in Phase I work using an i.v. preparation. The main toxicities seen in those studies were dose dependent and mainly neurocerebellar, with hematological toxicity being far less severe. This Phase I trial of p.o. penclomedine was conducted to potentially alter the toxicity profile and to avoid the neurological side effects seen with i.v. penclomedine. Eligibility criteria included microscopic confirmation of a solid malignancy or lymphoma with a lack of effective anticancer therapy. Twenty patients were enrolled. The median age was 60.5 years, and the median performance status was one. All but one patient had received prior systemic therapy. The starting dose of penclomedine was 200 mg/m2 p.o. for 5 days, and was escalated according to a traditional Fibonacci sequence until the maximum tolerated dose (MTD) was observed. No treatment-related deaths were observed during the study. The MTD was determined to be 800 mg/m2 p.o. for 5 days. Dose-limiting toxicities included mainly neurocerebellar symptoms such as ataxia and dysmetria, but neurocortical symptoms, such as confusion, were seen as well. Myelosuppression was less common and resulted in the discontinuation of therapy in only two patients. Pharmacokinetics show that the observed MTD is consistent with the i.v. preparations, and that the bioavailability of p.o. penclomedine is 49 ± 18%. This regimen can be considered for additional studies in patients with intracranial neoplasms, because good central nervous system penetration is evident. Further development of penclomedine metabolites, such as 4-O-demethylpenclomedine, should be considered to minimize dose-limiting neurotoxicity.
|
INTRODUCTION
|
|---|
Penclomedine (NSC-338720) was first identified by the NCI2
as a potential anticancer agent in 1981 using the P388 leukemia prescreen. This -picoline derivative [3,5-dichloro-2,4-dimethoxy-6-(trichloromethyl)pyridine] has since been found to have pronounced activity against CDF81 murine and MCF-7 human mammary adnenocarcinomas (1)
, as well as intracerebrally implanted human mammary adnenocarcinoma MX-1 in nude mice (2)
. Dow Chemicals initially synthesized penclomedine for use as a potential fungicide, but its antitumor activity has made it attractive for clinical studies in solid malignancies. Although the mechanism of action of penclomedine is unknown, P388 cell lines resistant to the alkylating agents melphalan, cyclophosphamide, and carmustine were also found resistant to penclomedine (2
, 3)
, which suggests that the mechanism of action was alkylation. In addition, DNA analysis in cells treated with penclomedine revealed chromosomal gaps, breaks, and exchanges supporting DNA interactions consistent with alkylating activity (3)
.
Preclinical studies suggest that penclomedine is metabolized by microsomal reduction into dm-pen. Because of the measured low bioavailability of p.o. penclomedine (although observed higher in in vivo than in in vitro activity), the compound dm-pen may be a prime mediator in the antitumor activity of penclomedine (4)
. Therefore, the low systemic bioavailability is likely attributable to presystemic conversion to reactive mediators and not attributable to poor gastrointestinal absorption. Subsequent bioavailability studies have identified fair levels of penclomedine in tissues after p.o. penclomedine was administered. This suggests that intact penclomedine is also absorbed but is then rapidly taken up by tissues (5)
.
Two parallel Phase I trials with i.v. penclomedine have already been completed, one at Johns Hopkins Oncology Center (6)
and the other here at the University of Wisconsin (7)
. The penclomedine was given i.v. daily for 5 days every 3 or 4 weeks. The MTD seen was 
410–415 mg/m2 with main toxicities being neurocerebellar (i.e., ataxia, dysmetria, vertigo, and confusion), with nausea/vomiting and myelosuppression being less severe. The observed neurocerebellar toxicities were found to be dose dependent because the unwanted effects appeared to be greatest just after the completion of treatment and resolved spontaneously within hours. Also, prolongation of the infusion time at Johns Hopkins resulted in a decreased severity of symptoms. A third Phase I trial with i.v. penclomedine has since been conducted in Europe. Again, neurocerebellar symptoms (dizziness and ataxia) were the dose-limiting toxicities. This was seen in all three of the three patients treated at 340 mg/m2 (8)
. Given this new information, all trials with i.v. penclomedine were stopped for safety issues. No objective responses were seen in the Hopkins or European trial. The University of Wisconsin trial did have one partial response in a woman with ovarian carcinoma.
Thus, this trial of p.o. penclomedine was designed to potentially improve the dose-limiting, dose-dependent neurocerebellar toxicity seen with i.v. penclomedine. Preclinical data in mice with p.o. penclomedine show equal efficacy when compared with i.v. or i.p. injections (2)
. This activity is present despite the low bioavailability (2%) of p.o. penclomedine in mice (4)
. The low plasma penclomedine levels seen with equivalent p.o./i.v. activity in vivo suggests either rapid first-pass metabolism or high tissue uptake. Likely, both mechanisms are true because extensive metabolites of penclomedine can be identified in the liver after injection, and high amounts of the intact penclomedine can be identified in fresh tumor samples and brain tissue (as one would expect from a lipophilic substance; Refs. 5
, 9
). Whether the metabolites of penclomedine or the parent compound itself is the active agent remains to be determined. On the basis of equally effective p.o. versus i.v. penclomedine administration in animal models and the observed dose-dependent neurotoxicity that is directly related to the serum concentration of intact penclomedine, the use of p.o. penclomedine deserves clinical evaluation. This Phase I trial was designed to determine the MTD of p.o. penclomedine. A concurrent bioavailability and pharmacokinetic study was also performed during this Phase I trial.
|
PATIENTS AND METHODS
|
|---|
Patients were eligible for this trial if they had histological evidence for a solid malignancy or lymphoma that was refractory to standard therapy or to which no known effective treatment exists. Age 
18 years with an Eastern Cooperative Oncology Group performance status of 0–2 and life expectancy greater than 12 weeks was required. At least 4 weeks must have elapsed since any prior therapy such as chemotherapy or radiation therapy (6 weeks if on mitomycin C or nitrosoureas), and patients needed to be fully recovered from any previous treatment-related toxicity. Adequate organ function was required and was defined as WBC, 4000/mm3; platelets, 100,000/mm3; absolute neutrophil count, 1500/mm3; as well as bilirubin, <1.5 mg/dl, and a normal serum creatinine. Patients with a history of brain metastasis or seizure disorder were excluded. Pregnant or nursing mothers were also not eligible for this study. A signed, written informed consent was obtained from all of the patients.
P.o. penclomedine was provided by the NCI, formulated in 100-mg opaque, soft-gelatin capsules that also contained capric/caprilic triglycerides. For the pilot bioavailability study, i.v. penclomedine was provided by the NCI in sterile 100-mg vials containing a 10-mg/ml emulsion. Each milliliter of this solution contained 100 mg of soybean oil, 30 mg of egg phospholipid, 20 mg of glycerin, and 10 mg of penclomedine. This solution was subsequently diluted in 5% dextrose as needed for infusion.
Trial Design.
The treatment schedule was p.o. daily for 5 days with morning dosing. Treatments were repeated every 4 weeks (28-day cycles). One patient was initially treated at 200 mg/m2/day, with subsequent doses defined according to a traditional modified Fibonacci sequence as follows: level 1, 400 mg/m2; level 2, 600 mg/m2; level 3, 800 mg/m2; level 4, 1067 mg/m2.
The study followed a standard design with patients enrolled in cohorts of three. On the basis of the number of patients who experienced a DLT, additional patients were added at the same or previous level. Once 2 of 6 patients experienced a DLT at a given level, the MTD was considered exceeded and the previous dose level was considered the MTD. Additional patients were then added at the MTD to further define toxicity at this dose.
Before initiation of therapy, a complete history and physical exam, with neurological exam, was obtained. Baseline tumor measurements were performed on all patients with measurable disease and at least every 4–8 weeks while on study. Blood counts were obtained weekly, with repeat chemistries (including liver function tests) performed biweekly.
All of the patients who completed at least one treatment course were considered evaluable. Patients were evaluated before initiation of each treatment cycle, with primary end points being the determination of the MTD as an appropriate Phase II dose for p.o. penclomedine. Clinical response was of secondary importance and analyzed only when available. Responses were characterized as a complete response if all clinical evidence of active tumor and symptoms disappeared for at least 4 weeks. A partial response was defined as a >50% decrease in the sum of the products of the perpendicular tumor diameters of all measurable lesions for 4 weeks. Stable disease was a response that was less than a partial response or that had a <25% increase in size of any measurable lesion for a minimum of 8 weeks. Progression was defined as 25% increase in the size of any measurable lesion or the appearance of any new lesions. A DLT was defined as any grade-4 hematological toxicity or any nonhematological toxicity of 3 according to the NCI Common Toxicity Criteria (except nausea and vomiting). Any grade 2 neurocerebellar event lasting more than 2 h was also considered a DLT in this study.
Pharmacokinetic Sampling and Methods.
The p.o. formulation of penclomedine and its major metabolite dm-pen were measured in plasma using a validated high-performance liquid chromatographic method. Intact penclomedine was assayed by a previously reported method (4
, 7)
and dm-pen by an assay reported by Hartman (5)
. Pure penclomedine for preparation of standards was provided by the Drug Synthesis and Chemistry Branch of the NCI (Bethesda, MD), and pure dm-pen was provided as a gift from Dr. Robert Struck at the Southern Research Institute (Birmingham, AL).
Patients received a single dose of p.o. penclomedine daily for 5 days every 4 weeks. During the initial p.o. cycle, plasma samples were collected on days 1 and 5 predose, and postdose at 15, 30, 60 min and 2, 3, 4, 6, 8, and 24 h. Additional samples were collected at 1 and 2 weeks after the 5-day dosing of penclomedine. These samples were assayed for both penclomedine and dm-pen. For the pilot bioavailability study, three patients at level 1 received a single 1-hour i.v. infusion of penclomedine followed by a 2-week washout period before starting the p.o. penclomedine. Again, plasma samples were collected both predose and postdose at the same times listed above. The sampling was again repeated during the p.o. penclomedine period for comparison to the single i.v. dose.
The PKs for botha penclomedine and dm-pen were examined by noncompartmental methods using PKAnalyst (MicroMath Research, Salt Lake City, UT). The AUC was calculated by linear trapezoidal rule. Penclomedine clearance was determined by dose/AUC. The maximum concentration and time-to-maximum concentration was determined by inspection of the data sets. P.o. bioavailability was determined by comparing the 0–24 h AUC of intact penclomedine for the day-1 p.o. dose with the 0–24-h AUC from the single i.v. dose.
|
RESULTS
|
|---|
Toxicity Data.
A total of 20 patients were enrolled onto this trial. Only one patient was later found ineligible with last chemotherapy <4 weeks prior, three others had incomplete data before drug initiation. Demographic information is listed in Table 1
. The total number of cycles given was 41 with the median number of treatment cycles was 2 (range, 1–6). Four patients were unevaluable, three because of early withdrawal from the study and one because of an insufficient dose received (received only 4 days of therapy). Table 2
lists the trial progression and DLTs observed. In general, neurocerebellar toxicity such as ataxia and dizziness were seen at all dose levels (Table 3)
. These symptoms were, however, brief (lasting <2 h after drug dosing) and mild (grade 1–2). Neurocortical toxicity was observed in some patients, manifesting mainly as transient (grade 1–2) confusion. Mild nausea and fatigue (grade 1–2) were also seen in some patients. Myelosuppression was also seen (grade 1–2) and prompted a discontinuation of therapy in only two patients (one patient had been heavily pretreated in the past with chemotherapy). See Table 4
for the main nonhematological toxicity observed.
One patient did develop grade-4 diarrhea with hypotension after only 1 day of drug. Telemetry at that time showed evidence of transient heart block, although follow-up Holter monitoring did not show any signs of arrhythmia or nodal irregularity. It is unknown whether this was drug related, but the patient was subsequently removed from the study and considered unevaluable.
Pharmacokinetic Analysis.
The small bioavailability study (three patients) that was performed showed extreme interpatient variability in measured serum intact penclomedine with the i.v. dosing and the p.o. dosing. The bioavailability of penclomedine given p.o. was AUC, 37 ± 26% (range, 5–87%) of that for the i.v. dose. The penclomedine Cmax (peak concentration) was 24 ± 15% (range, 12–38%) of that for the i.v. dose. Of note, the AUC for dm-pen after p.o. penclomedine administration was 48 ± 16% of the AUC of dm-pen after a single i.v. dose. Because dm-pen is a metabolite, it does not represent the true bioavailability of penclomedine. However, it does likely give us a good estimate the actual bioavailability of p.o. penclomedine.
In general, a dose-dependent AUC and plasma concentration of intact penclomedine was observed (see Fig. 1
and 2
), but with the interpatient variability, the correlation coefficient (r) was not high (day 1 r, 0.688; P < 0.001; day 5 r, 0.732; P = 0.008). During routine p.o. dosing, PKs were obtained for penclomedine and dm-pen on days 1 and 5. Peak concentrations of penclomedine occurred 2–4 h after dosing, whereas peak concentrations of dm-pen occurred 6–24 h after dosing. Some accumulation of intact penclomedine was seen from day 1 to day 5, but very significant accumulations of dm-pen occurred over the 5 days. PK parameters are shown in Tables 5
and 6
. Of interest, two patients had ascites that was measured for penclomedine and dm-pen on day 15. This showed ascites penclomedine/dm-pen concentration of 54 and 40% of the measured serum levels, respectively.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 1. AUC of penclomedine and dm-pen as a function of dose. The mean from the i.v. penclomedine MTD (410 mg/m2) is shown for comparison. At he highest dose of p.o. penclomedine studied (1067 mg/m2/day), the AUC approached the MTD seen in the i.v. preparation. D1, day 1; D5, day 5; Pen, penclomedine.
|
|
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 2. Peak concentration of penclomedine and dm-pen as a function of dose. The mean from the i.v. penclomedine MTD (410 mg/m2) is shown for comparison. Note that the peak concentration of penclomedine was below 10 µg/ml, the level associated with significant neurotoxicity in preclinical studies, in all patients. D1, day 1; D5, day 5; Pen, penclomedine; Cmax, peak concentration.
|
|
Response Data.
One patient with renal cell cancer had stable disease for 16 weeks, but ultimately progressed after six cycles of therapy. Another patient with colon cancer had stable disease for 8 weeks, but progressed after three cycles of therapy. All of the others progressed radiographically after two cycles of treatment except two patients, who were taken off-study for early clinical deterioration. No drug-related death occurred while on study.
|
DISCUSSION
|
|---|
Penclomedine is a synthetic -picoline derivative that has made its way to clinical trials because of the antitumor activity seen using the NCI screening program. Predominate activity was observed in breast cancer lines, with significant regression seen in a spontaneous murine mammary tumor CD8F1, as well as in a human mammary tumor MX-1. Because seeing activity in both CD8F1 and MX-1 cell-lines is an uncommon occurrence in these drug screening trials, specific activity for breast cancer is suggested. Also, the data showing regression of intracerebrally implanted MX-1 implies good CNS penetration. Responses in other tumors were seen as well, such as the P388 leukemia and M5076 sarcoma. On the basis of these preclinical studies, penclomedine was taken to early trials.
Central nervous system toxicity is the main difficulty with i.v. penclomedine therapy as evident in the three Phase I studies previously conducted. The two parallel Phase I studies, performed at Johns Hopkins and at our own center, suggest that the MTD of i.v. penclomedine is around 410–415 mg/m2 with neurocerebellar toxicity being dose dependent. Animal studies performed show that p.o. penclomedine has a clinical activity similar to i.v. penclomedine (2)
and that low plasma levels of intact penclomedine are seen when compared with the i.v. dosing (4)
. Thus, p.o. penclomedine was thought to be a feasible method to minimize neurotoxicity while maintaining efficacy.
We have shown in this Phase I study of p.o. penclomedine, that the MTD is between 800 mg/m2 and 1067 mg/m2 when administered daily for 5 days in 28 day cycles. This corresponds remarkably well to our Phase I results with i.v. penclomedine because at the highest dose-level studied (1067 mg/m2), the AUC of intact penclomedine approached that seen at the MTD level of the i.v. formulation (410 mg/m2). When comparing the AUC of p.o. penclomedine in this study to the predicted AUC of an equivalent i.v. dose from our previous study, we estimate the bioavailability of p.o. penclomedine to be 49 ± 18%. This estimate is in agreement with the results we obtained (37 ± 26%) from the small pilot bioavailability study that was conducted.
In the original toxicology studies, dogs with serum penclomedine levels between 6 and 9 µg/ml did not show significant neurotoxicity; however, when serum levels reached 10 µg/ml, seizure activity was seen that resolved rapidly on discontinuation of the drug (10, 11
). Early clinical studies with i.v. penclomedine did not show any grade 3 or higher neurotoxicity in patients again until the peak plasma concentration of penclomedine exceeded 10 µg/ml (7)
. During this study, all patients had a maximum plasma concentration of intact penclomedine that was <10 µg/ml. Not surprisingly, no grade 3 or higher neurotoxicity was observed. In general, the clinical experience with i.v. penclomedine showed that the neurological toxicity occurred soon after dosing and resolved rapidly within 2 h. The same held true for p.o. penclomedine because neurocerebellar symptoms occurred usually after dosing and resolved quickly. Our pharmacokinetic data suggest that penclomedine levels, not dm-pen, is responsible for the neurotoxic effects. This is supported by the fact that plasma dm-pen levels peaked late on day 5, after all neurotoxicity had been resolved.
We have shown that the defined MTD of p.o. penclomedine is 800 mg/m2 daily for 5ays, repeated every 28 days, and that the observed toxicity is mainly neurological as one would expect. Although no responses were seen in this small study, early preclinical data as well as in vitro studies are promising. Although instances of neurocerebellar toxicity were still observed, these were milder (grade 1–2) when compared with the effects seen with i.v. penclomedine at comparable doses. Given the more favorable toxicity profile, it would seem reasonable to pursue a Phase II evaluation of p.o. penclomedine in patients with intracranial neoplasms, because central nervous system penetration appears unusually good. Also, further evaluation of the metabolite dm-pen is warranted, because this agent is likely still a precursor to the active component of the drug. We would hope that the testing of this metabolite will avoid the dose-limiting neurotoxicity observed with the parent compound penclomedine.
|
FOOTNOTES
|
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 To whom requests for reprints should be addressed, at Department of Medicine, Medical Oncology Section, K6/5 CSC, 600 Highland Avenue, Madison, Wisconsin 53792. 
2 The abbreviations used are: NCI, National Cancer Institute; MTD, maximum tolerated dose; DLT, dose-limiting toxicity; PK, pharmacokinetic parameter; AUC, area under the curve; dm-pen, 4-O-demethylpenclomedine.
Received 6/15/01;
revised 11/20/01;
accepted 12/13/01.
|
REFERENCES
|
|---|
-
Plowman J., Harrison S. D., Jr., Dykes D. J., Paull K. D., Naranan V., Tobol H. K., Griswold D. P. Preclinical antitumor activity of an -picoline derivative, penclomedine (NSC-338720), on human and murine tumors. Cancer Res., 49: 1909-1915, 1989.[Abstract/Free Full Text]
-
Harrison S. D., Jr., Plowman J., Dykes D. J., Waud W. R., Griswold D. P. Preclinical antitumor activity of penclomedine in mice: cross-resistance, schedule dependence, and oral activity against tumor xenografts in brain. Cancer Res., 51: 1979-1983, 1991.[Abstract/Free Full Text]
-
Benvenuto J. A., Hittelman W. N., Zwelling L. A., Plunkett W., Pandita T. K., Farquhar D., Newman R. A. Biochemical pharmacology of penclomedine (NSC-338720). Biochem. Pharmacol., 50: 1157-1164, 1995.[CrossRef][Medline]
-
Reid J. M., Mathiesen D. A., Benson L. M., Kuffel M. J., Ames M. M. Murine pharmacokinetics and metabolism of penclomedine [3,5-dichloro-2,4-dimethyl-6-(trichloromethyl)pyridine, NSC 338720]. Cancer Res., 52: 2830-2834, 1992.[Abstract/Free Full Text]
-
Hartman N. R., O’Reilly S., Rowinsky E. K., Collins J. M., Strong J. M. Murine and human in vivo penclomedine metabolism. Clin. Cancer Res., 2: 953-962, 1996.[Abstract]
-
O’Reilly S., Grochow L. B., Donehower R. C., Chen T. L., Bowling K., Hartman N. R., Struck R. F., Rowinsky E. K. Phase I and pharmacologic study of penclomedine, a novel alkylating agent, in patients with solid tumors. J. Clin. Oncol., 15: 1874-1984, 1977.
-
Berlin J., Stewart J. A., Storer B., Tutsch K. D., Arzoomanian R. Z., Alberti D., Feierabend C., Simon K., Wilding G. Phase I clinical and pharmacokinetic trial of penclomedine using a novel, two-stage trial design for patients with advanced malignancy. J. Clin. Oncol., 16: 1142-1149, 1998.[Abstract]
-
Jodrell D. I., Bowman A., Stewart M., Dunlop N., French R., MacLellan A., Cummings J., Smyth J. F. Dose-limiting neurotoxicity in a Phase I study of penclomedine (NSC 388720, CRC 88-04), a synthetic -picoline derivative, administered intravenously. Br. J. Cancer, 77: 808-811, 1998.[Medline]
-
O’Reilly S., Hartman N. R., Grossman S. A., Stromg J. M., Struck R. F., Eller S., Lesser G. J., Donehower R. C., Rowinsky E. K. Tissue and tumor distribution of 14C-penclomedine in rats. Clin. Cancer Res., 2: 541-548, 1996.[Abstract]
-
Dixit R., Lopez R., Douglas T., Muellner P., Litle L., Stolz M., Stedham M., Smith A., Tomaszewski J. E. Toxicity of five-hour intravenous infusion of penclomedine (PEN, NSC-338720) in beagle dogs. Proc. Am. Assoc. Cancer Res., 33: 548 1992.
-
National Cancer Institute. . Investigator’s Brochure: Penclomedine (NSC-338720), Bethesda, MD National Cancer Institute 1992.
Top
ABSTRACT
INTRODUCTION
