Journal of Pharmacokinetics and Biopharmaceutics, Vol. 7, No. 1, 1979

Clinical Pharmacokinetics of Procainamide Infusions in Relation to Acetylator Phenotype John J. Lima, 1 David R. Conti, 2 Allen L. Goldfarb, 2 William J. Tilstone, 3 Lawrence H. Golden, 2 and William J. J u s k o 1'4 Received May 31, 1978--Final August 16, 1978

The pharrnacokinetics of procainamide was determined in 21 lidocaine-resistant patients who received the drug according to a pharmacokinetically designed double-infusion technique. Thirteen patients were phenotyped as slow acetylators, seven as fast, and one as intermediate. The total body clearances (Cl-r) of PA in slow and fast acetylators were 22.6 and 34.8 liters/hr, respectively. The fraction of PA cleared by the formation of N A P A in the corresponding acetylator group was 0.2 and 0.4. Renal impairment affected the pharmacokinetics of PA more profoundly as the CIT's of PA in patients with and without renal impairment were 1 Z 9 and 31.2 liters/hr, respectively. None of the calculated volumes of distribution was affected by acetylator phenotype or renal impairment. These data identify the contribution of at least two of the major factors accounting for variability in PA disposition in patients undergoing therapy.

KEY WORDS: procainamide; pharrnacokinetics; constant-rate infusion; acetylator phenotype; pharmacogenetics; renal impairment.

This work was supported by Grants 20852 and 24211 from the National Institutes of General Medical Sciences, National Institutes of Health. 1Department of Pharrnaceutics, School of Pharmacy, State University of New York at Buffalo, Amherst, New York 14260. 2Department of Medicine, School of Medicine, State University of New York at Buffalo, Buffalo, New York 14214. 3Department of Forensic Chemistry, University of Strathclyde, Glasgow, Scotland. 4Address correspondence to Dr. W. J. Jusko, Clinical Pharmacokinetics Laboratory, Millard Fillmore Hospital, 3 Gates Circle, Buffalo, New York 14209. 69 0090-466X/79/0200-0069503.00/0

(~) 1979 Plenum Publishing Corporation

70

Lima, Conti, Goldfarb, Tilstone, Golden, and Jusko

Glossary

PA NAPA LI MI Vp V, V~s VdB

C112 CIT C1R CIA CIAp CIM CpA

c, CpA v. CI~ Clo CIB~ CNAPA c~ ss

Procainamide N- Acetylprocainamide Loading infusion Maintenance infusion Volume of central compartment for PA Volume of tissue compartment for PA Volume of distribution of PA at steady state Volume of distribution of PA during postdistributive phase Intercompartmental clearance of PA Total body clearance of PA Renal clearance of PA Acetylation clearance of PA Apparent acetylation clearance of PA Metabolic (nonacetylation)clearance of PA Serum concentration of PA Tissue concentration of PA Steady-state serum concentration of PA Volume of distribution of NAPA Renal clearance of NAPA Nonrenal clearance of NAPA Total body clearance of NAPA Serum concentration of NAPA Steady-state serum concentration of NAPA

INTRODUCTION Procainamide (PA) has a narrow therapeutic range and its disposition varies considerably among patients (1). One factor which may account for its variable disposition is that P A is a substrate for polymorphic acetylation (2-5). Depending on the acetylator status of the patient, serum concentrations of the N-acetylated metabolite of P A ( N A P A ) approach or exceed serum concentrations of the parent c o m p o u n d at steady state (2,4). This is of clinical importance because N A P A appears to have antiarrhythmic properties equal to those of P A (6-8). Furthermore, slow acetylators appear to be at greater risk of developing PA-induced systemic lupus erythematosus (SLE) (9,10). Earlier pharmacokinetic studies in patients have failed to take into account the influence of acetylator status on the disposition kinetics of P A (1,11). More recently, Dutcher et al. (12) and Gibson et al. (13) studied the pharmacokinetics of P A and N A P A in a limited number of normal and anephric subjects of slow and fast acetylator phenotype. This report describes the influence of acetylator phenotype as well as other commonly encountered factors on the pharmacokinetics of P A in a larger group of cardiac patients who were undergoing acute therapy with the drug according to a pharmacokinetically designed dosage regimen (4).

Clinical Pharmacokinetics of Procainamide Infusions

71

EXPERIMENTAL Patients

Twenty-two males and seven females with potentially life-threatening ventricular arrhythmias who, according to the attending cardiologist, were unresponsive to conventional treatment with lidocaine were treated with PA according to our dosing regimen as described below (4). All patients were white, and their ages ranged between 38 and 82 years, with a mean of 61.3 years, and their weights ranged between 50 and 109 kg, with a mean of 77.5 kg. The degree of renal impairment in each patient was assessed by measuring serum creatinine (Cr) and BUN (SMAC Autoanalyzer, Techicon Instruments). Patients were grossly designated as having renal impairment if their serum creatinine exceeded 1.4 mg/dl or their BUN values exceeded 25 mg/dl. These studies occurred in the Cardiac Care Unit of the Millard Fillmore Hospital. Loading doses (LI) of PA HCI ranging between 500 and 1500 mg (mean of 986 mg) were added to 100 ml of dextrose %5 in water (D5W) and infused over 1 hr. Known amounts of PA HC1 were then added to 500 ml of D5W and infused over approximately 23 hr. A Holter pump (model No. 977, Extracorporeal Medical Specialties, King of Prussia, Pa.) was used to administer the two infusions. Because the volume of D5W averaged about 529 + 6 ml, and the dose of PA HCI increased the infusion volume, the final concentration of PA HC1 was lower than the initially calculated value. Since the pump was set to deliver 20 ml of infusion volume per hour, there was always some infusion solution remaining 24 hr after initiation of treatment. Therefore, a corrected maintenance infusion rate (MI) of PA HC1 was calculated by the following equation: MI =

V~-V,D t 2 - tl Vs

(1)

where Vs is the starting infusion volume (529 + 6 ml D5W plus the volume in which the dose of PA HC1 was dissolved), V, is the remaining volume (the volume of infusion remaining at time the infusion was stopped or a new infusion bottle was hung), tx is the time at which the maintenance infusion was started, t2 is the time at which the maintenance infusion was terminated or the infusion solution replaced, D is the dose or amount of PA HCI added to the bottle of D5W. The MI rates ranged between 95 and 375 mg of PA HCl/hr, with a mean of 206 mg/hr. Blood specimens (5 ml) were collected via an indwelling catheter at 0 time, and at 0.25, 0.5, 1, 2, 4, 8, and 24 hr after the initiation of PA treatment in most patients (an average of 7.3 samples of blood/patient were collected, range 5-8 samples/patient). Urine samples were obtained from

72

Lima, Conti, Goldfarb, Tilstone, Golden, and Jusko

most patients at various times during the infusion. Serum and urinary concentrations of PA and NAPA were analyzed by HPLC (14).

Pharmacokinetic Analysis PA serum concentration-time data were fitted to a two-compartment open model (Fig. 1), consisting of central (Vp) and tissue (Vt) volumes and intercompartmental rate constants (k12, k21). The total body clearance (CIT) of PA is the sum of its renal clearance (C1R), acetylation clearance (CIA), and the formation of other metabolites (C1M). Serum concentrations of NAPA were fitted to a one-compartment open model (Fig. 1) which included a volume of distribution (VN) and two elimination pathways: renal clearance (CIN) and nonrenal clearance (Clo). Although two or three compartments have been used to model the pharmacokinetics of NAPA (13,16), because of the limited number of blood samples and the insensitivity of our method of analyzing serum concentrations of NAPA during the loading infusion of PA a one-compartment model was assumed. Serum concentration-time data of PA and NAPA from 15 patients were fitted to the appropriate model by nonlinear least-squares iteration (15). The digital computer program NONLIN (15) was employed with the DFUNC subroutine designed to integrate three differential equations which describe serum concentrations of PA (CpA) (equation 2), tissue concentrations of PA (Ct) (equation 3) and serum concentrations of NAPA (C~APA) (equation 4). I-dose

aCpA/at=[---~--C1r . C p g - e l , z . Cpg +Cl~2 . Ct]/ Vp

(2)

acJat = [(CI12)(CpA-- Ct)]/V,

(3)

aCNAPA

dt

= [(Cla)(CpA) - (C1N + C10)(CNAPA)]/VN

(4)

where Clx2 is the intercompartmental clearance of PA. The equations were written in clearance terms rather than rate constants to facilitate assignment of initial parameter estimates in the NONLIN program. The program was designed to account for the two infusion rates (dose/Y) and residual drug in each compartment during transition from one dosing rate to another. Data values were weighted by use of their reciprocals in the least-squares iteration process, except for Ct values, which were weighted zero because of their use as "dummy" variables for operation of equation 3. The program provided the least-square estimates of C1r, Cl12, Vp, and Vt. In addition, where urinary excretion data were available, an estimate of

73

Clinical Pharmacokinetics of Procainamide Infusions

,

t

Dose/'r

Other t' MeteboLitesUrine

~

~

Urine

e l ' ~ o Other El iminati0n Fig. 1. Pharmacokinetic model summarizing the known disposition characteristics of procainamide and NAPA. Symbols for the model are described in the pharmacokinetic analysis section.

the true acetylation clearance (C1A) of P A was obtained by measuring CIN and assigning C10 as a constant. The inclusion of the nonrenal clearance of N A P A (2.6 liters/hr) was necessary to obtain a good fit of serum concentrations of N A P A and was based on the initial findings of Strong et aL (16), who reported that CIN was 85% of the total body clearance of N A P A in normal subjects. The apparent acetylation clearance of P A (ClAp), obtained as described below (equation 7), provided the initial estimate of CIA. In each case, C1A was always greater than ClAp. To further assess the validity of estimating CIA according to this method, the total body clearance of N A P A (CIB~) in patients with no apparent renal impairment was calculated by C1BN = ( C 1 A ) ( C ~ \ ) / C ~

(5)

and compared to the results obtained from the literature. The calculated CIBN and measured C1N in our patients were 0.130 and 0.113 liter/kg/hr, respectively, and were lower than the respective values of 0.185 and 0.151 liter/kg/hr reported by Strong et al. (16) in three normal volunteers. However, the percent N A P A cleared renally in our patients averaged 87%, which was in excellent agreement with the 81% value reported by these investigators and suggests that our method of estimating CIA is r e a s o n a b l e .

Because urine was not obtained in six patients, only serum concentrations of P A were fitted, and the least-square estimates of C1T, Cl12, Vp, and Vt were obtained (equations 2 and 3). The PA and N A P A serum concentration-time data from eight patients could not be fitted because either an insufficient number of blood samples were obtained (three patients), or the MI was changed (three patients),

74

Lima, Conti, Goldtarb, Tilstone, Golden, and Jusko

or there were inconsistencies in MI (two patients). The pharmacokinetics of N A P A are not included herein but will be the subject of a future report. The rate constants klo, k12, k21 were obtained from the ratio of Clr/Vp, Cl12/V~,and Cl12/Vt (the intercompartmental clearance of P A was assumed to be equal in both directions). T h e distribution and disposition hybrid constants (a,/3) were obtained from the rate constants as described elsewhere (17). The volume of distribution of P A at steady state (V~s ) was obtained by employing V~s = (k12 +

k21/k21)Vp

(6)

and Vdt3 was obtained by the ratio CIr//3. The acetylator phenotype of 15 patients was determined by measuring the apparent acetylation clearance (ClAp) of P A according to ClAp = ERNAPA/C~,

(7)

where ERNApA is the urinary excretion rate of N A P A at steady state and C~A is the steady state serum concentrations of P A (3). Patients were classified as slow or fast acetylators if their ClAp values were 80 ml/min or less, or 1 2 0 m l / m i n or more, respectively (3). Because urine was not collected in six patients, the ratio of serum concentrations of N A P A and PA (N/P) at 24 hr was used as the index of acetylator phenotype. Patients were classified as slow acetylators if steady-state NIP ratios were 0.85 or less and as fast acetylators if N / P ratios were 0.95 or more (2). This is a valid index of acetylator phenotype in patients with no renal impairment (2). There was no evidence of renal impairment in the six patients who were phenotyped according to this index of acetylation.

RESULTS Serum Concentrations of P A and N A P A

Typical serum concentration profiles of P A and N A P A during the course of initial therapy are showh for four patients in Fig. 2. During this pharmacokinetic analysis, calculated serum concentrations of P A were within 20% of the observed values in 80% of the determinations. Similarly, calculated serum concentrations of N A P A were within 20% of the observed value in 75% of the determinations. Most of the calculated serum concentrations of P A and N A P A that showed greatest deviation occurred during the first hour of the infusion. This was due to inconsistencies associated with the loading infusion. In general, a better fit of P A serum concentration data was obtained than of NAPA. This was due in part to assay insensitivity at very low serum concentrations of N A P A during the loading infusion of P A

Clinical P h a r m a c o k i n e t i c s o f P r o c a i n a m i d e Infusions

75

2

~

8

z 121 z

4

w

2

_o

PtAM

~

4

~Q;

2

L._~t

t

.•

9

|

~'

l.-Z

Pt JG

g o

t6 12

~ " ' ~ =

rr .

4 2

Pt EW

0

!

I

2

4

~

L-..

8

24

TIME, hours Fig. 2. Serum concentrations of procainamide (0) and N A P A (11) at various times in slow and fast acetylators with no renal impairment (patients L. A. and A. M.) and in slow and fast acetylators with renal impairment (patients J. G, and E_ W.). The lines are computer fitted, while the points are measured values.

76

Lima, Conti, Goidfarb, Tilstone, Golden, and Jusko

Table I. Characteristics and Procainamide Doses of Four Patients Patient L.A.

J.G.

A.M.

E.W.

Sex Age (yr) Acetylator phenotype BUN (mg/dl) Serum creatinine (mg/dl)

M 38 Slow 7 0.9

F 52 Slow 68 5.0

M 60 Fast 24 1.3

F 82 Fast 76 3.2

Infusion rates, mg/hr Load (1 hr) Maintenance (23 hr)

500 193

750 84

1050 144

750 110

4939 74.8 24.4 6.0 0.913 0.950

2682 39.0 13.6 3.5 0.958 0.973

4362 54.0 58.1 11.0 0.922 0.934

3280 61.0 15.5 3.1 0.914 0.985

Total dose mg/day 1 mg/kg CIT (liters/hr) SD r, PA r, NAPA

and the simplifying assumption of a one-compartment model for the metabolite. The pertinent patient factors and P A doses for the first 24 hr of treatment in the four patients whose data are depicted in Fig. 2 are contained in Table I. The least-square estimate of C1T, its standard deviation, and the correlation coefficients for the entire fit of the PA and N A P A serum concentration-time data are also included. In general, serum concentrations of P A were near to steady state at 2-4 hr following initiation of treatment. Therapeutic serum concentrations of PA (4-10 rag/liter) were achieved in 15-30 min and maintained in patients L. A. and E. W. when given doses of P A HCI according to our dosing guidelines (4). Therapeutic serum concentrations of P A were achieved during the first 2 hr of therapy in patient A. M., while they were subtherapeutic throughout the entire treatment period in patient J. G. Both patients received P A according to the dosing guidelines proposed in earlier literature (1): 50 m g / k g / d a y for the average patient with reduction in the presence of renal insufficiency. According to computer simulation studies, serum concentrations of N A P A are expected to be within 94% of steady state at 24 hr after initiation of intravenous P A in the patients without renal impairment (4). The ratios of serum concentrations of N A P A and PA (N/P) at 8 and 24 hr were identical in patient A. M., indicating that steady-state serum concentrations of N A P A obtained in 8 hr. The NIP ratios at 8 and 24 hr in patient J. G. were 0.75 and 1.3, respectively, while the corresponding values in patient E. W. were 0.83

Clinical Pharmacokinetics of Procainamide Infusions

77

and 2.0, indicating that steady state for N A P A was not achieved. It was not likely to have occurred within 24 hr since both patients had marked renal impairment. Despite the absence of any apparent renal impairment in patient L. A., his NIP ratios at 8 and 24 hr were 0.28 and 0.47, respectively. The reason for this anomaly is unknown, but is probably artifactual since his NIP ratio at 4 hr was 0.39. Acetylator Phenotype

The ClAp values of 15 patients are shown in this histogram of Fig. 3. Nine patients were classified as stow acetylators, five as fast, and one as intermediate. Five of the six patients in whom urine was not obtained had NIP ratios of 0.69 or less and were, therefore, classified as slow acetylators. The N/P ratio of the remaining patient was 1.5, and he was classified as a fast acetylator. Both NIP ratios and ClAp values were valid indices of acetylator phenotype in patients with no renal impairment. However, the NIP ratio in six of eight patients with renal impairment ranged between 1.0 and 2.4, consistent with fast acetylators, while the ClAp of P A in three of these six patients was less than 80 ml/min, indicating slow acetylator phenotype. The marked dependence of NIP ratios on renal function is illustrated in Fig. 2. The NIP ratios at 24 hr of patients A. M. and L. A. were 1.2 and 0.5, respectively, while their respective ClAp values were 183 and 55 ml/min. These are consistent with the Cla values of the entire group for fast and slow

w Z LLI

6

__V w UJ

Z

2

0

1 0

80

160

CLAp, m i / m i n

Fig. 3. Frequency distribution of the apparent acetylation clearance (C1Ae)of procainarnide.

78

Lima, Conti, Goldfarb, Tilstone, Golden, and Jusko

acetylators. In contrast, the NIP ratios at 24 hr for patients E. W. and J. G., both of whom had renal impairment, were 2.0 and 1.5, while their respective ClAp values were 43 ml/min and 25 ml/min. Thus both patients were classified as fast acetylators by using NIP ratios or slow acetylators by using ClAp as the index of acetylation. The renal clearances of N A P A and P A comprise 85% and 50% of their respective total body clearances. Consequently, serum concentrations of N A P A rise disproportionately to P A in patients with renal impairment and slow acetylators may be misclassified as fast acetylators when N / P ratios are employed, as was the case with patient J. G. The computer estimates of CIA for 15 patients are shown in the histogram of Fig. 4. Eight patients were classified as slow acetylators, six as fast acetylators, and one as intermediate. Patient E. W. was misclassified as a slow acetylator according to ClAp because N A P A had not yet achieved steady state due to renal impairment. The phenotype of one patient classified as intermediate by ClAp was similarly classified according to C1A, and according to NIP ratios (0.94). Because the CIT and CIA of P A normalized for body weight (105 kg) resembled slow acetylators as a group, the pharmacokinetic data obtained in this patient were treated accordingly for sake of brevity. This is valid since deletion of the pharmacokinetic data obtained in this patient did not alter

co l.z 4 84 U.

0 rr w 2-

7

I

0

I

120

i

f

I

240

36O

CL A. rnl/min

Fig. 4. Frequency distribution of the acetylation clearance (CIA)Ofprocainamide.

Clinical Pharmacokinetics of Procainamide Infusions

79

the relationships observed between slow and fast acetylators, and separate treatment was unremarkable. PA Pharmacokinetics

The pharmacokinetic p a r a m e t e r s for P A are summarized for the 21 patients in Table II. The interpatient variation was considerable, with a m i n i m u m coefficient of variation (sD/mean) of 48%. The results of a two-way analysis of variance ( A N O V A : two-way analysis with unequal numbers, ref. 18) of the pharmacokinetics of P A in relation to the two m a j o r factors are summarized in Table III. As described further below, both renal function and acetylator p h e n o t y p e accounted for much of the variability in P A clearances. Renal Function

Eight patients had renal impairment. T h e m e a n serum creatinine and B U N were 2.6 (1.3) and 48 (20) mg/dl, respectively, and were higher than the m e a n serum creatinine and B U N of 1.2 (0.16) and 18 (5) mg/dl, respectively, in patients without renal i m p a i r m e n t (p < 0.005). The effect of Table I!. Pharmacokinetic Parameters of Pro-

cainamide in 21 Patients ~ Volumes of distribution(liters/kg) Vp Vt V~s Vdt3

0.58 2.23 2.40 2.60

(0.27) (1.86) (1.2) (1.3)

Clearances (liters/kg/hr) Clr C112

0.35 2.52

(0.15) (1.24)

Rate constants (hr-l) a /~ kl 2 k21 kel

7.59 0.168 5.27 1.74 0.683

(4.0) (0.12) (3.28) (0.97) (0.34)

Half-life (hr) ll/2t3

6.72

(6.18)

~See text for definitions. Standard deviations in parentheses.

80

Lima, Conti, Goldiarb, Tilstone, Golden, and Jusko

Table IIl. Two-Way Analysis of Variance (ANOVA) of PA Pharmacokinetics in Relation to Renal Function and Acetylator Phenotype Renal function

(P value) Level of significance

Pharmacokinetic parameter mean (SD)

CIT(liters/kg/hr) Vp (liters/kg)

tl/213 (hr) V ~ (liters/kg) Vdt 3 (liters/ks) kl0 (hr -1) C1R(liters/kg/hr) a CIA (liters/kg/hr) b

Impairment Slow

No impairment

Fast

0.21 (0.07) 0.63 (0.22) 13,0 (11) 3,30 (1.8) 3,40 (2.1) 0.40 (0.21) 0.06 (0.03) 0.06 (0.02)

0.29(0.05) 0.63 (0.40) 5.84 (3.5) 2.10 (I,0) 2.30 (1.1) 0.58 (0.33) 0.08(0.04) 0.19 (0,03)

Slow

Renal

Acetylator status

Interaction

Clinical pharmacokinetics of procainamide infusions in relation to acetylator phenotype.

Journal of Pharmacokinetics and Biopharmaceutics, Vol. 7, No. 1, 1979 Clinical Pharmacokinetics of Procainamide Infusions in Relation to Acetylator P...
866KB Sizes 0 Downloads 0 Views