- Email: [email protected]
Pharmacodynamic and pharmacokinetic rationales for the development of an oral controlled-release amoxicillin dosage form
Journal of Controlled Release 54 (1998) 29–37
Pharmacodynamic and pharmacokinetic rationales for the development of an oral controlled-release amoxicillin dosage form a, b a a Amnon Hoffman *, Haim D. Danenberg , Ifat Katzhendler , Rivka Shuval , a a Dalia Gilhar , Michael Friedman a
Department of Pharmaceutics, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, P.O. Box 12065, Jerusalem 91120, Israel b Division of Medicine, Hadassah University Hospital, Jerusalem, Israel Received 4 March 1997; received in revised form 23 July 1997; accepted 30 July 1997
Abstract The goal of this investigation was to develop an oral sustained-release formulation for amoxicillin that would maximize the duration of active drug concentration in the extracellular fluid, thus increasing the dosing interval while assuring antimicrobial activity. This rationale is based on the pharmacodynamic properties of the drug which is non- concentration dependent on the one hand, while requiring long exposure of the pathogen to the drug with minimal post-antibiotic effect on the other. Due to pharmacokinetic constraints, including short biological half-life and limited ‘absorption window’ (confined to the small intestine) with poor colonic absorption, the new matrix tablet formulation, composed of hydrophilic (hydroxypropyl methyl-cellulose) polymer, was designed to release 50% of its contents within the first 3 h and to complete the drug release process over 8 h (under in vitro conditions). The pharmacokinetics of the new formulation was evaluated in 12 healthy volunteers and compared to a conventional gelatin capsule with both formulations containing 500 mg amoxicillin. The plasma concentrations of active amoxicillin and penicilloic acid were determined by an HPLC method with a fluorometric detector. It was found that the area under the concentration–time curve and maximal serum amoxicillin concentrations following the sustained release preparation were lower than the immediate release formulation. However, the time over the required threshold concentrations, i.e. the minimal inhibitory concentration (MIC) as well as the more clinically relevant parameter — four times MIC of the drug against susceptible pathogens, was found to be maintained for significantly longer periods. The results suggest that in order to achieve a twice daily dosing regimen that will provide therapeutic concentrations for the whole 12 h dosing intervals, a larger dose of the new formulation should be given (e.g. 750 mg or even 1g twice daily). This recommendation is based on the large interindividual differences of the extent of amoxicillin absorption found in this investigation, and is intended to assure that the ‘poor’ absorbers will also benefit from full antibiotic efficacy. This dosing regimen will lead to increased patient compliance and improved therapeutic outcome. 1998 Elsevier Science B.V. Keywords: Pharmacodynamics; Pharmacokinetics; Beta-lactam antibiotics; Amoxicillin; Controlled release; Emax model
*Corresponding author. Tel.: 1972 2 6758667; fax: 1972 2 6436246; e-mail: [email protected] 0168-3659 / 98 / $19.00 1998 Elsevier Science B.V. All rights reserved. PII: S0168-3659( 97 )00165-X
30
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
1. Introduction
1.1. Pharmacodynamic rationale
The development of an improved pharmaceutical dosage form, such as an oral sustained release preparation, should be based upon the pharmacokinetic and pharmacodynamic properties of the drug. In addition, issues such as minimization of adverse drug reactions, prevention of the development of resistant organisms (in the case of antimicrobial therapy), patient compliance factors and overall treatment cost must be considered. While much emphasis is given to pharmacokinetic parameters such as absorption characteristics, protein binding and clearance, less concern is given during the development stage to the pharmacodynamic profile of the drug, i.e. the concentration–effect relationship. In the case of antibiotic agents, this relationship depends on three elements: the pathogen, the host and the specific antimicrobial agent. The impact of the host, apart from the individual pharmacokinetic properties, depends mainly on its immune system. The relationship between drug concentration and its inhibitory effect on microbial growth, for a certain drug–pathogen combination, can be determined in vitro. While in certain cases (e.g. aminoglycoside antibiotics) elevation of drug concentration is associated with enhanced bactericidal potency, other antibiotic agents (such as betalactam antibiotics and erythromycin [1]) are not highly concentration dependent [2]. The post-antibiotic effect (PAE) is another pharmacodynamic parameter that has to be regarded in determining an optimal dosage regimen. The extrapolation of the in vitro data to an in vivo situation is less complex when the pathogen is located extracelluarly, as in the case of beta-lactam susceptible microorganisms. The specific goal of this investigation was to develop an improved oral dosage form for amoxicillin, a representative beta-lactam antibiotic. Based on the pharmacodynamic and pharmacokinetic properties of the drug (outlined below) it was concluded that an oral sustained-release formulation fits the therapeutic goals of antimicrobial therapy with this drug by assuring effective drug concentrations for a prolonged period. The rationale leading to this conclusion is based on the following points:
(1) Elevation of beta-lactam concentration demonstrates increased bacterial killing, only until a finite point which tends to be about four times the minimal inhibitory concentration (MIC). This value will be denoted in this paper as the ‘therapeutic concentration’. Further elevation is not associated with increased bactericidal potency [3]. It had been also suggested that at concentrations much greater than the MIC a paradoxical pattern may occur, i.e. decrease in bacterial kill potency [2,4]. (2) Tissue penetration of these drugs is not correlated with serum concentrations, i.e. elevation of serum drug concentration will not contribute much in cases where the pathogen is located intracellularly. (3) A direct correlation exists between the time the beta-lactam antibiotic concentrations are maintained above the therapeutic concentration and the clinical outcomes [5,6]. It has been confirmed that for beta-lactams, continuous infusion has clinical advantages over an intermittent mode of administration [2]. Furthermore, a smaller total antibiotic dose is required to achieve the same pharmacodynamic endpoint by continuous infusion in comparison to intermittent infusion [7,8]. (4) There is no correlation between the area under the concentration vs. time curve (AUC) and the magnitude of effect in vivo. (5) Unlike aminoglycosides, the kinetics of bactericidal effect is slow and requires prolonged maintenance of effective concentrations of the drug in order to achieve onset of effect. (6) Amoxicillin, like many other beta lactam antibiotics, exhibits no or only a very short PAE against Enterobacteriaceae or Pseudomonas species. Bacterial regrowth occurs rapidly after these antibiotic concentrations fall below the bacterial MIC [9,10]. Therefore, the goal of a dosage regimen for each individual beta-lactam should be to prevent the drug-free interval between doses from being long enough for the bacterial pathogen to resume growth. (7) Adverse drug reactions caused by beta-lactam antibiotics are not specifically related to constant serum or tissue concentrations. Furthermore, continuous administration may result in lower toxicity than
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
the administration of large and potentially toxic bolus doses.
1.2. Pharmacokinetic rationale (1) Beta-lactam antibiotics exhibit short half-life values, which demand frequent drug administration. Therefore, continuous infusion has been suggested as the most beneficial mode of administration for the majority of beta-lactams [2,5,6]. (2) Due to the fact that amoxicillin, like many other beta-lactams, is mostly active against extracelluarly located microbial infections, and its binding to serum proteins is low, there is a direct correlation between its concentration at the infection site at the extravascular fluids and its serum concentration. Therefore, a realistic link can be drawn between the in vivo serum concentration and the in vitro measured parameter MIC, in this case. Based on these arguments, continuous infusion was selected as the preferred dosing strategy for beta-lactam antibiotics (A Consensus Document by the PharmPK List [email protected], Valencia, Spain). However, this approach has certain limitations, of which the most problematic issues are the patient inconvenience due to decreased mobility, increased rate of bacteremias due to intravenous line infection, increased costs and the need for trained medical personnel for antibiotic administration and the overall cost due to the necessity for hospitalization during parenteral antimicrobial therapy. The idea of this investigation was to develop an oral controlled release preparation that would expand the time over MIC of amoxicillin in vivo in relation to a regular formulation of gelatin capsules, thereby obtaining all the advantages of continuous intravenous infusion while avoiding the above limitations. The development of such an oral formulation must take into account the characteristic gastrointestinal (GI) absorption properties of amoxicillin with poor colonic absorption [11] and non-linear absorption kinetics [12]. 2. Materials and methods
2.1. Preparation of tablets Cylindrical tablets were prepared by direct com-
31
pression of amoxicillin–polymer blends, using a laboratory press fitted with a 12-mm flat faced punch and die set, applying a pressure of 5 tons. The tablets contained 603 mg amoxicillin trihydrate (equivalent to 500 mg amoxicillin) and 17% Methocel K100LV, Dow, USA.
2.2. In vitro dissolution studies The dissolution rates of amoxicillin from the controlled release matrix tablets and from the regular gelatin capsule formulation were monitored using a tablet dissolution tester (Model 7ST Caleva, USA), according to the USP XXI basket method, under the following conditions: rotation speed was 100 rpm; release medium was 900 ml phosphate buffer pH 7.4 [13], maintained at 378C. The effect of the certain variables on the rate of drug release were examined including: (1) medium pH, 2.0, 6.8, 7.4 (2) compression pressure, 1, 3, 5 and 8 tons (3) basket rotation speed, 50, 100, 200.
2.3. Analytical procedure Amoxicillin levels were monitored spectrophotometrically (Uvikon 930 Kontron spectrophotometer, Switzerland) at 272 nm. Dissolution studies were performed in triplicate.
2.4. Pharmacokinetic evaluation Twelve healthy male volunteers, aged 24–30, participated in the in vivo evaluation of the new amoxicillin controlled release (AMOX-CR) formulation versus the standard amoxicillin 500 mg gelatin capsule formulation (Moxyvit Forte, Vitamed, Binyamina, Israel). All gave informed consent to participation and the study protocol was approved by the Helsinki Committee of the Hadassah Medical Center and Israeli Ministry of Health. The study was conducted in a randomized, two-way cross-over fashion. Each amoxicillin formulation was administered at seven a.m. following an overnight fast. Food was withheld for 5 h after the administration of the drug. Tea, coffee, cola and other caffeinated beverages and
32
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
food were not permitted from 1 day before until the end of each study phase. Venous blood samples (8 ml) were taken via an indwelling catheter in the forearm vein 0, 0.5, 1, 1.5, 2, 3, 4, 5, 6, 7, 8, 9, 10 and 13 h after dosing. The plasma was immediately separated by centrifugation at 3000 g for 15 min and stored at 2208C pending assay.
2.5. Analytical procedure of biological samples Amoxicillin plasma levels were determined by an HPLC method with a fluorometric detector. The assay was based on the methodology described by Miyazaki [14] with certain modifications. The detection was performed at excitation wavelength of 355 nm and emission wavelength of 435 nm. The separation was carried out through a C 18 column (Econosphere, 5m, 15034.6. mm, Alltech, Deerfield, IL, USA), at a temperature of 558C. The mobile phase consisted of methanol–distilled water (55:45, v / v) at a flow-rate of 1 ml / min. Samples were thawed at room temperature and the concentration of total amoxicillin (i.e. active amoxicillin1penicilloic acid) and penicilloic acid alone were determined. Active amoxicillin concentrations were calculated from the difference between the total amoxicillin and penicilloic acid concentrations. The procedure for formation of the fluorescent degradation products in plasma followed the specific procedure outlined by Miyazaki [14], apart from the fact that only half of the amounts of plasma and reagents were used. The method is specific and sensitive, with a quantifiable limit of 10 ng / ml. The precision is 5% relative standard deviation and accuracy is 610%.
the plasma concentration–time curve. The rate of amoxicillin absorption following administration of each of the amoxicillin formulations was assessed by the peak plasma concentration (Cmax ) and the time (t max ) to reach Cmax . Since it was suggested that in the case of sustained release preparations the Cmax value normalized by the AUC value (i.e. Cmax /AUC) is a more valid parameter than Cmax [15] for comparison between formulations, this parameter was also calculated. The time over certain therapeutic concentrations (defined according to specific MIC values of amoxicillin-sensitive microorganisms) was also calculated.
2.7. Statistical analysis The results of this investigation were analyzed statistically by the non-parametric paired Wilcoxon test. All calibration curves in the various analyses of the in vitro and in vivo evaluation processes had a correlation coefficient of at least 0.997.
3. Results The in vitro dissolution profiles of the new formulation AMOX-CR and the regular gelatin capsule formulation in phosphate buffer (pH 7.4) at 378C are illustrated in Fig. 1. The new formulation
2.6. Pharmacokinetic analysis The extent of amoxicillin and penicilloic acid absorption was assessed by the AUC from time 0 that was calculated by the standard equation: AUC(0 2 2 `) 5 AUC(0 2 2 t) 1 Ct /b in which AUC(0–t) is the area under the curve up to the last observable plasma concentration (Ct ) obtained by the linear trapezoidal method, and b is the slope of ln-linear regression of the terminal phase of
Fig. 1. Percent of amoxicillin released from gelatin capsule formulation (open squares) and from the controlled release matrix tablet preparation (full squares).
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
33
was designed to release 50% of its contents within the first 3 h and to complete the drug release process over 8 h under in vitro conditions according to the USP specifications [13]. The effect of various parameters on the dissolution rate of amoxicillin from the new formulation was also assessed in this investigation. No statistically significant differences were found between the release rates of amoxicillin from the controlled release formulation tested under different pH conditions. Similarly, the extent of compression pressure was not found to affect amoxicillin release rate from the new formulation. On the other hand, there was a significant effect of the basket rotation speed on the release rate from the tablets. Increasing the rotation speed from 50 rpm to 100 rpm increased the erosion constant of the tablet 1.6 times, and 2.9 times higher at 200 rpm.
3.1. In vivo evaluation The mean amoxicillin concentration vs. time profile of active amoxicillin is presented in Fig. 2, and the pharmacokinetic parameters are summarized in Table 1. As can be seen, lower Cmax values (about 56% of the standard preparation) where found for the controlled release preparation and these Cmax value were attained much slower. On the other hand, although the AUC values tend to be lower for the AMOX-CR preparation, this difference is not statistically significant and the mean of the individual ratios of new / regular formulations AUC values for each volunteer is 0.92. Thus, there are no substantial differences in the extent of amoxicillin absorption between the regular formulation and the controlled release matrix tablets. Amoxicillin is degraded to an inactive metabolite penicilloic acid. The mean plasma penicilloic acid concentration versus time following administration of the two amoxicillin formulations is presented in Fig. 3. The mean AUC value of the controlled release preparation was found to be 23% lower than that of the standard preparation. The MIC values of representative amoxicillinsusceptible pathogens are outlined in Table 2, together with the more clinically relevant parameter of four times MIC. The time period in which amoxicillin concentrations were found to be over certain
Fig. 2. Serum active amoxicillin concentration following administration of 500 mg amoxicillin in gelatin capsule formulation (open squares) and controlled release matrix tablet preparation (full squares). The stripped zone present the difference between the time over a theoretical MIC value (0.2 mg / ml) produced by the two formulations.
‘therapeutic concentration’ levels required against every pathogen in each volunteer were measured and the mean values are summarized in Tables 2 and 3. It was found that the controlled release preparation extended the ‘time over MIC’. This is also illustrated in Fig. 4 in which no real difference between the onset time (i.e. the time following administration Table 1 Minimum Inhibitory concentration (MIC) values of amoxicillin against common pathogens susceptible to the drug Pathogen
MIC (mg / ml)
MIC34 (mg / ml)
Strep. group A Strep. pneumoniae Staph. aureus H. influenzae
0.01 0.02 0.1 0.25
0.04 0.08 0.4 1
34
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
Fig. 3. Serum penicilloic acid concentration following administration of 500 mg amoxicillin in gelatin capsule formulation (open squares) and controlled release matrix tablet preparation (full squares).
Table 2 Mean time (h) over certain MIC values produced by the standard and the sustained release preparations a Pathogen
Sustained release
Gelatin capsule
Strep. group A Strep. pneumoniae Staph. aureus H. influenzae
13.461.4 * 13.561.5 * 11.761.7 * 9.962.3 *
12.061.9 12.061.9 10.261.4 8.561.8
a
Results reported as mean6S.D. (n512). * Statistically significant difference between the sustained release formulation and the gelatin capsule formulation by Wilcoxon test (P,0.05).
Table 3 Mean time (h) over certain MIC34 values produced by the standard and the sustained release preparations a Pathogen
Sustained release
Gelatin capsule
Strep. group A Strep. pneumoniae Staph. aureus H. influenzae
13.061.9 * 12.662.1 * 9.262.0 8.562.8 *
11.062.0 10.761.9 8.061.5 5.661.4
a
Results reported as mean6S.D. (n512). * Statistically significant difference between the sustained release formulation and the gelatin capsule formulation by Wilcoxon test (P,0.05).
Fig. 4. Semi-logarithmic plot of serum active amoxicillin concentration following administration of 500 mg amoxicillin in gelatin capsule formulation (open squares) and controlled release matrix tablet preparation (full squares). The plot exhibits the differences in time over certain MIC values.
required to get the ‘therapeutic concentration’) of the two preparations is observable.
4. Discussion The development of an improved amoxicillin formulation must follow the pharmacodynamic rationale which, in the case of beta-lactam antibiotics, is to extend the ‘time over MIC’. Achieving this goal in vivo is complicated by the pharmacokinetic constraints which include rapid elimination of amoxicillin with a half-life of about 1 h, and a limited ‘absorption window’ following peroral
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
administration. As shown by Barr et al. [11], in humans the drug is well absorbed in the duodenum and jejunum, but absorption is decreased and ratedependent in the ileum, where more drug is absorbed following slow infusion versus bolus, and absorption is poor in all colon regions. Since the kinetics of drug elimination does not depend on the pharmaceutical delivery system, a rational method of prolonging the ‘time over MIC’ is to extend the absorption phase by slowly releasing the drug in the upper parts of the GI tract. The goal of the development stage, therefore, was to prepare a controlled release matrix tablet that will release, in in vitro dissolution tests, 50% of the drug within 3 h, followed by a constant release rate which will be completed after about 8 h [16]. The rapid onset of drug release was designed to provide an initial ‘loading dose’ and to maximize the absorption phase in those parts of the intestine in which amoxicillin is absorbed by a carrier-mediated process [12,17]. As found in the in vivo evaluation of the new formulation, the extent of absorption of the new formulation is not much different than that of a regular soft gelatin capsule formulation. Furthermore, the time required to obtain therapeutic concentration (‘onset time’) was found to be identical for the two formulations. The matrix tablet formulation developed is based on a hydrophilic (hydroxypropyl methyl-cellulose) polymer. As previously described [18], the kinetics of drug release from these matrix tablets follows the erosion mechanism [19,20]. Accordingly, variations in the releasing medium, such as pH changes, do not affect the kinetics of drug release. Since the erosion mechanism is controlled by hydration and gelation of the outer matrix layer, the reduced porosity due to increased compression pressure does not affect the rate of erosion, and the rate of drug release is therefore also unaffected. On the other hand, the rate of drug release in this delivery system by the erosion mechanism is expected to be affected by the rotation speed, as was found experimentally in this work. The pharmacodynamics of the bactericidal activity of amoxicillin is not concentration-dependent [6]. The concentration–antimicrobial activity relationship of beta- lactam antibiotics can be theoretically described according to the sigmoidal Emax model equation:
35
Emax C n E 5 E0 1 ]]] EC n50 1 C n The magnitude of effect E reaches the maximal effect Emax at relatively low concentrations. E0 is the magnitude of effect without the drug, which in this case is the antimicrobial effect of the immune system. The parameter n is the shape factor which is derived from the distribution function of the sensitivity of the microorganisms to the bactericidal effect of the drug [21]. In this case the value of n is expected to be high (i.e. close to 5) according to two approaches: (a) the concentration range between the minimal effective concentration (MIC) and the minimal concentration required to achieve Emax (CEmax ) is known to be low for these drugs; (b) the relationship between concentration and effect of each individual effector unit, which in this case is each microorganism, is an all-or-none pattern, and therefore, since the variability in degree of sensitivity of the microorganism population to the drug is expected to be low, n value has to be high in order to construct a steep concentration effect profile [21]. A theoretical pharmacodynamic profile of beta-lactam antibiotics drawn according to these parameters is illustrated in Fig. 5. It this case EC50 (i.e. the concentration at which 50% of the maximal effect is achieved) is a concentration slightly above MIC (or more accurately, above the minimal bactericidal concentration (MBC)). The importance in realizing
Fig. 5. Theoretical pharmacodynamic profile of beta-lactam antibiotics against susceptible pathogens. The scheme illustrate the relative narrow margin between drug concentration that do not contribute antimicrobial effect (less than MIC) and the concentration required to achieve maximal activity (Emax ).
36
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
the pharmacodynamic profile is that it indicates the optimal dosing regiment of the drug. Clarifying that high concentrations exceeding CEmax do not hold any additional therapeutic effect. It emphasizes the fact that the significantly greater Cmax values found in this investigation for the gelatin capsules in comparison to the new formulation are not associated with a larger magnitude of effect. In many cases the pharmacodynamic properties of the drug are not taken into consideration appropriately during the development stage while the pharmacokinetic parameters are over-emphasized. A relevant example is the work of Hilton and Deasy [22] who developed controlled release tablets of amoxicillin trihydrate with hydroxypropyl methylcellulose acetate succinate enteric polymer matrix and criticized the new product because of its lower AUC values (64.4%) in comparison to a proprietary product. From a pharmacodynamic point of view the comparison between the conventional formulation and the new formulation has to focus on the differences in the area under the required therapeutic concentration, as previously suggested by Schentag et al. [23]. According to this approach a line was drawn at a concentration of 0.2 mg / ml (Fig. 2) (which is ten times the MIC of Strep. pneumoniae and five times the MIC of Staph. aureus) and the difference between the two products regarding the area under this set concentration was highlighted by striped lines. This striped zone is the essence of the difference between the therapeutic efficacy of the conventional and AMOX-CR formulations. The ‘time over MIC’ depends on the specific pathogen that has to be treated. For those pathogens with relatively low sensitivity to amoxicillin, e.g. MIC.0.25 mg / ml, it would be better to select a more suitable antibiotic medication, because the time over MIC34 with either formulation will be too short, which will lead to regrowth of resistant strains and resultant therapeutic failure. On the other hand, as shown in Tables 2 and 3 and in Fig. 4 for amoxicillin-sensitive pathogens, the new controlled release formulation provided extended time over MIC as well as MIC34 levels. Extending the residence time of the antibiotic drug in the GI tract by the sustained release formulation may increase, in theory, the GI adverse effects associated with amoxicillin therapy. Such phenomenon of increased risk of side effect would be an
obvious limitation for the development of AMOXCR formulation. However, previous clinical investigations have demonstrated that long acting amoxicillin formulations used for therapy of upper respiratory tract infections [24] and urinary tract infection [25] did not increase side effects. In addition, the unabsorbed portion of the dose that had prolonged transit time in the GI tract due to the sustained release formulation is captured in the matrix formulation and is not available to interact with the intestine epithelia and / or flora. This notion is based on the lower degree of presystemic degradation of amoxicillin in the sustained release formulation in comparison to gelatin capsule formulation, shown in Fig. 3. A single matrix tablet of 500 mg amoxicillin was compared in this investigation with an equivalent dose of conventional capsule. In order to achieve a twice daily dosing regimen that will provide therapeutic concentrations for the whole 12 h dosing intervals a larger dose of the AMOX-CR should be given (750 mg or even 1 g twice daily). This recommendation is based on the large interindividual differences in the extent of amoxicillin absorption found in this investigation, and is intended to assure that the ‘poor’ absorbers will also benefit from full antibiotic efficacy. However, the validity of this recommendation should be validated experimentally. There are clear advantages to twice daily versus thrice daily dosing regimens. Such a regimen significantly increases patient compliance and thereby improves therapeutic outcome [26]. Like other dosing regimens which incorporate pharmacodynamic principles, it may lead to decreased cost of therapy, reduce the incidence of adverse drug reactions and improve or at least equal clinical efficacy.
Acknowledgements The work was supported by Vitamed, Binyamina, Israel Ltd. Dr. A. Hoffman and Prof. M. Friedman are affiliated with the David R. Bloom Center for Pharmacy.
References [1] N. White, T. Davis, Anti-ineffective drugs, in: C. van Boxtel,
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
[2]
[3] [4] [5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13] [14]
N. Holford, M. Danhof (Eds.), The in vivo study of drug action, Elsevier, Amsterdam, 1992, pp. 311–334. W. Craig, S. Ebert, Continuous infusion of beta-lactam antibiotics. Antimicrob. Agent. Chemother. 36 (1992) 2577– 2583. W. Craig, S. Ebert, Killing and regrowth of bacteria in vitro: a review. Scand. J. Infect. Dis. 74(Suppl.) (1991) 63–70. S. Holm, I. OdenholtTornqvist, O. Cars, Paradoxical effects of antibiotics, Scand. J. Infect. Dis. 74 (1991) 113–117. G. Drusano, Role of pharmacokinetics in the outcome of infections, Antimicrob. Agent. Chemother. 32 (1988) 289– 297. B. Vogelman, S. Gudmundsson, J. Leggett, J. Turnidge, S. Ebert, W. Craig, Correlation of antimicrobial pharmacokinetic parameters with therapeutic efficacy in an animal model, J. Infec. Dis. 158 (1988) 831–847. R. Roosendaal, I. Bakker-Woundenberg, J. van der Berg, M. Michel, Therapeutic efficacy of continuous versus intermittent administration of ceftazidime in an experimental Klebsiella pneumonia pneumonia in rats, J. Infec. Dis. 152 (1985) 373–378. J. Zeisler, J. McCarthy, W. Richelieu, M. Nichol, Cefuroxime by continuous infusion: a new standard of care?, Infect. Med. 9 (1992) 54–60. W. Craig, S. Gudmundsson, The post-antibiotic effect, in: V. Lorian (Ed.), Antibiotics in laboratory medicine, Williams and Wilkins, Baltimore, 1986, pp. 515–536. J. Rotschafer, K. Walker, K. Madaras-Kelly, C. Sullivan, Antibiotic pharmacodynamics, in: N. Cutler, J. Sramek, P. Narang (Eds.), Pharmacodynamics and drug development: Perspectives in clinical pharmacology, John Wiley, West Sussex, 1994, pp. 315–343. W. Barr, M. Zola, E. Candler, H. Shie-Ming, A. Tendolkar, R. Shamburek, B. Parker, M. Hilty, Differential absorption of amoxicillin from the human small and large intestine, Clin. Pharmacol. Ther. 56 (1994) 279–285. F. Torres-Molina, J. Peris-Ribera, M. Garcia-Carboneli, J. Aristorena, J. Pla-Delfina, Nonlinearities in amoxycillin pharmacokinetics. II absorption studies in the rat, Biopharm. Drug Dispos. 13 (1992) 39–53. USP. XXI, 1985. K. Miyazaki, K. Ohtani, K. Sunada, T. Arita, Determination of ampicillin, amoxicillin, cephalexin and cephradine in plasma by high-performance liquid chromatography using fluorometric detection. J. Chromatog. 276 (1983) 478–482.
37
[15] M. Bialer, S. Sussan, O. Abu Salach, H.D. Danenberg, J. Ben David, Y. Gibor, A. Laor, Criteria to assess in vivo performance of sustained release products: application to diltiazem formulations, J. Pharm. Sci. 84 (1995) 1160–1163. [16] I. Katzhendler, A. Hoffman, M. Friedman, Controlled-release pharmaceutical preparations, Israeli patent application No. 119627 (1996). [17] P.J. Sinko, G.L. Amidon, Characterization of the oral absorption of beta- lactam antibiotics. II. Competitive absorption and peptide carrier specificity, J. Pharm. Sci. 78 (1989) 723–727. [18] Y. Katzhendler, A. Hoffman, A. Goldberg, M. Friedman, Modeling of drug release from erodible tablets, J. Pharm. Sci. 86 (1997) 110–115. [19] J. Heller, R. Baker, R. Gale, O. Rodin, Controlled drug release by polymer dissolution. I. Partial ester of maleic anhydride copolymers — properties and theory. J. Appl. Polym. Sci. 22 (1978) 1991–2009. [20] J. Heller, Controlled release of biologically active compounds from bioerodible polymers, Biomaterials 1 (1980) 51–57. [21] A. Hoffman, A. Goldberg, The relationship between receptor-effector unit heterogeneity and the shape of the concentration-effect profile: pharmacodynamic implications, J Pharmacokinet. Biopharm. 22 (1994) 449–468. [22] A.K. Hilton, P.B. Deasy, Use of hydroxypropyl methylcellulose acetate succinate in an enteric polymer matrix to design controlled-release tablets of amoxicillin trihydrate, J. Pharm. Sci. 82 (1993) 737–743. [23] J.J. Schentag, D.E. Nix, M.H. Adelman, Mathematical examination of dual individualization principles (I): Relationships between AUC above MIC and area under the inhibitory curve for cefmenoxime, ciprofloxacin and tobramycin, Drug Intel. Clin. Pharm. 25 (1991) 1050–1057. [24] N. Yoshihito, K. Masayoshi, Y. Naoyuki, S. Rinzo, Clinical studies of long acting amoxicillin granules. Japan J. Antibiotics (1983) 428–432. [25] M. Ktsuyuki, N. Akira, T. Akihiko, Clinical study on long acting amoxicillin in urinary tract infection. Japan J. Antibiotics (1983) 408–413. [26] J. Cockburn, A.L. Reid, J.A. Bowman, R.W. Sanson Fisher, Effects of intervention on antibiotic compliance in patients in general practice, Med. J. Aust 147 (1987) 324–328.
Pharmacodynamic and pharmacokinetic rationales for the development of an oral controlled-release amoxicillin dosage form a, b a a Amnon Hoffman *, Haim D. Danenberg , Ifat Katzhendler , Rivka Shuval , a a Dalia Gilhar , Michael Friedman a
Department of Pharmaceutics, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, P.O. Box 12065, Jerusalem 91120, Israel b Division of Medicine, Hadassah University Hospital, Jerusalem, Israel Received 4 March 1997; received in revised form 23 July 1997; accepted 30 July 1997
Abstract The goal of this investigation was to develop an oral sustained-release formulation for amoxicillin that would maximize the duration of active drug concentration in the extracellular fluid, thus increasing the dosing interval while assuring antimicrobial activity. This rationale is based on the pharmacodynamic properties of the drug which is non- concentration dependent on the one hand, while requiring long exposure of the pathogen to the drug with minimal post-antibiotic effect on the other. Due to pharmacokinetic constraints, including short biological half-life and limited ‘absorption window’ (confined to the small intestine) with poor colonic absorption, the new matrix tablet formulation, composed of hydrophilic (hydroxypropyl methyl-cellulose) polymer, was designed to release 50% of its contents within the first 3 h and to complete the drug release process over 8 h (under in vitro conditions). The pharmacokinetics of the new formulation was evaluated in 12 healthy volunteers and compared to a conventional gelatin capsule with both formulations containing 500 mg amoxicillin. The plasma concentrations of active amoxicillin and penicilloic acid were determined by an HPLC method with a fluorometric detector. It was found that the area under the concentration–time curve and maximal serum amoxicillin concentrations following the sustained release preparation were lower than the immediate release formulation. However, the time over the required threshold concentrations, i.e. the minimal inhibitory concentration (MIC) as well as the more clinically relevant parameter — four times MIC of the drug against susceptible pathogens, was found to be maintained for significantly longer periods. The results suggest that in order to achieve a twice daily dosing regimen that will provide therapeutic concentrations for the whole 12 h dosing intervals, a larger dose of the new formulation should be given (e.g. 750 mg or even 1g twice daily). This recommendation is based on the large interindividual differences of the extent of amoxicillin absorption found in this investigation, and is intended to assure that the ‘poor’ absorbers will also benefit from full antibiotic efficacy. This dosing regimen will lead to increased patient compliance and improved therapeutic outcome. 1998 Elsevier Science B.V. Keywords: Pharmacodynamics; Pharmacokinetics; Beta-lactam antibiotics; Amoxicillin; Controlled release; Emax model
*Corresponding author. Tel.: 1972 2 6758667; fax: 1972 2 6436246; e-mail: [email protected] 0168-3659 / 98 / $19.00 1998 Elsevier Science B.V. All rights reserved. PII: S0168-3659( 97 )00165-X
30
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
1. Introduction
1.1. Pharmacodynamic rationale
The development of an improved pharmaceutical dosage form, such as an oral sustained release preparation, should be based upon the pharmacokinetic and pharmacodynamic properties of the drug. In addition, issues such as minimization of adverse drug reactions, prevention of the development of resistant organisms (in the case of antimicrobial therapy), patient compliance factors and overall treatment cost must be considered. While much emphasis is given to pharmacokinetic parameters such as absorption characteristics, protein binding and clearance, less concern is given during the development stage to the pharmacodynamic profile of the drug, i.e. the concentration–effect relationship. In the case of antibiotic agents, this relationship depends on three elements: the pathogen, the host and the specific antimicrobial agent. The impact of the host, apart from the individual pharmacokinetic properties, depends mainly on its immune system. The relationship between drug concentration and its inhibitory effect on microbial growth, for a certain drug–pathogen combination, can be determined in vitro. While in certain cases (e.g. aminoglycoside antibiotics) elevation of drug concentration is associated with enhanced bactericidal potency, other antibiotic agents (such as betalactam antibiotics and erythromycin [1]) are not highly concentration dependent [2]. The post-antibiotic effect (PAE) is another pharmacodynamic parameter that has to be regarded in determining an optimal dosage regimen. The extrapolation of the in vitro data to an in vivo situation is less complex when the pathogen is located extracelluarly, as in the case of beta-lactam susceptible microorganisms. The specific goal of this investigation was to develop an improved oral dosage form for amoxicillin, a representative beta-lactam antibiotic. Based on the pharmacodynamic and pharmacokinetic properties of the drug (outlined below) it was concluded that an oral sustained-release formulation fits the therapeutic goals of antimicrobial therapy with this drug by assuring effective drug concentrations for a prolonged period. The rationale leading to this conclusion is based on the following points:
(1) Elevation of beta-lactam concentration demonstrates increased bacterial killing, only until a finite point which tends to be about four times the minimal inhibitory concentration (MIC). This value will be denoted in this paper as the ‘therapeutic concentration’. Further elevation is not associated with increased bactericidal potency [3]. It had been also suggested that at concentrations much greater than the MIC a paradoxical pattern may occur, i.e. decrease in bacterial kill potency [2,4]. (2) Tissue penetration of these drugs is not correlated with serum concentrations, i.e. elevation of serum drug concentration will not contribute much in cases where the pathogen is located intracellularly. (3) A direct correlation exists between the time the beta-lactam antibiotic concentrations are maintained above the therapeutic concentration and the clinical outcomes [5,6]. It has been confirmed that for beta-lactams, continuous infusion has clinical advantages over an intermittent mode of administration [2]. Furthermore, a smaller total antibiotic dose is required to achieve the same pharmacodynamic endpoint by continuous infusion in comparison to intermittent infusion [7,8]. (4) There is no correlation between the area under the concentration vs. time curve (AUC) and the magnitude of effect in vivo. (5) Unlike aminoglycosides, the kinetics of bactericidal effect is slow and requires prolonged maintenance of effective concentrations of the drug in order to achieve onset of effect. (6) Amoxicillin, like many other beta lactam antibiotics, exhibits no or only a very short PAE against Enterobacteriaceae or Pseudomonas species. Bacterial regrowth occurs rapidly after these antibiotic concentrations fall below the bacterial MIC [9,10]. Therefore, the goal of a dosage regimen for each individual beta-lactam should be to prevent the drug-free interval between doses from being long enough for the bacterial pathogen to resume growth. (7) Adverse drug reactions caused by beta-lactam antibiotics are not specifically related to constant serum or tissue concentrations. Furthermore, continuous administration may result in lower toxicity than
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
the administration of large and potentially toxic bolus doses.
1.2. Pharmacokinetic rationale (1) Beta-lactam antibiotics exhibit short half-life values, which demand frequent drug administration. Therefore, continuous infusion has been suggested as the most beneficial mode of administration for the majority of beta-lactams [2,5,6]. (2) Due to the fact that amoxicillin, like many other beta-lactams, is mostly active against extracelluarly located microbial infections, and its binding to serum proteins is low, there is a direct correlation between its concentration at the infection site at the extravascular fluids and its serum concentration. Therefore, a realistic link can be drawn between the in vivo serum concentration and the in vitro measured parameter MIC, in this case. Based on these arguments, continuous infusion was selected as the preferred dosing strategy for beta-lactam antibiotics (A Consensus Document by the PharmPK List [email protected], Valencia, Spain). However, this approach has certain limitations, of which the most problematic issues are the patient inconvenience due to decreased mobility, increased rate of bacteremias due to intravenous line infection, increased costs and the need for trained medical personnel for antibiotic administration and the overall cost due to the necessity for hospitalization during parenteral antimicrobial therapy. The idea of this investigation was to develop an oral controlled release preparation that would expand the time over MIC of amoxicillin in vivo in relation to a regular formulation of gelatin capsules, thereby obtaining all the advantages of continuous intravenous infusion while avoiding the above limitations. The development of such an oral formulation must take into account the characteristic gastrointestinal (GI) absorption properties of amoxicillin with poor colonic absorption [11] and non-linear absorption kinetics [12]. 2. Materials and methods
2.1. Preparation of tablets Cylindrical tablets were prepared by direct com-
31
pression of amoxicillin–polymer blends, using a laboratory press fitted with a 12-mm flat faced punch and die set, applying a pressure of 5 tons. The tablets contained 603 mg amoxicillin trihydrate (equivalent to 500 mg amoxicillin) and 17% Methocel K100LV, Dow, USA.
2.2. In vitro dissolution studies The dissolution rates of amoxicillin from the controlled release matrix tablets and from the regular gelatin capsule formulation were monitored using a tablet dissolution tester (Model 7ST Caleva, USA), according to the USP XXI basket method, under the following conditions: rotation speed was 100 rpm; release medium was 900 ml phosphate buffer pH 7.4 [13], maintained at 378C. The effect of the certain variables on the rate of drug release were examined including: (1) medium pH, 2.0, 6.8, 7.4 (2) compression pressure, 1, 3, 5 and 8 tons (3) basket rotation speed, 50, 100, 200.
2.3. Analytical procedure Amoxicillin levels were monitored spectrophotometrically (Uvikon 930 Kontron spectrophotometer, Switzerland) at 272 nm. Dissolution studies were performed in triplicate.
2.4. Pharmacokinetic evaluation Twelve healthy male volunteers, aged 24–30, participated in the in vivo evaluation of the new amoxicillin controlled release (AMOX-CR) formulation versus the standard amoxicillin 500 mg gelatin capsule formulation (Moxyvit Forte, Vitamed, Binyamina, Israel). All gave informed consent to participation and the study protocol was approved by the Helsinki Committee of the Hadassah Medical Center and Israeli Ministry of Health. The study was conducted in a randomized, two-way cross-over fashion. Each amoxicillin formulation was administered at seven a.m. following an overnight fast. Food was withheld for 5 h after the administration of the drug. Tea, coffee, cola and other caffeinated beverages and
32
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
food were not permitted from 1 day before until the end of each study phase. Venous blood samples (8 ml) were taken via an indwelling catheter in the forearm vein 0, 0.5, 1, 1.5, 2, 3, 4, 5, 6, 7, 8, 9, 10 and 13 h after dosing. The plasma was immediately separated by centrifugation at 3000 g for 15 min and stored at 2208C pending assay.
2.5. Analytical procedure of biological samples Amoxicillin plasma levels were determined by an HPLC method with a fluorometric detector. The assay was based on the methodology described by Miyazaki [14] with certain modifications. The detection was performed at excitation wavelength of 355 nm and emission wavelength of 435 nm. The separation was carried out through a C 18 column (Econosphere, 5m, 15034.6. mm, Alltech, Deerfield, IL, USA), at a temperature of 558C. The mobile phase consisted of methanol–distilled water (55:45, v / v) at a flow-rate of 1 ml / min. Samples were thawed at room temperature and the concentration of total amoxicillin (i.e. active amoxicillin1penicilloic acid) and penicilloic acid alone were determined. Active amoxicillin concentrations were calculated from the difference between the total amoxicillin and penicilloic acid concentrations. The procedure for formation of the fluorescent degradation products in plasma followed the specific procedure outlined by Miyazaki [14], apart from the fact that only half of the amounts of plasma and reagents were used. The method is specific and sensitive, with a quantifiable limit of 10 ng / ml. The precision is 5% relative standard deviation and accuracy is 610%.
the plasma concentration–time curve. The rate of amoxicillin absorption following administration of each of the amoxicillin formulations was assessed by the peak plasma concentration (Cmax ) and the time (t max ) to reach Cmax . Since it was suggested that in the case of sustained release preparations the Cmax value normalized by the AUC value (i.e. Cmax /AUC) is a more valid parameter than Cmax [15] for comparison between formulations, this parameter was also calculated. The time over certain therapeutic concentrations (defined according to specific MIC values of amoxicillin-sensitive microorganisms) was also calculated.
2.7. Statistical analysis The results of this investigation were analyzed statistically by the non-parametric paired Wilcoxon test. All calibration curves in the various analyses of the in vitro and in vivo evaluation processes had a correlation coefficient of at least 0.997.
3. Results The in vitro dissolution profiles of the new formulation AMOX-CR and the regular gelatin capsule formulation in phosphate buffer (pH 7.4) at 378C are illustrated in Fig. 1. The new formulation
2.6. Pharmacokinetic analysis The extent of amoxicillin and penicilloic acid absorption was assessed by the AUC from time 0 that was calculated by the standard equation: AUC(0 2 2 `) 5 AUC(0 2 2 t) 1 Ct /b in which AUC(0–t) is the area under the curve up to the last observable plasma concentration (Ct ) obtained by the linear trapezoidal method, and b is the slope of ln-linear regression of the terminal phase of
Fig. 1. Percent of amoxicillin released from gelatin capsule formulation (open squares) and from the controlled release matrix tablet preparation (full squares).
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
33
was designed to release 50% of its contents within the first 3 h and to complete the drug release process over 8 h under in vitro conditions according to the USP specifications [13]. The effect of various parameters on the dissolution rate of amoxicillin from the new formulation was also assessed in this investigation. No statistically significant differences were found between the release rates of amoxicillin from the controlled release formulation tested under different pH conditions. Similarly, the extent of compression pressure was not found to affect amoxicillin release rate from the new formulation. On the other hand, there was a significant effect of the basket rotation speed on the release rate from the tablets. Increasing the rotation speed from 50 rpm to 100 rpm increased the erosion constant of the tablet 1.6 times, and 2.9 times higher at 200 rpm.
3.1. In vivo evaluation The mean amoxicillin concentration vs. time profile of active amoxicillin is presented in Fig. 2, and the pharmacokinetic parameters are summarized in Table 1. As can be seen, lower Cmax values (about 56% of the standard preparation) where found for the controlled release preparation and these Cmax value were attained much slower. On the other hand, although the AUC values tend to be lower for the AMOX-CR preparation, this difference is not statistically significant and the mean of the individual ratios of new / regular formulations AUC values for each volunteer is 0.92. Thus, there are no substantial differences in the extent of amoxicillin absorption between the regular formulation and the controlled release matrix tablets. Amoxicillin is degraded to an inactive metabolite penicilloic acid. The mean plasma penicilloic acid concentration versus time following administration of the two amoxicillin formulations is presented in Fig. 3. The mean AUC value of the controlled release preparation was found to be 23% lower than that of the standard preparation. The MIC values of representative amoxicillinsusceptible pathogens are outlined in Table 2, together with the more clinically relevant parameter of four times MIC. The time period in which amoxicillin concentrations were found to be over certain
Fig. 2. Serum active amoxicillin concentration following administration of 500 mg amoxicillin in gelatin capsule formulation (open squares) and controlled release matrix tablet preparation (full squares). The stripped zone present the difference between the time over a theoretical MIC value (0.2 mg / ml) produced by the two formulations.
‘therapeutic concentration’ levels required against every pathogen in each volunteer were measured and the mean values are summarized in Tables 2 and 3. It was found that the controlled release preparation extended the ‘time over MIC’. This is also illustrated in Fig. 4 in which no real difference between the onset time (i.e. the time following administration Table 1 Minimum Inhibitory concentration (MIC) values of amoxicillin against common pathogens susceptible to the drug Pathogen
MIC (mg / ml)
MIC34 (mg / ml)
Strep. group A Strep. pneumoniae Staph. aureus H. influenzae
0.01 0.02 0.1 0.25
0.04 0.08 0.4 1
34
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
Fig. 3. Serum penicilloic acid concentration following administration of 500 mg amoxicillin in gelatin capsule formulation (open squares) and controlled release matrix tablet preparation (full squares).
Table 2 Mean time (h) over certain MIC values produced by the standard and the sustained release preparations a Pathogen
Sustained release
Gelatin capsule
Strep. group A Strep. pneumoniae Staph. aureus H. influenzae
13.461.4 * 13.561.5 * 11.761.7 * 9.962.3 *
12.061.9 12.061.9 10.261.4 8.561.8
a
Results reported as mean6S.D. (n512). * Statistically significant difference between the sustained release formulation and the gelatin capsule formulation by Wilcoxon test (P,0.05).
Table 3 Mean time (h) over certain MIC34 values produced by the standard and the sustained release preparations a Pathogen
Sustained release
Gelatin capsule
Strep. group A Strep. pneumoniae Staph. aureus H. influenzae
13.061.9 * 12.662.1 * 9.262.0 8.562.8 *
11.062.0 10.761.9 8.061.5 5.661.4
a
Results reported as mean6S.D. (n512). * Statistically significant difference between the sustained release formulation and the gelatin capsule formulation by Wilcoxon test (P,0.05).
Fig. 4. Semi-logarithmic plot of serum active amoxicillin concentration following administration of 500 mg amoxicillin in gelatin capsule formulation (open squares) and controlled release matrix tablet preparation (full squares). The plot exhibits the differences in time over certain MIC values.
required to get the ‘therapeutic concentration’) of the two preparations is observable.
4. Discussion The development of an improved amoxicillin formulation must follow the pharmacodynamic rationale which, in the case of beta-lactam antibiotics, is to extend the ‘time over MIC’. Achieving this goal in vivo is complicated by the pharmacokinetic constraints which include rapid elimination of amoxicillin with a half-life of about 1 h, and a limited ‘absorption window’ following peroral
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
administration. As shown by Barr et al. [11], in humans the drug is well absorbed in the duodenum and jejunum, but absorption is decreased and ratedependent in the ileum, where more drug is absorbed following slow infusion versus bolus, and absorption is poor in all colon regions. Since the kinetics of drug elimination does not depend on the pharmaceutical delivery system, a rational method of prolonging the ‘time over MIC’ is to extend the absorption phase by slowly releasing the drug in the upper parts of the GI tract. The goal of the development stage, therefore, was to prepare a controlled release matrix tablet that will release, in in vitro dissolution tests, 50% of the drug within 3 h, followed by a constant release rate which will be completed after about 8 h [16]. The rapid onset of drug release was designed to provide an initial ‘loading dose’ and to maximize the absorption phase in those parts of the intestine in which amoxicillin is absorbed by a carrier-mediated process [12,17]. As found in the in vivo evaluation of the new formulation, the extent of absorption of the new formulation is not much different than that of a regular soft gelatin capsule formulation. Furthermore, the time required to obtain therapeutic concentration (‘onset time’) was found to be identical for the two formulations. The matrix tablet formulation developed is based on a hydrophilic (hydroxypropyl methyl-cellulose) polymer. As previously described [18], the kinetics of drug release from these matrix tablets follows the erosion mechanism [19,20]. Accordingly, variations in the releasing medium, such as pH changes, do not affect the kinetics of drug release. Since the erosion mechanism is controlled by hydration and gelation of the outer matrix layer, the reduced porosity due to increased compression pressure does not affect the rate of erosion, and the rate of drug release is therefore also unaffected. On the other hand, the rate of drug release in this delivery system by the erosion mechanism is expected to be affected by the rotation speed, as was found experimentally in this work. The pharmacodynamics of the bactericidal activity of amoxicillin is not concentration-dependent [6]. The concentration–antimicrobial activity relationship of beta- lactam antibiotics can be theoretically described according to the sigmoidal Emax model equation:
35
Emax C n E 5 E0 1 ]]] EC n50 1 C n The magnitude of effect E reaches the maximal effect Emax at relatively low concentrations. E0 is the magnitude of effect without the drug, which in this case is the antimicrobial effect of the immune system. The parameter n is the shape factor which is derived from the distribution function of the sensitivity of the microorganisms to the bactericidal effect of the drug [21]. In this case the value of n is expected to be high (i.e. close to 5) according to two approaches: (a) the concentration range between the minimal effective concentration (MIC) and the minimal concentration required to achieve Emax (CEmax ) is known to be low for these drugs; (b) the relationship between concentration and effect of each individual effector unit, which in this case is each microorganism, is an all-or-none pattern, and therefore, since the variability in degree of sensitivity of the microorganism population to the drug is expected to be low, n value has to be high in order to construct a steep concentration effect profile [21]. A theoretical pharmacodynamic profile of beta-lactam antibiotics drawn according to these parameters is illustrated in Fig. 5. It this case EC50 (i.e. the concentration at which 50% of the maximal effect is achieved) is a concentration slightly above MIC (or more accurately, above the minimal bactericidal concentration (MBC)). The importance in realizing
Fig. 5. Theoretical pharmacodynamic profile of beta-lactam antibiotics against susceptible pathogens. The scheme illustrate the relative narrow margin between drug concentration that do not contribute antimicrobial effect (less than MIC) and the concentration required to achieve maximal activity (Emax ).
36
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
the pharmacodynamic profile is that it indicates the optimal dosing regiment of the drug. Clarifying that high concentrations exceeding CEmax do not hold any additional therapeutic effect. It emphasizes the fact that the significantly greater Cmax values found in this investigation for the gelatin capsules in comparison to the new formulation are not associated with a larger magnitude of effect. In many cases the pharmacodynamic properties of the drug are not taken into consideration appropriately during the development stage while the pharmacokinetic parameters are over-emphasized. A relevant example is the work of Hilton and Deasy [22] who developed controlled release tablets of amoxicillin trihydrate with hydroxypropyl methylcellulose acetate succinate enteric polymer matrix and criticized the new product because of its lower AUC values (64.4%) in comparison to a proprietary product. From a pharmacodynamic point of view the comparison between the conventional formulation and the new formulation has to focus on the differences in the area under the required therapeutic concentration, as previously suggested by Schentag et al. [23]. According to this approach a line was drawn at a concentration of 0.2 mg / ml (Fig. 2) (which is ten times the MIC of Strep. pneumoniae and five times the MIC of Staph. aureus) and the difference between the two products regarding the area under this set concentration was highlighted by striped lines. This striped zone is the essence of the difference between the therapeutic efficacy of the conventional and AMOX-CR formulations. The ‘time over MIC’ depends on the specific pathogen that has to be treated. For those pathogens with relatively low sensitivity to amoxicillin, e.g. MIC.0.25 mg / ml, it would be better to select a more suitable antibiotic medication, because the time over MIC34 with either formulation will be too short, which will lead to regrowth of resistant strains and resultant therapeutic failure. On the other hand, as shown in Tables 2 and 3 and in Fig. 4 for amoxicillin-sensitive pathogens, the new controlled release formulation provided extended time over MIC as well as MIC34 levels. Extending the residence time of the antibiotic drug in the GI tract by the sustained release formulation may increase, in theory, the GI adverse effects associated with amoxicillin therapy. Such phenomenon of increased risk of side effect would be an
obvious limitation for the development of AMOXCR formulation. However, previous clinical investigations have demonstrated that long acting amoxicillin formulations used for therapy of upper respiratory tract infections [24] and urinary tract infection [25] did not increase side effects. In addition, the unabsorbed portion of the dose that had prolonged transit time in the GI tract due to the sustained release formulation is captured in the matrix formulation and is not available to interact with the intestine epithelia and / or flora. This notion is based on the lower degree of presystemic degradation of amoxicillin in the sustained release formulation in comparison to gelatin capsule formulation, shown in Fig. 3. A single matrix tablet of 500 mg amoxicillin was compared in this investigation with an equivalent dose of conventional capsule. In order to achieve a twice daily dosing regimen that will provide therapeutic concentrations for the whole 12 h dosing intervals a larger dose of the AMOX-CR should be given (750 mg or even 1 g twice daily). This recommendation is based on the large interindividual differences in the extent of amoxicillin absorption found in this investigation, and is intended to assure that the ‘poor’ absorbers will also benefit from full antibiotic efficacy. However, the validity of this recommendation should be validated experimentally. There are clear advantages to twice daily versus thrice daily dosing regimens. Such a regimen significantly increases patient compliance and thereby improves therapeutic outcome [26]. Like other dosing regimens which incorporate pharmacodynamic principles, it may lead to decreased cost of therapy, reduce the incidence of adverse drug reactions and improve or at least equal clinical efficacy.
Acknowledgements The work was supported by Vitamed, Binyamina, Israel Ltd. Dr. A. Hoffman and Prof. M. Friedman are affiliated with the David R. Bloom Center for Pharmacy.
References [1] N. White, T. Davis, Anti-ineffective drugs, in: C. van Boxtel,
A. Hoffman et al. / Journal of Controlled Release 54 (1998) 29 – 37
[2]
[3] [4] [5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13] [14]
N. Holford, M. Danhof (Eds.), The in vivo study of drug action, Elsevier, Amsterdam, 1992, pp. 311–334. W. Craig, S. Ebert, Continuous infusion of beta-lactam antibiotics. Antimicrob. Agent. Chemother. 36 (1992) 2577– 2583. W. Craig, S. Ebert, Killing and regrowth of bacteria in vitro: a review. Scand. J. Infect. Dis. 74(Suppl.) (1991) 63–70. S. Holm, I. OdenholtTornqvist, O. Cars, Paradoxical effects of antibiotics, Scand. J. Infect. Dis. 74 (1991) 113–117. G. Drusano, Role of pharmacokinetics in the outcome of infections, Antimicrob. Agent. Chemother. 32 (1988) 289– 297. B. Vogelman, S. Gudmundsson, J. Leggett, J. Turnidge, S. Ebert, W. Craig, Correlation of antimicrobial pharmacokinetic parameters with therapeutic efficacy in an animal model, J. Infec. Dis. 158 (1988) 831–847. R. Roosendaal, I. Bakker-Woundenberg, J. van der Berg, M. Michel, Therapeutic efficacy of continuous versus intermittent administration of ceftazidime in an experimental Klebsiella pneumonia pneumonia in rats, J. Infec. Dis. 152 (1985) 373–378. J. Zeisler, J. McCarthy, W. Richelieu, M. Nichol, Cefuroxime by continuous infusion: a new standard of care?, Infect. Med. 9 (1992) 54–60. W. Craig, S. Gudmundsson, The post-antibiotic effect, in: V. Lorian (Ed.), Antibiotics in laboratory medicine, Williams and Wilkins, Baltimore, 1986, pp. 515–536. J. Rotschafer, K. Walker, K. Madaras-Kelly, C. Sullivan, Antibiotic pharmacodynamics, in: N. Cutler, J. Sramek, P. Narang (Eds.), Pharmacodynamics and drug development: Perspectives in clinical pharmacology, John Wiley, West Sussex, 1994, pp. 315–343. W. Barr, M. Zola, E. Candler, H. Shie-Ming, A. Tendolkar, R. Shamburek, B. Parker, M. Hilty, Differential absorption of amoxicillin from the human small and large intestine, Clin. Pharmacol. Ther. 56 (1994) 279–285. F. Torres-Molina, J. Peris-Ribera, M. Garcia-Carboneli, J. Aristorena, J. Pla-Delfina, Nonlinearities in amoxycillin pharmacokinetics. II absorption studies in the rat, Biopharm. Drug Dispos. 13 (1992) 39–53. USP. XXI, 1985. K. Miyazaki, K. Ohtani, K. Sunada, T. Arita, Determination of ampicillin, amoxicillin, cephalexin and cephradine in plasma by high-performance liquid chromatography using fluorometric detection. J. Chromatog. 276 (1983) 478–482.
37
[15] M. Bialer, S. Sussan, O. Abu Salach, H.D. Danenberg, J. Ben David, Y. Gibor, A. Laor, Criteria to assess in vivo performance of sustained release products: application to diltiazem formulations, J. Pharm. Sci. 84 (1995) 1160–1163. [16] I. Katzhendler, A. Hoffman, M. Friedman, Controlled-release pharmaceutical preparations, Israeli patent application No. 119627 (1996). [17] P.J. Sinko, G.L. Amidon, Characterization of the oral absorption of beta- lactam antibiotics. II. Competitive absorption and peptide carrier specificity, J. Pharm. Sci. 78 (1989) 723–727. [18] Y. Katzhendler, A. Hoffman, A. Goldberg, M. Friedman, Modeling of drug release from erodible tablets, J. Pharm. Sci. 86 (1997) 110–115. [19] J. Heller, R. Baker, R. Gale, O. Rodin, Controlled drug release by polymer dissolution. I. Partial ester of maleic anhydride copolymers — properties and theory. J. Appl. Polym. Sci. 22 (1978) 1991–2009. [20] J. Heller, Controlled release of biologically active compounds from bioerodible polymers, Biomaterials 1 (1980) 51–57. [21] A. Hoffman, A. Goldberg, The relationship between receptor-effector unit heterogeneity and the shape of the concentration-effect profile: pharmacodynamic implications, J Pharmacokinet. Biopharm. 22 (1994) 449–468. [22] A.K. Hilton, P.B. Deasy, Use of hydroxypropyl methylcellulose acetate succinate in an enteric polymer matrix to design controlled-release tablets of amoxicillin trihydrate, J. Pharm. Sci. 82 (1993) 737–743. [23] J.J. Schentag, D.E. Nix, M.H. Adelman, Mathematical examination of dual individualization principles (I): Relationships between AUC above MIC and area under the inhibitory curve for cefmenoxime, ciprofloxacin and tobramycin, Drug Intel. Clin. Pharm. 25 (1991) 1050–1057. [24] N. Yoshihito, K. Masayoshi, Y. Naoyuki, S. Rinzo, Clinical studies of long acting amoxicillin granules. Japan J. Antibiotics (1983) 428–432. [25] M. Ktsuyuki, N. Akira, T. Akihiko, Clinical study on long acting amoxicillin in urinary tract infection. Japan J. Antibiotics (1983) 408–413. [26] J. Cockburn, A.L. Reid, J.A. Bowman, R.W. Sanson Fisher, Effects of intervention on antibiotic compliance in patients in general practice, Med. J. Aust 147 (1987) 324–328.