Accumulation of lovastatin, but not pravastatin, in the blood of cyclosporine-treated kidney graft patients after multiple doses

ol+ctim To study pravastatin and lovastatin pharmacoldnetic and p-c e&cts and their inter- actions with cyclosporine (INN, cickqorin) in kidney transplant patients after single and multiple doses. Su,b+& and tactbo~: The pharmacokinetic and pharmacodynamic effects of administration of 20 mg/day oral pravastatin and lovastatin for 28 days and their interactions with cyclosporine (2 to 6 mg/kg/day) were studied in a double-blind, double-dummy, randomized, parallel-group multicenter trial in 44 stable kidney graft recipients. Rcsnlts: The median area under the curve [AUC(O-24)] of pravastatin was 249 pg - hr/L (range, 104 to 1026 pg * hr/L) after a single dose (day 1) and 241 pg * hr/L (114 to 969 pg - hr/L) after multiple doses (day 28) and was fivefold higher than values reported in the absence of cyclosporine. The median AUC(O-24) of lovastatin was 243 pg. hr/L (105 to 858 pg. hr/L) on day 1 and 459 pg * hrfi (140 to 1508 pg * hrfi) on day 28. Beaides a sign&ant accumulation during the study period (p c O.OOl), the lovastatin AUC(O-24) vahtes were twentyfold higher than values qxxted without cyclosporine. Coadmin- istration of pravastatin or lovastatin did not alter cyclosporine pharmacokinetics. In this study, 20 mg/day doses of both drugs resulted in a sign&ant improvement of the lipid profile and were well tolerated. Conclwsions: ln contrast to lovastatin, pravastaGn did not accumulate over the study period, which is probably one of the reasons rhabdomyolysis has been reported in lovastatht-treated but not pravastatin-treated trans- plant patients recehkg cyclospo&e immunosuppression. (Clin Pharmacol Ther 1997;62:311-21.)

Christoph Olbricht, MD, PhD, Christoph Wanner, MD, PhD, Thomas Eisenhauer, MD, PhD, Volker Kliem, MD, Rainer Doll, MD, Michel Boddaert, MSc, Padraig O’Grady, PhD, Michael Krekler, PiiD, Bernhard Mangold, MD, PliD, and Uwe Christians, MD, PhD I4iwwwvcr, W&bug, K0blen.q and Munich, Germany, Watzrloo, Be&m, and San Franc&a, calif:

Elevated low-density lipoprotein (LDL) choles- olemia after transplantation, besides cyclosporine terol is often present in kidney graft patients treated (INN, ciclosporin) administration, include age, male with cyclosporine, and cardiovascular complications sex, diabetes, graft impairment, obesity, antihyper- are the most frequent cause of death in this group of tensive therapy, and corticosteroids.‘*2 3-Hydroxy-3- patients. The major risk factors for hypercholester- methylglutaryl coenzyme A (HMG-CoA) reductase

From the Abteilung Nephrologie, Medizinische Hochschule Hannover, Harmover; the Abteihmg Nephrologie, Mediz- inische Universitatsklinik, Wtirzburg; the Medizinische Klinik II, Stiidtisches Krankenhaus Kemperhof, Koblenz; the Depart- ment of Clinical Pharmacology and the Department of Biosta- tistics and Data Management, Bristol-Myers Squibb Pharma- ceutical Research Institute, Waterloo; Cardiovascular Research, Bristol-Myers Squibb, Munich; and the Department of Biopharmaceutical Sciences, School of Pharmacy, Univer- sity of California at San Francisco, San Francisco.

Supported in part by Bristol-Myers Squibb and the Deutsche Forschungsgemeinschaft grants SFB265/A7 and Ch95/6-1.

Presented in part at the Annual Meeting of the American Society of Nephrology, San Diego, Calif., November 5-8, 1995, and New Trends in Clinical and Experimental Immunosuppression, Geneva, Switzerland, February 15-18,1996.

Received for publication Feb. 21, 1997; accepted June 4, 1997.

Reprint requests: Uwe Christians, MD, PhD, Department of Biopharmaceutical Sciences, School of Pharmacy, University of California at San Francisco, 513 Pamassus Ave., S-834, San Francisco, CA 94143-0446. E-mail: uwec@itsa.ucsf.edu

Copyright 0 1997 by Mosby-Year Book, Inc. 0009-9236/97/$5.00 + 0 13/l/83708

311

3 12 Olbricht et al. CLINICAL PHARMA COLOGY & THERAPEUTICS

SEPTEMBER 1997

inhibitors such as pravastatin, lovastatin, and simvasta- tin have been shown to effectively lower cholesterol serum concentrations in kidney graft recipients receiv- ing long-term cyclosporine immunosuppression3‘7 and are therefore the drugs of choice in kidney transplant patients.

Lovastatin is mainly metabolized by enzymes of the cytochrome P450 (CYP) 3A subfamily.’ The CYP enzymes in the liver responsible for pravastatin metabolism have not yet been identified. Several drugs known to be CYP3A substrates, inducers, and inhibitors also affect or are affected by cyclosporine metabolism.’ Increased plasma concentrations of lo- vastatin have been reported in transplant patients receiving concomitant cyclosporine. These in- creased plasma concentrations have been associated with acute and severe rhabdomyolysis and kidney failure.““’ Altered disposition of pravastatin in heart transplant patients receiving concomitant cy- closporine has also been reported.12 In this study, an increase in pravastatin area under the curve (AUC) suggested a possible interaction with cyclosporine as a result of reduced plasma clearance or a diminished first-pass effect.12

It was therefore our objective to compare prava- statin and lovastatin pharmacokinetic and pharma- codynamic effects and their interactions with cyclo- sporine in kidney transplant patients after single and multiple doses.

PATIENTS AND METHODS Stua’y design. This study was a double-blind,

double-dummy, randomized, parallel-group multi- center comparison of pravastatin and lovastatin in stable kidney allograft recipients. Patients received a single 20 mg oral daily dose of pravastatin or lova- statin for 28 days. Pharmacokinetic profiles for single- and multiple-dose pravastatin and lovastatin were obtained after 1 day and 28 days of dosing, respectively. Similarly, pharmacokinetic profiles of cyclosporine were collected in the absence of prav- astatin or lovastatin 1 day before the start of the study (day -l), as well as during the study to obtain single-dose (day 1) and multiple-dose (day 28) prav- astatin or lovastatin data. The lipid profile of the patients was determined on days -1, 14, and 29. Pharmacokinetic evaluation of the study drugs in the presence of cyclosporine was defined as the primary end point of the study and pharmacody- namic effects as the secondary end point. Patients were enrolled in three study centers (Hannover, Gottingen, and Freiburg, Germany). The study was

organized, conducted, monitored, and analyzed in compliance with the rules of “good clinical prac- tice.“13

Patients. Male and female kidney graft recipients without childbearing potential were included. Their ages were between 18 and 65 years, their body mass index was between 17 and 32 kg/m2, and their trans- plantation had been performed more than 6 months earlier. All patients had stable kidney graft function. Total cholesterol was between 6.2 and 11.7 mmol/L, LDL cholesterol was >4.1 mmol/L (measured on two occasions with an interval of 2 weeks), and fasting triglyceride serum concentrations were lower than 4.5 mmol/L (measured on two occasions with an interval of 2 weeks). The immunosuppressive regimen was based on cyclosporine administered as oral capsule formulation twice daily and prednisone. The cyclosporine dose had to be between 2 and 6 mg/kg/day and not have been changed by more than 30% in the 2 months before enrollment. Prednisone doses had to be between 5 to 10 mg/day for the 2 months before enrollment. Patients had to be in a stable condition without any acute disease. Patients with diabetes and patients with hypertension were allowed to enter the study.

Patients were not allowed to have taken the fol- lowing drugs: barbiturates for 4 weeks before enroll- ment, probucol for 12 months before enrollment, and the investigational drugs for 30 days before enrollment. In addition, patients with histories of alcohol or drug abuse or with positive markers for hepatitis B or human immunodeficiency virus were excluded. Patients were informed according to the Declaration of Helsinki and its revisions and had to give written consent before enrollment. The study was approved by the ethics committees of all three study centers. Forty-four patients were recruited and 23 patients were randomized to receive prava- statin and 21 patients to receive lovastatin. One patient in the pravastatin group discontinued partic- ipation in the study on day 21 because of a >50% rise of serum creatinine concentration caused by a complication unrelated to the study medication; that patient’s data were not included in the pharmacoki- netic analysis. All patients in the lovastatin group completed the study. The demographic data of the patients in the two study groups are listed in Table I. There were no differences between the two groups.

Two weeks before randomization, a medical his- tory was recorded, and the patients underwent a physical and ophthalmologic examination and a 12- lead ECG. Vital signs were measured and clinical

CLINICAL PHARMACOLOGY &THERAPEUTICS VOLUME 62, NUMBER 3 Olbricht et al. 3 13

Table I. Demographic data and important laboratory tests before begin of the study

Parameter SUY group

Pravastatin (n = 23) Lovastatin (n = 21)

Age (yr) Body mass index (kg/m’) Height (m) Weight (kg) Males Females Total cholesterol (mmol/L) LDL cholesterol (mmol/L) HDL cholesterol (mmol/L) VLDL cholesterol (mmol/L) Triglycerides (mmol/L) Creatinine clearance (ml/min/1.73 m2) Urinary protein (mg/24 hr) Hemoglobin (gm/dl) Blood urea (mmol/L) Creatinine ( pmol/L) Uric acid (mmol/L) Total bilirnbin (pmol/L) Alkaline phosphatase (U/L) Alanine aminotransferase (U/L) Aspartate aminotransferase (U/L) y-Glutamyl transferase (U/L) Lactate dehydrogenase (U/L)

46.9 + 11.9 24.6 2 2.4

170.0 c 9.0 71.1 k 10.8

13

SE 0 9 5.6 t 0:8 1.8 2 0.6 0.6 + 0.3 2.2 ? 0.9

81.8 +- 26.8 201.8 +- 233

14.6 + 1.6 9.7 r+ 4.0

123.8 + 35.5 1.2 r+ 0.3

13.8 + 8.7 129.0 2 50.5

9.0 + 2.3 8.6 t- 3.5

16.5 + 8.6 206.4 2 32.6

42.7 2 12.8 24.8 t 3.1

172.1 t 10.3 73.9 t 12.5

11 10

8.4 2 1.3 5.9 t 1.3 1.6 2 0.6 0.6 2 0.3 2.5 ‘-’ 1.0

17.5 +- 26.4 372.0 2 567.1

13.7 + 1.5 11.3 + 4.3

123.8 2 26.5 1.3 ? 0.3

12.0 2 5.1 126.9 + 39.7

8.7 + 2.5 10.5 2 6.0 21.6 -c 16.3

195.3 2 27.3

Data are. mean values t SD. LDL, Low-density lipoprotein; HDL, high-density lipoprotein; VLDL, very low-density lipoprotein.

chemical and biochemical blood and urine parame- ters were checked. The parameters and screening results are given in Table I. Laboratory tests, vital signs, and creatinine clearance were checked at the visits during the study. The same clinical screening program was repeated after the study (day 29). The patients were asked to refrain from strenuous activities during the study. Caffeine and alcoholic beverages were not allowed for 24 hours before and fasting was required for 12 hours before phar- macokinetic sampling was begun. During sample collection a standardized diet was served. Patients were not allowed to receive any changes in dose of concomitant medication or of medication started during the study. In addition to immunosuppres- sants, 11 subjects in the lovastatin group received calcium antagonists (nifedipine [eight subjects], nitrendipine [two subjects], and diltiazem [one subject]) and 18 subjects in the pravastatin group received calcium antagonists (nifedipine [lo sub- jects], verapamil [three subjects], and nitrendipine [five subjects]).

Blood sample collection. Blood samples for the se- rum pravastatin or lovastatin assays were drawn from a cubical vein into a tube without anticoagu-

lant. Clotting was allowed for 30 minutes on ice. Samples were centrifuged at +4” C and the serum was deep-frozen. Pravastatin and lovastatin were extracted and quantified with use of a modification of a gas chromatography-mass spectrometry method described previously.14 During the study, the pravastatin assay had the following specifica- tions: lower limit of quantification, 0.5 kg/L; range of reliable response, 0.5 to 150 &I; the calibration curve was linear with y [l~,g/L] = 0.0073 - x - 0.0003 (I’ = 0.99); within-run precision was 6.0% at 5.2 l~,g/L and 3.7% at 104 l.rg/L; between-run precision was 0.9% at 5.2 pg/L and 4.4% at 104 kgJL; accuracy was 10.2% at 5.2 &L and 2.7% at 104 l~,g/L. Lova- statin samples were hydrolyzed with potassium hy- droxide before analysis to convert the lovastatin lactone to the corresponding acid. During the study, the method had the following specifications: lower limit of quantification, 0.6 l.@L; range of reliable response, 0.6 to 114 l.@L; the calibration curve was linear with y [P&L] = 0.026 * x + 0.006 (I” = 0.99); within-run precision was 7.5% at 5.4 P&L and 12.4% at 75.3 pg/L; between-run precision was 7.9% at 5.4 pg/L and 7.8% at 75.3 &L; accuracy was 3% at 5.4 &L and -1.6% at 75.3 l~,g/L.

3 14 Olbricht et al. CLINICAL PHARMACOLQGY & THERAPEUTICS

SEPTEMBER 1997

Blood samples for the determination of cyclospor- ine (2.5 ml) were anticoagulated with ethylenedia- minetetraacetic acid (EDTA) and stored at -20“ C. Cyclosporine was analyzed by means of a previously described HPLC assay,15 cyclosporin D as the inter- nal standard, with the following specifications: lower limit of quantification, 25 &L; range of reliable response, 25 to 5000 t&L; the calibration curve was linear with y [lo@.,] = 0.97x + 28 (? = 0.98); within-run precision was 13% at 25 pg/L and 8.2% at 250 p&L..; between-run precision was 8% at 50 u&L and 14.5% at 1000 @L; accuracy was +1.9% at 50 l.rg/L and +5% at 1000 l&L. Method valida- tion, sample analysis, quality assurance, and docu- mentation followed the rules of “good laboratory practice.“16 In addition, the cyclosporine assay was externally validated by monthly participation in the U.K. cyclosporine quality assessment scheme.17

Triglyceride and total cholesterol concentrations in serum were determined with use of standard enzymatic assays and electrophoresis. Very-low- density lipoprotein (VLDL) and LDL cholesterol were estimated by application of the formula of Friedewald et all8 This was possible because none of the patients had a marked hypertriglyceridemia (>4.5 mmol/L) or type III hyperlipoproteinemia.lg

Phamucokinetic a~lysi~. Pharmacokinetic pa- rameters of pravastatin and lovastatin were calcu- lated on the basis of noncompartmental analysis*’ and those of cyclosporine were calculated on the basis of two-compartmental analysis*l with use of the algorithms implemented in the respective soft- ware programs: the maximum concentration (C,,), and the time to reach G, (t,,&, AUC(O-12) for cyclosporine and the AUC(O-24) for pravastatin and lovastatin, and the concentration before the next dose (C,,). In addition, the accumulation index was calculated for pravastatin and lovastatin and was defined as the AUC(O-24) at day 28/AUC(O-24) at day 1.

SWal analysis. Demographic data were com- pared for both groups with use of the two-sample t test. Distribution of male and female patients in both groups was assessed by Fisher’s exact test. Quantitative measurements in the two study groups were analyzed, if not stated otherwise, with the Wil- coxon rank-sum test. For the analysis of AUC(O-24) and C,, of lovastatin and pravastatin within each study group, a mixed-effect linear model with “day” as the fixed effect and both “subject” and the resid- uals as the random effects was used. Two-sided 95% confidence intervals of the geometric mean prava-

statin/lovastatin ratio and of the geometric mean day 28/day 1 ratio were calculated. The estimation method was the restricted maximum likelihood. For t of lovastatin and pravastatin, the Wilcoxon srg;ed rank test was used to test for a “day” effect. A nonparametric two-sided 95% confidence interval for the median difference between day 28 and day 1 was calculated.** C,, was used to prove that steady state was reached during the study.

The pharmacodynamic effects were tested with use of ANOVA. The last available value before the start of the study was considered to be the baseline value.

RESULTS Based on analysis of C,, values, steady-state con-

ditions of both lovastatin and pravastatin were reached between day 2 and day 7. The mean phar- macokinetic profiles of pravastatin after a single dose (day 1) and multiple doses (day 28) are shown in Fig. 1, A, and the respective pharmacokinetic parameters are shown in Table II. The geometric mean values of the pravastatin C,,,, were 84.3 p&L and 82.8 &L on day 1 and on day 28, respectively. This corresponded to a 1% decrease over the study period without reaching statistical difference: the geometric mean day 28/day 1 ratio was 0.98, and the 95% confidence interval was 0.72 to 1.35. The geo- metric mean AUC(O-24) was 248 pg/ * hr/L on day 1 and 271 kg - hr/L after day 28, corresponding to a 9.3% increase over the study period without reach- ing statistical significance: the geometric mean day 28lday 1 ratio was 1.09, with a 95% confidence interval of 0.85 to 1.41. Changes of the individual pravastatin and lovastatin AUC values are depicted in Fig. 2. The median value of t,, at day 1 of 1% hour (range, 1 to 3 hours) was the same as that at day 28 of 1% hour (range, 1/2 to 4 hours). The median day 28 to day 1 difference was 0 hours (p = 0.98; 95% confidence interval, -1/4 to Y, hour).

The lovastatin mean pharmacokinetic profiles at days 1 and 28 are shown in Fig. 1, B, and the pharmacokinetic parameters during the study are shown in Table II. The mean serum concentrations indicated increased absorption of lovastatin. C,, values were not different during the study period. The geometric mean lovastatin C,, was 48.2 &L after a single dose (day 1) and 67.4 t~,g/L (day 28) after multiple doses, corresponding to a 40% in- crease. The difference was statistically significant (p = 0.003): the the geometric mean day 28/day 1 ratio was 1.4, with a 95% confidence interval of 1.14

CLINICAL PHARMACOLOGY & THERAPEUTICS VOLUME 62, NUMBER 3 Olbricht et al. 315

80

80

80

80

40

20

0

0 4 8 12 18 20 24 time [h]

I I I I I I I

0 4 8 12 18 20 24

time [h]

Fig. 1. Pharmacokinetic profiles of pravastatin (A) and lovastatin (B) in serum of stable kidney graft recipients after a single (day 1) and after multiple (day 28) 20 mg/day doses. Pravastatin and lovastatin were quantified with use of a gas chromatography-mass spectrom- etry assay on the basis of chemical ionization and detection of negative ions. All data points are given as mean 2 SE (pravastatin, n = 22, lovastatin, n = 21). Circles, Single-dose pharmacokinetics (day 1); squares, multiple-dose pharmacokinetics (day 28).

to 1.72. The geometric mean AUC of 276 ug . hr/L to day 1 difference was Y2 hour (p = 0.36: 95% at day 1 was lower than that on day 28 of 405 confidence interval, -1/2 to 1% hour). pg * hr/L (p < 0.001; Fig. 2). The geometric mean The accumulation index of lovastatin (1.47; p = day 28/day 1 ratio was 1.5 and the 95% confidence 0.001; 95% confidence interval, 1.17 to 1.84) was interval was 1.22 to 1.76. The median &,, was the higher than that of pravastatin (1.09;~ = 0.41; 95% same on day 1(4 hours; range, 1 to 6 hours) and day confidence interval, 0.88 to 1.35). The difference 28 (4 hours; range, 1 to 8 hours); the median day 28 between the accumulation index of lovastatin and

3 16 Olbricht et al. CLINICAL P HARMACOLOGY & THERAPEUTICS

SEPTEMBER 1997

Table II. Pharmacokinetic parameters of lovastatin and pravastatin

Pravastatin Lovastatin

Median Range Median Range

Day 28 80*

Day 28 1.5* AUC(O-m) (pg * hr)

Day 1 249* Day 28 241*

(Lin (IV@) Day 1 o.o* Day 28 0.3*

29-325* 29-456*

l.O-3.0* 0.5-4.0*

104-1026* 115-969*

O.O-1.6* O.O-2.5*

467 75t

4.0t 4.OP

243$ 4591-

1.77 o.o-11.5t 3.67 O.O-16.7.t

16-149t 22-298t

l.O-6.0t l.O-8.Ot

105-859$ 140-1508t

C,,,,, Observed maximum blood/serum drug concentration; bax, time from drug intake to C,,; AUC, area under the curve; Cmin, observed minimum blood/serum drug concentration.

*n = 22; tn = 21; $n = 20.

that of pravastatin approached statistical signifi- cance: geometric mean pravastatin/lovastatin ra- tio = 0.74;~ = 0.062; 95% confidence interval, 0.55 to 1.02.

The steady state of cyclosporine was already es- tablished when the study was started. The cyclospor- ine trough levels (Cmin) are listed in Table III. Com- parison of the pharmacokinetic parameters of cyclosporine in the absence of or in the presence of single or multiple lovastatin or pravastatin doses showed no differences (Table III). In addition, there were no differences in the pharmacokinetic param- eters of cyclosporine between the study groups.

During the study period, VLDL, LDL, and total cholesterol decreased in the majority of subjects during the study period (Fig. 3). LDL cholesterol concentrations were reduced by a clinically signifi- cant amount in 19 (82.6%) subjects in the pravasta- tin group and in 20 (95.2%) subjects in the lovastatin group. For LDL cholesterol concentrations, the de- crease from baseline was statistically significantly smaller for subjects in the pravastatin group than in the lovastatin group: The mean difference between the groups was 0.53 mmol/L (p = 0.039; 95% con- fidence interval, 0.03 to 1.04 mmol/L). In addition, the difference from baseline total cholesterol was statistically significantly smaller in subjects in the pravastatin than in the lovastatin group: The mean difference between the groups for the change from baseline 0.62 mmol/L (p = 0.014; 95% confidence interval, 0.14 to 1.11 mmol/L). Mean high-density

lipoprotein (HDL) cholesterol in all three centers together did not change over the study period. How- ever, there was an increase of HDL in two centers and a mean decrease from baseline by 0.15 mmol/L in the third center. The study group-by-center inter- action was statistically significant for VLDL choles- terol (p = 0.002). In one center, the decrease from baseline was significantly greater for the lovastatin group, whereas in the other two centers there was no significant difference between the study groups. There was a statistically significant decrease in tri- glycerides between day 14 and day 29: mean differ- ence was 0.29 mmol/L @ = 0.028; 95% confidence interval, 0.03 to 0.57 mmoliL). There was no statis- tically significant difference between the study groups.

Adverse events. Vital signs were essentially un- changed during the study. During the study one seri- ous adverse advent that required prolonged hospital- ization was recorded in a patient in the lovastatin group. This adverse event was a sleeping disorder that was known before study enrollment and that was un- related to the study drug. Twenty-three adverse effects were reported in the pravastatin group and 20 in the lovastatin group. One adverse event in the pravastatin group that was not related to the study medication was severe (increase of creatinine serum concentration more than 50%) and the patient ceased participation in the study. Activity of the creatine kinase was not different from the baseline values. The most frequently reported events were increased serum concentration of myoglobin (four patients in the pravastatin group and three in the lovastatin group), nausea (two in the pravastatin group and one in the lovastatin group), muscle ache (two in the pravastatin group and one in the lovastatin group), and abdominal pain (two in the pravastatin group and one in the lovastatin group).

DISCUSSION This study showed that, after multiple doses, lo-

vastatin, but not pravastatin, significantly accumu- lated in the blood of cyclosporine-treated kidney graft patients. The effect of cyclosporine on lova- statin pharmacokinetics was readily explicable: Lo- vastatin and cyclosporine are substrates of the same CYP enzymes8,9 and are therefore potential com- petitive inhibitors of their metabolism. Metabolism of CYP3A substrates in the small intestine reduces their absorption,23,24 and metabolism in the liver contributes to first-pass metabolism and to systemic elimination. Thus coadministration of cyclosporine leads to an increased bioavailability of lovastatin as a result of higher absorption, to a reduced first-

CLINICAL P HABMACOLOGY & THERAPEUTICS VOLUME 62, NUMBER 3

Iovastatin

100

Olbricht et al. 317

pravastatin

day 1 day 28 day 1 day 28

Fig. 2. Lovastatin and pravastatin area under the curve from 0 to 24 hours [AUC(O-24)] values of individual subjects after single (day 1) and multiple doses (day 28).

pass effect and to reduction of its systemic elimina- tion, resulting in the significant accumulation ob- served in this study. It can be anticipated that this also applies to the combination of lovastatin with other CYP3A substrates and inhibitors, such as cal- cium channel blockers, macrolide antibiotics, or an- timycotics.’ In this study, calcium channel blockers were obviously not responsible for the pharmacoki- netic differences between pravastatin and lovastatin because more subjects in the pravastatin group (18 subjects) than in the lovastatin group (11 subjects) received concomitant antihypertensive therapy based on these drugs.

The affinity of pravastatin to CYP enzymes, as determined by differential ultraviolet spectroscopy, is almost 2000 times lower (Ks = 35,000 p,mol/L) than that of lovastatin (Ks = 18 p.mol/L) in the same assay (Jacobsen W, Sewing KIT, Christians U. Un- published data, May 1996). This explains the lack of detectable pravastatin metabolism when pravastatin

was incubated with human liver and small intestinal microsomes (Jacobsen W, Sewing KF, Christians U. Unpublished data, May 1996). In the same study, lovastatin was readily metabolized to its major metabolites. In contrast to lovastatin, the major pravastatin metabolites are generated by non- CYP-dependent processes: 3ol-isopravastatin and metabolite 6-epipravastatin are either formed by acidic degradation of pravastatin in the stomach25 or by sulfation at the 6’p-hydroxy group by sulfo- transferases, followed by a nucleophilic attack of hydroxy anions at the 3’~ or 6’a-position.25 Be- cause of the lack of detectable in vitro metabolism by liver microsomes, a possible involvement of CYP enzymes in pravastatin metabolism has not been confirmed. At present, it can be concluded that CYP-mediated biotransformation of pravastatin does not contribute to the elimination of pravastatin to a significant extent. Metabolism is only a minor route of elimination for pravastatin, whereas lova-

3 18 Olbricht et al. CLINICAL PHARMACOLOGY & THERAPEUTICS

SEPTEMBER 1997

Table III. Pharmacokinetic parameters of cyclosporine (INN, ciclosporin)

PraVUSfUtiU Lovastatin Median Range Median Range

Day 1 Day 28

tm~~~~l Day 1 Day 28

AUC(O-00) (pg - hr/L) Day -1 Day 1 Day 28

crl~a~“~’

Day 1 Day 28

471* 445* 368*

3.9t 3.7* 4.0*

4,700-t 4,570* 4,490*

113* 114* 142*

153-1,120* 146-1,190* 195-1,140*

0.7-6.9-t 0.8-7.3* 0.5-8.2*

2,070-24,800t 1,570-21,700* 2,200~28,900*

64-290* 49-274* 52-212*

443t 475$ 466$

3.97 3.5$ 4.5$

4,220-f 4,255$ 4,295$

112-f 121t 15ot

146-1,090t 176-781$ 225-794$

1.0-9.3-f 1.6-6.0$ 0.8-7.8$

2,710-38,900t 2,430~44,3OO$ 2,590-49,600$

78-282-f 65-391-f 91-310t

f&, Observed maximum blood/serum drag concentration; t,,- time from drug intake to C&,, AUC, area under the curve; C,, observed miniimn blood/serum drug concentration.

‘n = 22; tn = 21; $n = 20.

statin is eliminated mostly as metabolites. There- fore, as shown in this study, interaction with cyclo- sporine has a higher impact on lovastatin than on pravastatin pharmacokinetics.

The AUC(O-24) values of pravastatin and lova- statin in kidney graft patients, as found in this study, were severalfold greater than those reported in pa- tients without cyclosporine therapy. Comparison of historic pharmacokinetic data of lovastatin and pravastatin with the results of this study is justified because kidney function in the patients included in this study was stable, and renal elimination is not the major route of elimination for pravastatin or lova- statin. Pravastatin mean AUC(O-24) values were ap- proximately fivefold to sevenfold higher in renal graft recipients receiving cyclosporine compared with historic data: the mean AUC(O-24) after single and multiple doses in this study were 300 and 314 pg * hr/L, respectively, compared with steady-state AUC(O-24) values of 43.7 Fg - hr/L26 and 57.1 Fg * hr/L.i’ However, the mean AUC(O-24) values in the present study compare favorably with the mean steady-state AUC of 1304 kg * hr/L reported after a single 20 mg dose. l2 It should be noted that this report12 was not based on a controlled study. Even more important, compared with the control group, heart transplant patients received a multitude of additional drugs and were generally in far worse condition than the patients in the control group.12 Because of the poor study design, the question has

been raised whether or not a valid cyclosporine- pravastatin drug interaction could be established on the basis of these data.” Because it is unlikely that CYP enzymes significantly contribute to pravastatin elimination (vide supra), it can be speculated that cyclosporine interacts with pravastatin transport pro- cesses. The exact mechanism of the cyclosporine- pravastatin interaction remains to be evaluated.

Unfortunately, there is only limited information on lovastatin pharmacokinetics available because HMG-CoA reductase inhibitory activity has been determined without identification of lovastatin acid or active metabolites in most studies. Steady-state AUC(O-24) in healthy subjects after a single 40 mg dose was 44.9 pg - hr/L. Assuming dose-linear kinet- ics, this suggests an AUC(O-24) value of 22.5 pg - hr/L for a single 20 mg lovastatin dose. Lovastatin mean AUC(O-24) values after single and multiple doses in the present study were 326 and 492 pg * hr/L, respectively, showing a twentyfold increase of lovastatin AUC(O-24) values in the presence of cy- closporine. This estimate is supported by a recent study that evaluated the clinical interaction of the CYP3A enzyme inhibitor itraconazole and lova- statin in healthy volunteers.28 Compared with the pharmacokinetics after administration of placebo instead of itraconazole, in the presence of itracon- azole, the C,, of the active metabolite, lovastatin acid, was increased thirteenfold and the AUC(O-24) was increased twentyfold.=

CLINICAL PHARMACOUXY &THERAPEUTICS VOLUME 62, NUMBER 3 Olhicht et al. 319

6

total cholestefol

T

Fig. 3. Lipid profiles in blood of renal allograft recipients receiving 20 mg/day pravastatin (n = 23, open bars) or lovastatin (n = 21, hatched bars). All data are presented as mean values + SE.

The lack of an effect of lovastatin on cyclosporine pharmacokinetics could be explained by the more than tenfold higher concentrations of cyclosporine and by a higher affinity of cyclosporine to CYP enzymes (cyclosporine: Ks = 2.5 mmol/L).2g

Pharmacodynamic analysis revealed a slightly but significantly greater lipid-lowering effect of lova- statin compared with pravastatin, although the same doses were administered. This might at least par- tially be caused by the higher systemic availability of lovastatin as a result of the interaction with cyclo- sporine. In contrast to lovastatin and other HMG-

CoA reductase inhibitors, pravastatin selectively in- hibits HMG-CoA reductase in the liver. The reason may be its high affinity to a sodium-independent bile acid carrier system in the hepatocyte.30 Although the AUC of pravastatin was almost fivefold higher than in the absence of cyclosporine, the resulting lipid-lowering effect in the present study was less than it could be expected from studies based on patients not receiving cyclosporine therapy who re- ceived higher pravastatin doses. It remains to be evaluated whether cyclosporine interferes with the carrier system responsible for pravastatin uptake

320 Olbricht et al.

into the hepatocyte and thus interacts with prava- statin pharmacokinetics as well as pharmacodynam- its.

CLINICAL I’HARMA COLOGY & THERAPEUTICS SEFTEMBER 1997

7.

Accumulation of lovastatin bears the potential of severe side effects, such as rhabdomyolysis, which have been reported after combination with the CYP3A substrates cyclosporine and erythromycin.11*31”4 Myopathies in patients treated with lovastatin as monotherapy occurs in 0.1% to 0.2% of patients.35 This incidence can rise to 30% with coadministra- tion of cyclosporine.10 This has led to dose- reduction recommendations for lovastatin in trans- plant patients. 36 This safety issue has not been reported in similar studies with pravastatin677S36 and might be explained by the less significant effect of cyclosporine on pravastatin pharmacokinetics estab- lished in the present study.

effects of pravastatin on hyperlipidemia in renal trans- plant recipients. Transplantation 1992;53:94-9. Yoshimura N, Ohmori Y, Tsuji T, Oka T. Effect of pravastatin on renal transplant recipients treated with cyclosporine- 4 year follow-up. Transplant Proc 1994; 26:2632-3.

8. Wang RW, Kari PH, Lu AYH, Thomas PE, Guengerich FP, Vyas KP. Biotransformation of lova- statin; IV: identification of cytochrome P450 3A proteins as the major enzymes responsible for oxi- dative metabolism of lovastatin in rat and human liver microsomes. Arch Biochem Biophys 1991;290: 355-61.

9.

In conclusion, this study shows that, compared with lovastatin, pravastatin pharmacokinetics is less affected by coadministration of cyclosporine. This may provide an additional safety margin and there- fore may be an advantage in long-term treatment of kidney graft recipients receiving cyclosporine immu- nosuppression.

10.

11.

12.

We thank Alex Giuliano, Wolfang Grotz, Annick Le- roy, Annette Linck, Andrea Roth, Heike Spitznagel, Marion Tscherwinsky, and Kan-Fat Yee for their valuable contributions to this study. The study was part of Dr. Doll’s thesis.

Pichard L, Fabre I, Fabre G, Domergue J, Saint Aub- ert B, Mourad G, et al. Cyclosporin A drug interac- tions: screening for inducers and inhibitors of cyto- chrome P-450 (cyclosporin A oxidase) in primary cultures of human hepatocytes and in liver micro- somes. Drug Metab Dispos 1990;18:595-606. Tobert JA. Efficacy and long-term adverse effect pat- tern of lovastatin. Am J Cardiol 1988;62:28J-345. East C, Alivizatos PA, Grundy SM, Jones PH, Farmer JA. Rhabdomyolysis in patients receiving lovastatin. N Engl J Med 1988;318:47-8. Regazzi MB, Iacona I, Campana C, Raddato V, Lesi C, Perani G, et al. Altered disposition of pravastatin following concomitant drug therapy with cyclosporin A in transplant recipients. Transplant Proc 1993;25: 2732-4.

13.

14.

References 1. Pirsch JD, D’Alessandro AM, Sollinger HW,

Knechtle SJ, Reed A, Kalayoglu M, et al. Hyperli- pemia and transplantation: etiologic factors and ther- apy. J Am Sot Nephrol 1992;2:S238-42.

2. Kasiske BL. Risk factors for accelerated atheroscle- rosis in renal transplant recipients. Am J Med 1988; 84:985-92.

15.

3. Castelao AM, Grinyo JM, Castineiras MJ, Fiol C, Gilvemet S, Seron D, et al. Effect of pravastatin in the treatment of hypercholesterolemia after renal trans- plantation under cyclosporine and prednisone. Trans- plant Proc 1995;27:2217-20.

16.

17.

EEC/CPMP Working Party. Good clinical practice for trials on medical products in the European Commu- nity. III/3976/88-EN draft No 3; August 17, 1989. Morris MJ, Gilbert DJ, Hsieh JYK, Matuszewski BK, Ramjit HG, Bayne WF. Determination of the HMG- CoA reductase inhibitors simvastatin, lovastatin, and pravastatin in plasma by gas chromatography/chemi- cal ionization mass spectrometry. Biol Mass Spectrom 1993;22:1-8. Christians U, Zimmer KO, Wonigeit K, Maurer G, Sewing KF. Liquid-chromatographic measurement of cyclosporin and its metabolites in blood, bile and urine. Clin Chem 1988;34:34-9. Food and Drug Administration. Good laboratory practice regulations: final rule. Federal Register 1987; 52:33768-82. Holt DW, Marsden JT, Johnson A. Quality assess- ment of cyclosporine measurements: comparison of

4. Rehman MA, Al-Sulaiman MH, Mousa DH, Al-Hawas current methods. Transplant Proc 1990;22:1234-9. FA, Abdalla AH, Rassoul Z, et al. Effects of simvastatin 18. Friedewald WT, Levy RI, Fredrickson DS. Estimation in hyperlipidemic renal transplant patients receiving cy- of the concentration of low-density lipoprotein cho- closporine. Transplantation 1995;27:397-9. lesterol in plasma, without use of the preparative

5. Cheung AK, De Vault GA, Gregory MC. A prospec- ultra-centrifuge. Clin Chem 1972;18:499-502. tive study on treatment of hypercholesterolemia with 19. Thompson GR. Quantitation of blood lipids and li- lovastatin in renal transplant patients receiving cyclo- poproteins. In: A handbook of hyperlipidaemia. Lon- sporine. J Am Sot Nephrol 1993;3:1884-91. don: Current Science; 1989. p. 43-58.

6. Yoshimura N, Oka T, Okamoto M, Ohomori Y. The 20. Farmer RH, Muniak JF, Pittman KA. Management of

CLINICAL PHARMA COLOGY &THERAPEUTICS VOLUME 62, NUMBER 3 Olbricht et al. 32 1

PK data using HP3357 mainframe IBM interfacing. Drug Info J 1987;21:141-52.

21. Heinzel G, Woloszczak R, Thomann P. TopFit 2.0: pharmacokinetic and pharmacodynamic data anal- ysis for the PC. Stuttgart: Gustav Fischer Verlag; 1993.

22. Campbell MJ, Gardner MJ. Calculating confidence intervals for some non-parametric analyses. Br Med J 1988;296:1454-6.

23. Kolars JC, Stetson PL, Rush BD, Ruwart MJ, Schmiedlin-Ren P, Duel1 EA, et al. Cyclosporin A metabolism by P45OIIIA in rat enterocytes-another determinant of oral bioavailability? Transplantation 1992;53:596-602.

24. Wacher VJ, Salphati L, Benet LZ. Active secretion and enterocyte drug metabolism barriers to drug ab- sorption. Adv Drug Delivery Rev 1996;20:99-112.

25. Kitazawa E, Tamura N, Iwabuchi H, Uchiyama M, Muramatsu S, Tkahagi H, et al. Biotransformation of pravastatin sodium; I: mechanism of enzymic trans- formation of epimerization of an allylic hydroxy group of pravastatin sodium. Biochem Biophys Res Com- mun 1993;192:597-602.

26. Pan HY, Funke PT, Willard DA, MC Kinstry DN. Pharmacokinetics, pharmacodynamics and safety of pravastatin sodium, a potent inhibitor of HMG-CoA reductase, in healthy volunteers. In: LaRosa C. New advances in the control of lipid metabolism: focus on pravastatin. London: Royal Society of Medicine Ser- vices; 1989. p. 9-21.

27. Quion JAV, Jones PH. Clinical pharmacokinetics of pravastatin. Clin Pharmacokinet 1994;27:94-103.

28. Neuvonen PJ, Jalava MD. Itraconazole drastically in-

creases plasma concentrations of lovastatin and lova- statin acid. Clin Pharmacol Ther 1996;60:54-61.

29. Combalbert J, Fabre I, Fabre G, Daleti I, Derancourt J, Cano JP, et al. Metabolism of cyclosporin A, IV: purification and identification of the rifampicin induc- ible human liver cytochrome P-450 (cyciosporin A oxidase) as a product of P450IIIA gene subfamily. Drug Metab Dispos 1989;17:197-207.

30. Ziegler K, Blumrich M, Hummelsiep S. The trans- porter for the HMG-CoA reductase inhibitor prava- statin is not present in HEP G2 cells: evidence for the nonidentity of the carrier for pravastatin and certain transport systems for BSP. Biochim Biophys Acta 1994;1223:195-201.

31. Corpier CL, Jones PH, Suki WN, Lederer ED, Qui- nones MA, Schmidt SW, et al. Rhabdomyolysis and renal injury with lovastatin use: report of two cases in cardiac transplant recipients. JAMA 1988;8:239-41.

32. Spach DH, Bauwens JE, Clark CD, Burke WG. Rhab- domyolysis associated with lovastatin and erythromy- tin use. West J Med 1991;154:213-5.

33. Alejandro DS, Petersen J. Myoglobinuric acute renal failure in a cardiac transplant patient taking lovastatin and cyclosporine. J Am Sot Nephrol 1994;5:153-60.

34. Marinella MA. More on lovastatin. West J Med 1995; 162:176-7.

35. Tobert JA, Shear CL, Chremos AN, Mantel1 GE. Clinical experience with lovastatin. Am J Cardiol 1990;65:23F-6F.

36. Hsu I, Spinler SA, Johnson NE. Comparative evaluation of the safety and efficacy of HMG-CoA reductase inhib- itor monotherapy in the treatment of primary hypercho- lesterolemia. Ann Pharmacother 1995;29:743-59.