Vancomycin dosing in morbidly obese patients

Transcription

1 Eur J Clin Pharmacol (1998) 54: 621±625 Ó Springer-Verlag 1998 PHARMACOKINETICS AND DISPOSITION L. A. Bauer á D. J. Black á J. S. Lill Vancomycin dosing in morbidly obese patients Received: 9 March 1998 / Accepted in revised form: 30 June 1998 Abstract Objectives and methods: Vancomycin hydrochloride dosing requirements in morbidly obese patients with normal renal function were computed to determine the dose of vancomycin necessary to achieve target steady-state peak and trough concentrations and compared with a normal weight population. Results: Morbidly obese patients [total body weight (TBW) 165 kg, ideal body weight (IBW) 63 kg] required 31.2 mg á kg )1 á d )1 TBW or 81.9 mg á kg )1 á d )1 IBW to achieve the target concentrations. Normal weight patients (TBW 68.6 kg) required 27.8 mg á kg )1 á d )1 to achieve the same concentrations. Because of altered kinetic parameters in the morbidly obese patients (obese: t1/2 ˆ 3.3 h, V ˆ 52 L, ˆ 197 ml á min )1 ; normal: t1/ 2 ˆ 7.2 h, V ˆ 46 L, ˆ 77 ml á min )1, 20 of 24 patients required q8h dosing (1938 mg q8h) compared with q12h dosing (954 mg q12h) in all normal weight patients in order to avoid trough concentrations that were too low for prolonged periods. There was a good correlation between TBW and, but only fair correlation between TBW and V. Conclusion: Doses required to achieve desired vancomycin concentrations are similar in morbidly obese and normal weight patients when TBW is used as a dosing weight for the obese (approximately 30 mg á k- g )1 á d )1 ). Shorter dosage intervals may be needed when dosing morbidly obese patients so that steady-state trough concentrations remain above 5 lg á ml )1 in this population. Because of the large amount of variation in required doses, vancomycin serum concentrations should be obtained in morbidly obese patients to ensure L.A. Bauer (&) á D.J. Black á J.S. Lill Department of Pharmacy, School of Pharmacy, H375T Health Sciences Building, Box , University of Washington, Seattle, WA , USA labauer@u.washington.edu Tel.: , Fax: L.A. Bauer Department of Laboratory Medicine, School of Medicine, University of Washington, Seattle, WA, USA that adequate doses are being administered. Dosage requirements for morbidly obese patients with renal dysfunction require further study. Key words Vancomycin á Obesity á Pharmacokinetic variability Introduction Dosage guidelines for vancomycin hydrochloride in morbidly obese patients with normal renal function were derived in the early 1980s before de nitive steady-state peak and trough target concentrations were available [1]. Doses based on total body weight (23.4 mg á kg )1 á d )1 TBW) were suggested based on the goal of achieving average steady-state concentrations of 15 lg á ml )1 [2]. At that time it was not fully appreciated that vancomycin exhibited time-dependent bacterial killing and that maintaining adequate steady-state trough concentrations may be important for this antibiotic [3]. The purpose of this study was to reassess vancomycin dosing needs in this population using de ned steady-state vancomycin concentrations. Materials and methods Twenty-four morbidly obese patients receiving intravenous vancomycin hydrochloride (Vancocin, Lilly, Indianapolis, IN, USA), who had steady-state vancomycin concentrations ordered by the clinicians taking care of them, comprised the study group. The total body weight (TBW) of these patients was more than 90% over their ideal body weight (IBW), and each patient had normal renal function (serum creatinine ˆ 0.4±1.5 mg á dl )1 ; Table 1). Normal weight patients (within 15% of IBW) receiving intravenous vancomycin, who had steady-state vancomycin concentrations obtained during the course of their treatment, were matched to the morbidly obese patients by gender, age ( 5 years), IBW ( 5 kg), and serum creatinine concentration ( 0.3 mg á dl )1 ; Table 1). Ideal body weight was computed for males and females using the following equations: IBW male (in kg) ˆ (Height )60) and IBW female (in kg) ˆ (Height )60), where height is in

2 622 Table 1 Patient characteristics mean with (SD) Patient class and gender Age (years) Ideal body weight (kg) Total body weight (kg) Creatinine clearance a (ml á min )1 ) Serum creatinine (mg á dl )1 ) Morbidly obese (14M/10F) 41 (7) 63 (14) 165 (46)* 209 (35)* 0.6 (0.2) Normal weight (14M/10F) 40 (7) 64 (9) 68 (6) 110 (17) 0.7 (0.2) a Estimated using Salazar-Corcoran method for obese, Cockcroft-Gault for normal * P < compared with normal weight inches [4]. Estimated creatinine clearance was calculated using the Salazar-Corcoran [5] method for morbidly obese patients and the Cockcroft-Gault [6] method for normal weight patients. Patients in both weight categories were seriously ill individuals with known or suspected methicillin-resistant Staphylococcus aureus or Staphylococcus epidermidis infections. Steady-state vancomycin peak and trough concentrations were measured in each patient after dosing had commenced for at least 5 estimated half-lives [7]. Blood samples were usually taken after 48± 72 hours of therapy. Trough concentrations were obtained within 30 min of the start of the next infused dose; peak concentrations were obtained 1 h after the end of a 1 h infusion from that same dose. Medication administration sheets were checked to ensure that previous doses had been given on schedule. Serum samples were assayed for vancomycin concentrations in the clinical laboratory (interday and intraday CV <10%) using a commercially available uorescence polarization immunoassay (TDx; Abbott Laboratories, Irving, TX). Vancomycin pharmacokinetic constants were computed using a one-compartment intermittent IV infusion model [7, 8] at steadystate. Doses (rounded to the nearest 250 mg) and dosage intervals (rounded to clinically acceptable times of every 6, 8, 12, etc. hours) required to achieve target steady-state peak concentrations of 25± 35 lg á ml )1 and trough concentrations (5±10 lg á ml )1 ) were computed for each patient group. Statistical comparisons were conducted using the paired student's t-test and linear regression. To con rm the accuracy of the pharmacokinetic parameters computed in the patients, follow-up concentrations were compared with predicted concentrations using the kinetic constants in the onecompartment model. This study was approved by the university's institutional review board. Results Vancomycin clearance was 2.5 times greater in the morbidly obese patients (mean value 197 ml á min )1 ) compared with matched normal weight subjects (77 ml á min )1, Table 2). Clearance normalized for weight using IBW was greater for the morbidly obese group (3.1 ml á min )1 á kg )1 IBW) than for the normal weight patients (1.0 ml á min )1 á kg )1 IBW). However, when TBW was used to normalize clearance values there was no di erence between patient groups (1.2 ml á min )1 á kg )1 TBW for morbidly obese patients vs 1.1 ml á min )1 á kg )1 TBW for normal weight patients). There was a good correlation between TBW and clearance in these patients with normal renal function (r ˆ 0.948, P < ; Fig. 1). A poor correlation existed between IBW and vancomycin clearance (r ˆ 0.204). Volume of distribution was statistically similar in the morbidly obese patients (mean value 52 l) compared with matched normal weight subjects (46 l), but there was a modest correlation between TBW and volume of distribution (r ˆ 0.490, P < 0.05; Fig. 2). No correlation existed between IBW and vancomycin volume of distribution (r ˆ 0.170). Normalized for weight, volume of distribution was larger for the normal weight group (0.68 l á kg )1 TBW) when TBW was used, compared with the value found in the morbidly obese group (0.32 l á kg )1 TBW). When IBW was used to normalize volume of distribution values, the morbidly obese patients had larger volumes of distribution (0.83 l á kg )1 IBW) compared with normal weight patients (0.64 l á kg )1 IBW). Due to the relatively large change in clearance found in the morbidly obese patients without a commensurate change in volume of distribution, half-life was shorter in the morbidly obese group (3.3 h) compared to the weight patients (7.2 h). Because of the pharmacokinetic changes, morbidly obese patients required larger doses and shorter dosage intervals compared with normal weight patients to achieve the target steady-state vancomycin concentrations. Twenty of 24 morbidly obese patients required 8 hourly dosing to avoid trough concentrations that were too low [1938 (549) mg every 8 h]. In the remaining four morbidly obese patients 12 hourly dosing was adequate [1250 (289) mg every 12 h]. These doses produced mean steady-state peak and trough concentrations equal to 29.5 (2.3) lg á ml )1 and 7.2. (1,4) lg á ml )1, respectively. In contrast, normal weight patients required 954 (309) mg every 12 h to achieve mean peak and trough Table 2 Pharmacokinetic parameters mean with (SD). IBW ideal body weight, TBW total body weight Patient class (ml á min )1 ) (ml á min )1 á kg )1 IBW) (ml á min )1 á kg )1 TBW) V(l) V (l á kg )1 IBW) V (l á kg )1 TBW) Half-life (h) Morbidly obese 197 (77)* 3.1 (0.7)* 1.2 (0.2) 52 (13) 0.83 (0.08)* 0.32 (0.05)* 3.3 (0.8)* Normal weight 77 (22) 1.0 (0.4) 1.1 (0.3) 46 (16) 0.64 (0.26) 0.68 (0.24) 7.2 (2.2) * P < compared with normal weight

3 623 Fig. 1 Relationship between total body weight and vancomycin clearance for morbidly obese and normal weight patients with normal renal function (r ˆ 0.948, P < ) Fig. 2 Relationship between total body weight and vancomycin volume of distribution for morbidly obese and normal weight patients with normal renal function (r ˆ 0.490, P < 0.05) steady-state concentrations of 25.8 (5.2) lg á ml )1 and 9.0 (1.7) lg á ml )1, respectively. When total daily doses were normalized to weight, morbidly obese patients required 31.2 (6.3) mg á kg )1 á d )1 TBW or 81.9 (19.9) mg á kg )1 á d )1 IBW, while the normal weight group required 27.8 (8.6) mg á kg )1 á d )1 TBW or 29.5 (9.2) mg á kg )1 á d )1 IBW. Follow-up vancomycin concentrations were available in 22 obese and 20 normal weight patients; a strong correlation was found between model-predicted and actual follow-up concentrations (r ˆ 0.975, P<0.0001; Cpredicted ˆ 0.95 * Cactual ). Discussion Morbid obesity is known to alter the pharmacokinetics of several drugs [9], including the aminoglycoside antibiotics [10±13] and vancomycin [1, 8, 14]. The initial dosing guidelines for vancomycin in this patient group were made before vancomycin was identi ed as exhibiting time-dependent bacterial killing and target peak and trough steady-state concentrations were suggested [1]. Thus, these recommendations were made at a time when vancomycin was routinely given every 6 h and desired average steady-state concentrations were 15 lg á ml )1 [2]. This reappraisal of vancomycin dosing requirements in morbidly obese patients is necessary now that di erent target steady-state concentrations have been identi ed and the importance of maintaining minimum e ective vancomycin trough concentrations to maintain bacterial killing is recognized. Although routine monitoring of vancomycin concentrations in patients has been questioned [15], recent research studies have shown that concentration monitoring is useful in critically ill patients [16, 17]. Some of the physiologic changes that alter drug pharmacokinetics are evident in morbidly obese patients. These patients have large amounts of excess adipose tissue, but body organs also increase in size and weight so that lean body mass is greater and blood volume is larger [18, 19]. Also, even though adipose cells contain predominately fat, adipose tissue contains extracellular uid to which drugs may distribute. Because

4 624 of this, the volume of distribution for many, but not all, drugs is larger in morbidly obese patients. This is because the volume of distribution of a drug is a function of the physiologic volume of blood and organs as well as the drug binding in the blood and organs. Therefore, if a drug does not penetrate into adipose tissue to a signi cant extent, the volume of distribution for a drug may be the same in morbidly obese and normal weight patients as is the case for digoxin [20, 21] and cimetidine [22]. However, for drugs that are very polar and water-soluble the volume of distribution may increase in morbidly obese patients, because the drug distributes to the additional extracellular uid contained in large stores of adipose tissue, as is the situation for the aminoglycoside antibiotics [10±13]. Finally, if the drug is very lipid soluble, like diazepam, the volume of distribution can increase to a large extent in morbidly obese patients [23]. Other physiologic alterations that change drug pharmacokinetics in morbidly obese patients are more subtle. Creatinine clearance, a surrogate marker for glomerular ltration rate, is greater in morbidly obese patients compared with normal weight individuals given equal serum creatinine concentrations [24]. This change is presumably due to larger kidneys in morbidly obese patients yielding more functional nephrons [19]. Because of this, the clearance of renally eliminated drugs, like aminoglycosides and vancomycin, is accelerated in morbidly obese patients [1, 8, 10±14]. Also, a specialized method of estimating creatinine clearance in morbidly obese patients has been devised since standard methods such as the Cockcroft-Gault equation are not accurate in this patient population [5, 24]. Two additional studies have examined the e ect of obesity on vancomycin pharmacokinetics using multiple linear regression techniques. The rst investigation studied 230 adult patients with a variety of estimated creatinine clearance values (computed using the method of Cockcroft and Gault) and body weights [14]. Patients were strati ed into nine groups according to weight; the heaviest group was 60% or more above their IBW. The mean age of this group was 60 years, and it was comprised of four males and ten females. Renal function was not reported for this group of obese patients; however, the reported vancomycin clearance of 0.6 ml á min )1 á kg )1 TBW suggests that their renal function was below normal. The mean volume of distribution and half-life was 128 l and 22.8 h, respectively. Based on the results of their regression analysis, the authors concluded that vancomycin doses should be based on TBW. The other paper concerned a population of 704 patients with a large range of ages and renal function [8]. Patients with serum creatinine concentrations <0.7 mg á dl )1 were excluded from analysis and the Cockcroft-Gault equation was used to estimate creatinine clearance. The patients were strati ed into three weight groups, with obesity de ned as TBW/IBW >1.3. One hundred eight patients were in the obese group (mean TBW 94 kg, mean IBW 60 kg), with a mean age of 50 years and mean estimated creatinine clearance of 60 ml á min )1. The mean clearance, volume of distribution, and half-life for these patients were 74 ml á min )1, 52 l, and 7 h. The authors recommend that vancomycin doses be adjusted for obesity in addition to age and gender using their regression equations. The ndings in our patients suggest that vancomycin should be dosed based on TBW in morbidly obese patients with normal renal function. In both seriously ill normal weight and morbidly obese patients with normal renal function, total daily doses of 30 mg á kg )1 á d )1 TBW were required to produce target steady-state peak concentrations of 25±35 lg á ml )1 and trough concentrations of 5±10 lg á ml )1. The principal pharmacokinetic change that occurred in the morbidly obese patients was an increased vancomycin clearance of 197 ml á min )1 compared with 77 ml á min )1 in normal weight patients. However, when TBW was used to normalize clearance, the values were virtually identical (1.2 ml á min )1 á kg )1 TBW in morbidly obese, 1.1 ml á min )1 á kg )1 TBW in normal weight). Additionally, there was a good correlation between TBW and clearance (r ˆ 0.948, P < ; Fig. 1). The reason for the increased clearance rate (in ml á min )1 ) found in morbidly obese patients is an increased glomerular ltration rate as evidenced by the large estimated creatinine clearance computed in these patients and large measured creatinine clearance values in other morbidly obese individuals [24]. While the volume of distribution was larger in the morbidly obese patients (mean 52 l) than in the normal weight patients (mean 46 l), the values were not signi cantly di erent. This is in contrast to ndings in other studies with morbidly obese individuals [12] and obese patients [8], where similar mean values for volume of distribution produced statistically signi cant di erences when compared with people of normal weight. It is possible, given the large amount of variability for volume of distribution in our patients (Table 2, Fig. 2), that we did not have su cient power to nd a signi cant di erence for this parameter. In any case, the mean di erence between volumes of distribution for the two populations were relatively modest (13%) compared with the mean di erence in clearances (156%). Because clearance changed to such a large extent compared with volume of distribution, half-life was much less in the morbidly obese patients (mean 3.3 h) compared with the value found in normal weight patients (mean 7.2 h). This result produced the other major nding of this study, which is that dosage intervals may need to be shorter in morbidly obese patients with normal renal function. Twenty of 24 patients required an 8-h dosage interval in order to avoid trough concentrations that were below 5 lg á ml )1. The nding is especially important when one considers that vancomycin exhibits time-dependent bacterial killing, and that long periods of time with sub-therapeutic trough concentrations may allow bacterial regrowth to occur. In conclusion, the results of our study support the use of 30 mg á kg )1 á d )1 TBW as a starting dosage in

5 625 morbidly obese patients with normal renal function. In order to avoid low trough concentrations for prolonged periods, dosing every 8 h may be needed in this patient population. Because of pharmacokinetic variability, vancomycin serum concentrations should be measured in this group of patients to keep trough concentrations within the desired therapeutic range and to identify those patients for whom dosing every 12 h would be appropriate. Dosage requirements for morbidly obese patients with renal dysfunction require further study. References 1. Blouin RA, Bauer LA, Miller DD, Record KE, Gri en Jr WO (1982) Vancomycin pharmacokinetics in normal and morbidly obese subjects. Antimicrob Agents Chemother 21: 575± Moellering Jr RC, Krogstad DJ, Greenblatt DJ (1981) Vancomycin therapy in patients with impaired renal function: a nomogram for dosage. Ann Intern Med 94: 343± James JK, Palmer SM, Levine DP, Rybak MJ (1996) Comparison of conventional dosing versus continuous-infusion vancomycin therapy for patients with suspected or documented gram-positive infections. Antimicrob Agents Chemother 40: 696± Devine BJ (1974) Gentamicin therapy. Drug Intell Clin Pharm 8: 650± Salazar DE, Corcoran GB (1988) Predicting creatinine clearance and renal drug clearance in obese patients from estimated fat-free body mass. Am J Med 84: 1053± Cockroft DW, Gault MH (1976) Prediction of creatinine clearance from serum creatinine. Nephron 16: 31±41 7. Matzke GR, McGory RW, Halstenson CE, Keane WF (1984) Pharmacokinetics of vancomycin in patients with various degrees of renal function. Antimicrob Agents Chemother 25: 433± Ducharme MP, Slaughter RL, Edwards DJ (1994) Vancomycin pharmacokinetics in a patient population: e ect of age, gender, and body weight. Ther Drug Monit 16: 513± Abernethy DR, Greenblatt DJ (1986) Drug disposition in obese humans ± an update. Clin Pharmacokinet 11: 199± Bauer LA, Blouin RA, Gri en Jr WO, Record KE, Bell RM (1980) Amikacin pharmacokinetics in morbidly obese patients. Am J Hosp Pharm 37: 519± Bauer LA, Edwards WAD, Dellinger EP, Simonowitz DA (1983) In uence of weight on aminoglycoside pharmacokinetics in normal weight and morbidly obese patients. Eur J Clin Pharmacol 24: 643± Blouin RA, Mann HJ, Gri en, Jr. WO, Bauer LA, Record KE (1979) Tobramycin pharmacokinetics in morbidly obese patients. Clin Pharmacol Ther 26: 508± Schwartz SN, Pazin GJ, Lyon JA, Ho M, Pasculle AW (1978) A controlled investigation of the pharmacokinetics of gentamicin and tobramycin in obese subjects. J Infect Dis 138: 499± Vance-Bryan K, Guay DRP, Gilliland SS, Rodvold KA, Rotschafer JC (1993) E ect of obesity on vancomycin pharmacokinetic parameters as determined by using a Bayesian forecasting technique. Antimicrob Agents Chemother 37: 436± Cantu TG, Yamanaka-Yuen NA, Leitman PS (1994) Serum vancomycin concentrations ± reappraisal of their clinical value. Clin Infect Dis 18: 533± Welty TE, Copa AK (1994) Impact of vancomycin therapeutic drug monitoring on patient care. Ann Pharmacother 28: 1335± Zimmermann AE, Katona BG, Plaisance KI (1995) Association of vancomycin serum concentrations with outcomes in patients with gram-positive bacteremia. Pharmacotherapy 15: 85± Alexander JK, Dennis EW, Smith WG, Amad KH, Duncan WC, Austin RC (1962±1963) Blood volume, cardiac output, and distribution of systemic blood ow in extreme obesity. Cardiovasc Res Cent Bull (Houston) 1: 39± Naeye RL, Roode P (1970) The size and numbers of cells in visceral organs in human obesity. Am J Clin Pathol 54: 251± Abernethy DR, Greenblatt DJ, Smith TW (1981) Digoxin disposition in obesity: clinical pharmacokinetic investigation. Am Heart J 102: 740± Ewy GA, Groves BW, Ball MF, Nimmo L, Jackson B (1971) Digoxin metabolism in obesity. Circulation 44: 810± Bauer LA, Wareing-Tran C, Edwards WAD, Raisys V, Ferrer L, Jack R, Dellinger EP, Simonowitz D (1985) Cimetidine clearance in the obese. Clin Pharmacol Ther 37: 425± Abernethy DR, Greenblatt DJ, Divoll M, Harmatz JS, Shader RI (1981) Alterations in drug distribution and clearance due to obesity. J Pharmacol Exp Ther 217: 681± Dionne RE, Bauer LA, Gibson GA, Gri en, Jr. WO, Blouin RA (1981) Estimating creatinine clearance in morbidly obese patients. Am J Hosp Pharm 38: 841±844