Secondary Logo

Journal Logo

Influence of Morbid Obesity on the Clinical Pharmacokinetics of Various Anti-Infective Drugs

Reappraisal Using Recent Case Studies—Issues, Dosing Implications, and Considerations

Srinivas, Nuggehally R. PhD, FCP*

doi: 10.1097/MJT.0000000000000401
Systematic Review and Clinical Guidelines
Free

Owing to availability of scanty pharmacokinetic data, dosing decisions in morbid obesity is increasingly challenging in the field of anti-infective drugs. However, in recent years data are emerging that describe the pharmacokinetics of anti-infective drugs in morbidly obese subjects. The objectives of the present work were: (1) to collate the recent reports pertaining to the pharmacokinetics in morbidly obese subjects for several anti-infective drugs and provide an overview of the pharmacokinetic data along with the applicable pharmacodynamics and/or clinical outcome; (2) to perform regression analysis on limited dataset for a few drugs to verify the existence of relationships between Cmax/Ctrough versus steady-state volume of distribution (Vss)/clearance to enable data prediction in morbid obesity subjects; (3) to provide a general discussion on issues and dosing implications. The key findings of this review were: (a) drugs such as vancomycin, ethambutol, and fluconazole, where the VSS is substantially greater in morbidly obese patients, need a dosing strategy with the appropriate body mass descriptors; (b) other drugs such as moxifloxacin, linezolid, doripenem, meropenem, voriconazole, oseltamivir, tigecycline, levofloxacin may not ordinarily need dosing adjustments; (c) using regression analysis of limited data, it seemed that drugs such as vancomycin, voriconazole, oseltamivir, and oseltamivir carboxylate may be amenable for the prediction of VSS and/or clearance in morbid obesity patients.

Suramus Bio, Drug Development, Bangalore, Karnataka, India.

Address for correspondence: Nuggehally R. Srinivas, PhD, FCP, Suramus Bio, Drug Development, 29th main, 10th cross, JP Nagar I Phase, Bangalore 560078, India. E-mail: srini.suramus@yahoo.com

The authors have no conflicts of interest to declare.

Back to Top | Article Outline

INTRODUCTION

The epidemic of obesity has become a major health issue in the developing world since it is associated with comorbidity of other diseases such as hypertension, cardiovascular disease, diabetes etc.1 The reports of Flegal et al2,3 has suggested that in a period spanning 4 decades, there was an almost 3-fold increase in the prevalence of extreme obesity [body mass index (BMI) ≥ 30 kg/m2] in the United States alone. Therefore, the onus of dosing patients with extreme obesity is still based on the body weight (BW) considerations and applicable renal function data for certain category of antimicrobials. Owing to the enormous change in the BW distribution, especially the fat component in such extremely obese patients, the question always remains whether or not dosing based on BW of the individual or the recommended therapeutic dose per dosing label without dosage adjustments would provide the required threshold levels of the drug for an expected pharmacodynamics and/or efficacy outcome. Furthermore, obesity and morbid obesity can directly alter physiological process such as cardiac output, hepatic blood flow, splanchnic blood flow which may perhaps lead to altered drug pharmacokinetics and disposition.4 Despite a large growing population of obese individuals, still the pharmacokinetic data are scanty when it comes to providing appropriate dosing recommendation for obese patients across many therapeutic areas including anti-infectives. In a recent review on the impact of obesity on cardiovascular medicines used for the management of cardiovascular risk factors, it was stated that one needs to first look at the clinical outcome as the critical step between obese and nonobese individuals. If clinical outcome is different, then one needs to look at secondary steps such as pharmacokinetics, pharmacodynamics to explain the disparity in the clinical outcome observed between obese and nonobese individuals.5 However, the same consideration proposed for cardiovascular medicines may not be practical in the case of anti-infective class, where a clinical outcome may be dependent on the adequacy of the parent drug and/or active metabolite distribution at the site of action and overall pharmacokinetic disposition of the drug in consideration.

In the last few years, however, clinical investigations into the pharmacokinetic and pharmacodynamics aspects have been performed with the key objectives of addressing and understanding the impact of morbid obesity on drug disposition parameters for several important drugs belonging to the anti-infective area.

Back to Top | Article Outline

SCOPE

This review covers several anti-infective drugs where pharmacokinetic data have been published in the patients with morbid obesity subjects. The individual examples reviewed in this work include drugs such as vancomycin, linezolid, voriconazole, ethambutol, meropenem, tigecycline, levofloxacin, doripenem, moxifloxacin, fluconazole, oseltamivir carboxylate, and oseltamivir. The study design, objectives, and overall pharmacokinetic data gathered in the morbidly obese or very obese subjects for the above stated drugs are tabulated along with pertinent remarks (Table 1). In addition, the comparative illustrations between morbidly obese versus nonobese subjects for the key pharmacokinetic parameters such as steady-state volume of distribution (VSS) and total body clearance (CL) for the 12 drugs reviewed in the present work are provided in Figures 1, 2, respectively. The general issues, perspectives, and considerations including dosing implications as a result of morbid obesity are discussed.

Table 1

Table 1

Table 1

Table 1

Table 1

Table 1

Table 1

Table 1

Table 1

Table 1

Table 1

Table 1

Table 1

Table 1

Table 1

Table 1

Table 1

Table 1

FIGURE 1

FIGURE 1

FIGURE 2

FIGURE 2

Back to Top | Article Outline

INDIVIDUAL CASE STUDIES

Vancomycin

Adane et al6 performed a prospective pharmacokinetic study in extremely obese patients (BMI ≥40 kg/m2) who were administered vancomycin for Staphylococcus aureus infections. The vancomycin dosing had to be for at least 3 consecutive days to be included in the investigation. The method involved the determination of population pharmacokinetics for vancomycin using a single compartment intravenous infusion model which was fit to 3 distinct serum concentrations (peak, mid-time point, and a trough concentration) of vancomycin by applying nonlinear mixed-effects modeling.6 It was found that BW and serum creatinine CL influenced the pharmacokinetic parameter such as volume of distribution (VD) and CL of vancomycin. By the modeling of all the serum concentration data, the population parameters for VD and CL for vancomycin were 0.51 L/kg, and 6.54 L/h, respectively. The simulations with a probability of ≥93% predicted that daily doses of vancomycin between 4000 to 5000 mg would be necessary in this population to provide a 24-hour area under the curve (AUC)/minimum inhibitory concentration (MIC) ratio of ≥400 at a MIC level of 1 μg/mL.6 The data from this population modeling and simulation analysis suggested that extreme obese patients would need a higher dose of vancomycin and the determination of the initial dose of vancomycin should be made by both total body weight (TBW) and renal function considerations. Although, adjusted body weight (ABW) was also tried for correlations with pharmacokinetic parameters, TBW seemed to be better correlated. The need for therapeutic drug monitoring of vancomycin serum levels was also recommended for dose titration during the therapy.6 Some 30 years ago, the work of Blouin et al7 had recommended higher dosing of vancomycin in morbidly obese patients relative to normal BW subjects. In addition, it was also proposed to consider for an increased dosing frequency of lowered vancomycin doses in morbidly obese patients to blunt the peak concentration from eliciting any untoward safety issues. More recently, in an interesting pediatric study with oncology patients, Mahmoud et al8 proposed consideration of obesity as a critical factor in the dosing considerations to ensure adequacy of vancomycin threshold levels. As suggested by Figures 1, 2,7 morbidly obese subjects exhibited substantially higher VSS and CL as compared with nonobese subjects.

Back to Top | Article Outline

Linezolid

Bhalodi et al (2012) have delineated the pharmacokinetics of linezolid in a patient population ranging (n = 10) from moderately obese to morbidly obese. The pharmacokinetic data between moderately obese and morbidly obese patients, either model independent or model dependent, were found to be similar. Although AUCtau value was numerically higher for the moderately obese group (130.3 mg·h·L−1), the removal of an outlier who showed a high value of 294.2 mg·h·L−1 brought the mean AUCtau value to 112.1 mg·h·L−1 almost near the morbidly obese patients (109.2 mg·h·L−1). In another study that involved similar linezolid dosing regimen in 95 obese patients, AUCtau value of 105.2 mg·h·L−1 was obtained showing the consistent behavior of linezolid across obese population.10 In contrast to obese patients, healthy volunteers who received linezolid (625 mg, every 12 hours interval) showed a slightly lower AUCtau of linezolid (93.43 mg·h·L−1).11

Although previous studies of linezolid had suggested that obese to extremely obese patients may have lower exposure to the drug because of higher systemic CL,12–14 it was not evident from the data reported by Bhalodi et al (2013). Also, it was postulated that lower exposure observed in previous studies may be an artifact of sample collection times, sparse sampling strategy, and/or application of the 1-comparmental model, which may not be appropriate to fit the plasma concentrations of linezolid. Also, another confounding factor that may have influenced the disposition of linezolid was related to the ill-health of the patients in such studies.12–14

Furthermore, Bhalodi et al (2013) established the correlation of the total VD with TBW (R2 = 0.524), ideal body weight (IBW) (R2 = 0.398), lean body weight (LBW), (R2 = 0.495) and ABW (R2 = 0.587); however, BMI (R2 = 0.171) failed to correlate with VD. Both systemic CL and VD of central compartment (Vc) were also poorly correlated with the various body mass descriptors. Based on the work of Bhalodi et al (2013), it was evident that morbidly obese patients should have adequate exposure from a pharmacodynamic perspective delivered by intravenous doses of 600 mg linezolid given every 12 hours. In extremely obese patients (>150 kg), one could use the various body mass descriptors for dose optimization if deemed necessary.9 As suggested by Figures 1, 2,9,11 both VSS and CL were not affected by morbid obesity as compared with nonobese individuals.9,11

Back to Top | Article Outline

Ethambutol

The work reported by Hall et al15 was geared toward developing a population pharmacokinetic model of ethambutol using serum concentrations obtained from subjects representing nonobese, obese, and morbidly obese categories. The key parameters developed from the model such as total body CL of ethambutol and VD of ethambutol (central compartment and peripheral compartment) were analyzed with host of covariates that included age, sex, creatinine CL, BMI, and BW. With the exception of BW, none of the other covariates had any influence on the pharmacokinetics of ethambutol. There was a strong correlation established between the BW versus total CL of ethambutol and it indicated increased CL of ethambutol in relation to an increase in BW in the study population included in the study. The relationship15 was defined by the equation: ethambutol CL = 42.6 × (BW/45.6)3/4. In this investigation, it was surprising that creatinine CL had no influence on the total CL of ethambutol because almost 80% of the administered dose of ethambutol is preferentially excreted by the renal pathway,16 whereas the remainder of the administered ethambutol dose undergoes metabolism by alcohol dehydrogenase (ADH) enzyme.16 It was postulated that ADH activity may increase as a function of BW to possibly explain the increased ethambutol CL observed in the study. Although there was a suggestion of increased ADH expression in obese individuals that had nonalcoholic hepatic steatosis as compared with the control group,17 further studies are required to confirm this finding to unequivocally provide a mechanism for increased CL of ethambutol in obese patients. The data from this study, when put in context to a previously reported work,18 suggested that obese and morbidly obese patients may be potentially at risk of efficacy failure because the desired AUC0–24/MIC levels may not be achieved in such patients because of increased CL of ethambutol as a function of BW. Although it may be possible to increase the dose ethambutol in both obese and morbidly obese patients based on the findings of this study,15 there are certain constraints: (1) the maximum recommended daily dose of first-line antituberculosis (TB) drugs has been capped at 1.6 g by World Health Organization (WHO) (World Health Organization guidelines for TB treatment)19; (2) high doses of ethambutol have been shown to be associated with an increased risk of optic neurophathy.20 Despite study limitations such as sample size and single dose data, the work of Hall et al (2012) provided clarity on the influence of BW on the CL and VD of ethambutol. The key recommendation to factor BW as a key variable in optimizing the dose size of ethambutol for the treatment of TB needs to be explored. As suggested in Figures 1, 2,15,21,22 the CL values observed in extremely obese individuals was almost comparable with that observed in nonobese subjects; but VSS was about 2–3-fold greater in obese subjects. Hence, it may be prudent to consider BW descriptors in the dosing decision of ethambutol in morbidly obese subjects.

Back to Top | Article Outline

Meropenem

Cheatham et al23 have assessed the pharmacokinetics of meropenem in morbidly obese patients (n = 9) that were hospitalized for medical intervention. The steady-state data derived from this study were used in various simulations of dosing regimens to determine the likely impact on the pharmacodynamics aspects of meropenem in patients with morbid obesity. The various volumes of distribution parameters (central, peripheral, and steady state) for meropenem were correlated with variety of body mass parameters such as TBW, IBW, LBW, and BMI.23 Interestingly, the systemic CL of meropenem at steady state was highly correlated with creatinine CL in morbidly obese patients (R2 = 0.489); whereas the peripheral VD of meropenem (R2 = 0.436) but not the volume of central compartment or volume at steady state was correlated with LBW in this patient population. The Cmax achieved in this study (28.8 mg/L) was similar to Cmax achieved in nonmorbidly obese patients (29.2 mg/L) that used dosing regimen of 500 mg every 6 hours by a 30-minute infusion; whereas the trough concentrations of meropenem achieved in morbidly obese patients was about 1.7-fold higher relative to nonmorbidly obese patients.23,24 The disposition of meropenem in morbidly obese patients seemed to be comparable with nonmorbidly obese patients who were hospitalized (half-life value of 3.1 hours vs. 2.1–2.5 hours).23,25,26 The uncorrected systemic CL of meropenem was comparable between morbidly obese versus nonmorbidly obese patients (10.5 vs. 9.3–11.5 L/h); however, after normalization of BW, the CL of meropenem was at least 1.5–2.2-fold slower relative to the CL in nonmorbidly obese patients.23,25,26 Similarly, morbidly obese patients had a somewhat higher steady-state volume of meropenem (1.29–1.74-fold higher relative to nonmorbidly obese patients), which when normalized by TBW made it smaller (0.25 L/kg vs. 0.28–0.38 L/kg) as compared with the nonmorbidly obese patients.23,25,26

From a pharmacodynamics perspective, various dosing regimens simulated by Cheatham et al (2014) achieved the required MIC target of 1 μg/mL for enterobacteriaceae with probability of target attainment (PTA) >90%. It was recommended for resistant pathogens which may require MIC of 4 μg/mL, the dosing regimen needs to be doubled (1 g every 6 hours or 2 g every 8 hours) to obtain a PTA >90%. However, the most important observation made from this study was that for a hypothetical MIC requirement of 8 μg/mL, none of the regimens achieved >90% PTA. In such situations, to achieve a higher target MIC of 8 μg/mL at >90%PTA, the available options were much higher dosing of meropenem (2 g every 6 hours) or continuous infusion for 3 hours (1 g every 6 hours).23 Using the data, it was established that owing to similarities in the pharmacokinetic profile of meropenem between nonobese and morbidly obese subjects, there was no need for dose titration studies in morbidly obese patients because the standard dosing schedules of meropenem were able to provide adequate pharmacodynamic endpoint for sensitive bacterial pathogens. However, if the infection is suspected to be due to resistant or less susceptible strains of pathogens, the option of either continuous infusion or higher dose of meropenem could be explored. As enumerated in Figures 1, 2,23,27,28 although there appeared to be a modest 1.2–1.5-fold difference in the CL and VSS parameters between obese versus nonobese subjects, it had no consequences in the dosing consideration for meropenem.

Back to Top | Article Outline

Tigecycline

Pai29 has reported the pharmacokinetics of tigecycline in a study that involved both morbidly obese patients (n = 8) and normal weight subjects (n = 4). In this single-dose intravenous study, both plasma pharmacokinetics and urinary excretion profiles of tigecycline were delineated. The pharmacokinetic disposition of tigecycline was comparable between the 2 cohorts, despite a 2.2-fold difference in the TBW among the patients. Similarly, the urinary excretion profiles matched between morbidly obese versus normal weight subjects. The population pharmacokinetic data derived from all subjects in this study were closely matching with literature values reported for tigecycline. The average peak concentration of 1.6 mg/L obtained in this study was in line with the reported value of 1.45 mg/L in the product label. The other parameters derived for tigecycline in this study using population pharmacokinetic approach concurred with the values reported for tigecyline.29–31 For example, the AUC0–96 (6.78 mg·h·L−1), systemic CL (14.2 L/h), and renal CL (3.03 L/h) values for tigecycline obtained in this study30 were in agreement with corresponding values of 6.40 mg·h·L−1, 14.6 L/h, and 2.6 L/h, respectively, reported previously.30 Interestingly, extrapolations and simulations performed for adequate pharmacodynamic exposure for relevant pathogens with predetermined MIC values and PTA >95%, suggested that morbidly obese patients may be at the risk of therapy failure for certain gram negative pathogens; however, it could not be rationalized solely based on the exposure of tigecycline which was not compromised in morbidly obese patients. Hence, the work of Pai29 suggested that despite similar pharmacokinetic profile, morbidly obese patients may run the risk of not adequate clinical response and therefore necessitated additional pharmacokinetic and pharmacodynamic studies in morbidly obese patients to provide clarity on these aspects. Furthermore, as displayed in Figures 1, 2,29,30 there seemed to be modest differences observed between VSS and CL parameters between morbidly obese and nonobese subjects.

Back to Top | Article Outline

Levofloxacin

Cook et al32 have evaluated the pharmacokinetics of levofloxacin in morbidly obese patients (n = 9) relative to ambulatory patients (n = 3). Similarities in the pharmacokinetics were observed for parameters such as peak concentration and VD for levofloxacin between morbidly obese and ambulatory patients; such data were comparable with other patient pool including healthy volunteers33–35 However, morbidly obese patients seemed to have a slower CL of levofloxacin leading to a considerably higher exposure of the drug and shorter half-life values when compared with the ambulatory patients (2.6 vs. 6.9 hours in morbidly obese and ambulatory, respectively). As suggested by Figures 1, 2,32,33,35 the VSS and CL for levofloxacin seemed to be at the lower end of the values observed for nonobese subjects suggesting that BW consideration may need to be factored in dosing decisions for morbidly obese patients.

Although it seemed that the clinically accepted target of AUC/MIC of 30 for levofloxacin could be achieved in most of the hospitalized patients, it was not the case with the ambulatory patients. However, owing to limited sample size, it was difficult to extrapolate the observed data to the general population. Given the high variability observed in the exposure of levofloxacin, due to variety of factors, it was recommended that one needs to be pragmatic in the dosing of levofloxacin to morbidly obese and obese populations. One such key factor was the renal CL of levofloxacin; patients who have exceptional renal CL of levofloxacin among the morbidly obese or ambulatory patients may be at the risk of subtherapeutic levels of levofloxacin.32 Therefore, it was recommended that a more detailed investigation may be necessary to determine the effects of renal CL as a key determinant of exposure of levofloxacin to establish appropriate dose size strategy in morbidly obese and obese patients to ensure clinical success for achieving the right AUC/MIC targets. However, it is important to point out that there was a single case report in a morbidly obese subject where no dosage adjustment was necessary in the dosing of levofloxacin.36

Back to Top | Article Outline

Doripenem

Kays et al (2014) have reported the pharmacokinetics of doripenem in morbidly obese subjects (n = 10) and have performed further population pharmacokinetic modeling. Also, Monte–Carlo simulations were used to delineate the pharmacodynamic response for the various dosing regimens of doripenem.37 The CL of doripenem was slower in obese patients as compared with nonobese patients.38 The VD of doripenem was somewhat smaller in morbidly obese patients as compared with nonobese patients.39

It was important to design suitable simulations to ensure the adequacy of pharmacodynamic exposures. In this regard, doripenem dosing regimen of 500 mg every 8 hours infused for 1 hour adequate systemic exposures to target pathogen, which had a target of MICs ≤ 2 μg/mL and ≤ 4 μg/mL.37 It was possible to achieve the required target of PTA >90% for most of the organisms with either 500 mg or 1 g doripenem dose every 8 hours over either 1- or 4-hour infusion. However, it was noted that doripenem dosing regimens could not adequately cover Acinetobacter species even after a longer 4-hour infusion and therefore, there was the need for another antifungal agent.37 As displayed in Figures 1, 2,37–39 the CL for doripenem was slower by >1.5-fold and the VSS was modestly lower in obese subjects as compared with nonobese subjects.

Back to Top | Article Outline

Fluconazole

In an interesting case report, an extremely morbidly obese patient (BMI: 84 kg/m2) with other complications was treated for fungal infection.40 The patient was receiving continuous venovenous hemofiltration (CVVH) given the deterioration of the renal function. Because fluconazole pharmacokinetic disposition is determined by efficient renal elimination with hepatic metabolism having little role to play and with CVVH in place, it was anticipated that a higher dose would be necessary to treat this morbidly obese patient. As fluconazole distribution is fairly restricted to the central compartment without excessive drug penetration into adipose tissues, the dosing decision was based on LBW rather than TBW. Accordingly, using LBW, an appropriate loading dose followed by maintenance dose of fluconazole was calculated and administered. In this case report, the desired target of AUC/MIC of >25 for fluconazole was achieved with the mix of a high starting loading dose and a maintenance dose. It was recommended that the basis of dosing using LBW was appropriate in dosing morbidly obese patients with fluconazole, and it may be necessary to have daily doses of 800 mg fluconazole in morbidly obese patients undergoing CVVH. As suggested by Figures 1, 2,40–42 at least 2–3-fold differences were observed in the CL and VSS values in the morbidly obese versus nonobese subjects and therefore, it supported the strategy of using BW descriptors to dose fluconazole in morbidly obese subjects.

Back to Top | Article Outline

Moxifloxacin

The pharmacokinetics of moxifloxacin was evaluated in 12 morbidly obese subjects; the design had a sequential oral dosing for 3 days followed by an intravenous dosing on day 4.43 Hitherto, no pharmacokinetic data were reported on the disposition of moxifloxacin in morbidly obese subjects. The key pharmacokinetic parameters such as Cmax, Tmax, half-life, CL, and VSS were comparable with that observed in healthy nonobese subjects.44 A single critical observation that was made in this study related to slower declining levels of moxifloxacin after 48 hours in morbidly obese subjects. This may be attributable to the initial adipose uptake of moxifloxacin and subsequent slower release into systemic circulation in the morbidly obese subjects. Because the 2 key parameters, namely CL and VSS that determine dosing, were not altered in morbidly obese as compared with other nonobese patients or healthy subjects, the authors concluded that the 400-mg dose of moxifloxacin was adequate in morbidly obese subjects.43

Back to Top | Article Outline

Oestelmevir carboxylate

Jittamala et al45 reported the pharmacokinetics of oseltamivir and oseltamivir carboxylate in nonobese and very obese subjects. The study consisted of low (75 mg) and high (150 mg) of oseltamivir carboxylate administered orally. The key pharmacokinetic parameters calculated in this study inclusive of CL and VSS of both oseltamivir carboxylate and oseltamivir suggested no significant differences between the 2 subject groups regardless of the doses administered in the study. Furthermore, it was found that trough levels of the active metabolite exceeded the threshold level required for the activity of the drug in either nonobese or very obese subjects. The authors concluded that the dosage adjustment for oseltamivir carboxylate was not necessary in dosing extremely obese populations.

An earlier study conducted by Pai and Lodise46 reported the pharmacokinetics of oseltamivir carboxylate and oseltamivir in morbidly obese patients (n = 21) that received a standard 75-mg oral dose of oseltamivir carboxylate. A population modeling approach was followed to define the pharmacokinetic disposition of the parent drug and the metabolite in the morbidly obese population. Various BW descriptors such as TBW, ABW, and LBW were also used for correlation with CL and VSS of both oseltamivir carboxylate and oseltamivir to probe for any casual relationships. However, none of the body mass descriptors yielded significant relationships with the pharmacokinetic parameters for either parent or the metabolite; and therefore, it was concluded that morbid obesity had no bearing on the disposition of oseltamivir carboxylate.46

Thorne–Humphrey et al47 compared the single dose and steady-state pharmacokinetics of both oseltamivir carboxylate and oseltamivir in morbidly obese (n = 10) and nonobese (n = 10) subjects using the standard oral dose of 75 mg. Although there appeared to be somewhat lower concentrations of both oseltamivir carboxylate and oseltamivir in the morbidly obese subjects as compared with nonobese individuals, it did not alter the VSS and CL of either of the 2 drugs. Moreover, the analysis of key pharmacokinetic parameters (VSS and total body clearance) with body size descriptors such as TBW and LBW yielded nonsignificant correlations either at single dose or at steady state for both parent and metabolite.

As displayed in Figures 1, 2,45 the VSS and CL parameters in the obese subjects were similar to those observed in nonobese subjects for both parent drug and the metabolite, with the exception of the formation CL of the metabolite, which appeared to be 1.5–1.7-fold higher in obese patients. However, this difference did not have any consequence on the pharmacodynamics of the drug as the Ctrough levels of the metabolite were found to be higher than the threshold efficacy levels.45

Back to Top | Article Outline

Voriconazole

In a pilot 2-way crossover study, the steady-state pharmacokinetics of voriconazole was evaluated in class II obese subjects along with appropriate nonobese controls.48 The pharmacokinetic parameters for voriconazole were similar between the obesity versus the nonobesity cohorts as evident in Figures 1, 2.48 Additional analysis performed with various body size descriptors such as TBW, LBW, ABW had no significant correlations suggesting the utility of any of the body size descriptors for dosing of voriconazole in subjects with obesity. The authors concluded that avoidance of oral dosing of voriconazole dosing on TBW was a good strategy because this may lead to a disproportional increase in the exposure of obese patients to voriconazole.48 Interestingly, in a case report in an obese allogeneic hematopoeitic cell transplant patient, the dosing of voriconazole was performed without any dosage adjustments.49 The chosen dose of voriconazole in this report yielded plasma concentrations and other key parameters similar to the values reported in nonobese individuals.49 Another recently reported case report of an obese patient dosed with voriconazole presented a real dilemma for developing an acceptable dosing strategy. In this study, the obese patient was genotyped to be a poor metabolizer of cytochrome P450 (CYP) 2C9 enzyme and therefore was expected to have a higher exposure to voriconazole, which is a substrate to the polymorphic CYP2C9 enzyme.50 Given the patient's obesity status, ABW method was followed to dose this patient (340 mg intravenous voriconazole, q12h was initiated). After dosing for a few days, the dose of the patient was reduced because of very high exposure of voriconazole observed in the systemic circulation.50 The new reduced dose (280 mg intravenous voriconazole, q12h) was continued for a few days; however, the patient was discontinued from the study because of an episode of corrected interval between start of the Q wave and end of the T wave prolongation, which may be as a consequence of high exposure to voriconazole.50 The authors recommended the avoidance of the use of TBW in dosing voriconazole. It was difficult to judge the usefulness of ABW in the dosing of voriconazole because the genotype status of the patient (poor metabolizer) was a confounding factor in this case report. Therefore, it may be prudent to monitor the plasma levels of voriconazole and also examine the genotype status if ABW strategy is being pursued for dosing of voriconazole in morbidly obese patients. As enumerated by Figures 1, 2,48,51 the VSS and CL values for voriconazole seemed to be similar between obese and nonobese subjects.

Back to Top | Article Outline

DATA ANALYSIS

Vancomycin

Blouin et al7 reported individual concentration data for vancomycin for various time points in morbidly obese and nonobese subjects. The concentration data at Cmax (end of infusion) and Ctrough (6-hour time point) were considered for the analysis. A simple linear regression analysis of the type,

was performed using the individual concentration data (both Cmax and Ctrough) with the corresponding individual value for the reported CL and VSS in the subjects. As suggested by Figure 3, there was an inverse correlation observed between the concentration data (Cmax or Ctrough) with the respective CL and VSS parameters. Typically, the morbidly obese patients had lower concentration levels and exhibited higher VSS and larger CL relative to nonobese subjects. Interestingly, the concentration data for the 2 points showed strong correlation and were defined by the following regression equations:

FIGURE 3

FIGURE 3

The distribution pattern of individual concentration data points in correlations of either VSS or CL for vancomycin showed clear demarcation of the values between morbidly obese versus nonobese subjects with obese patients having relatively lower levels of vancomycin at both peak and trough time points as compared with nonobese controls. Furthermore, the observed lower peak and trough concentrations for vancomycin in morbidly obese patients was further supported by the recent work of Mahmoud et al8 who found lower concentrations of vancomycin at peak and trough in obese children as compared with normal weight children. Based on the limited data, it may be possible to develop regression models to predict the VSS or CL values in morbidly obese patients using either Cmax or Ctrough data points (Figure 3). It also seemed that based on the analysis that both VSS and CL parameters are affected by morbid obesity and therefore dosing consideration in patients should not be empirical in nature. In this context, the work of Leong et al52 has emphasized the use of ABW for dosing consideration in extremely obese subjects.

Back to Top | Article Outline

Moxifloxacin

Using data reported by Kees et al43 and Stass et al,44 the individual Cmax concentration data were compiled for intravenous moxifloxacin at a high dose of 400 mg. The Cmax data gathered from obese subjects and nonobese subjects were subject to linear regression analysis with the corresponding individual VSS and CL values reported in the 2 studies.43,44 The analysis suggested that there was a poor correlation between Cmax versus VSS or Cmax versus CL (Figure 4). The 2 regression equations that defined VSS and CL of moxifloxacin were:

FIGURE 4

FIGURE 4

The distribution pattern of the Cmax values suggested an intermix of the values from morbidly obese and nonobese subjects. Hence, moxifloxacin may not be ideally suited for the development of linear regression models for predicting CL or VSS parameters from the Cmax data point. Our analysis is consistent with the findings and recommendations of Kees et al (2013) suggesting that CL and VSS would not be significantly altered because of morbid obesity.

Back to Top | Article Outline

Voriconazole

Using data reported by Pai and Lodise48 and Purkins et al,51 the individual Cmax concentration data were compiled for high oral dose of voriconazole. The Cmax data gathered from both obese subjects and nonobese subjects were subject to linear regression analysis with the corresponding individual VSS and CL values reported in the 2 studies.48,51 The analysis suggested that there was a strong correlation between Cmax versus VSS or Cmax versus CL (Figure 5). The 2 regression equations that defined VSS and CL of voriconazole were:

FIGURE 5

FIGURE 5

The distribution pattern of the Cmax suggested that the values from morbidly obese and nonobese subjects blended together in the model, although there were limited data from the morbidly obese subjects. However, unlike vancomycin, the Cmax data for voriconazole were positively correlated for both VSS and total body clearance regression models. Based on the limited data, it may be possible to develop linear regression models to predict the VSS or CL values in morbidly obese patients using Cmax data.

Back to Top | Article Outline

Oseltamivir

Using data reported by Jittamala et al,45 median (minimal, maximal) Cmax concentration data were compiled for both 75 and 150 mg oseltamivir doses in nonobese and obese subjects. The analysis was performed separately for the parent oseltamivir carboxylate and the active metabolite, oseltamivir. The regression analysis was performed using Cmax versus VSS and Cmax versus CL for both the parent and the metabolite. The parent drug showed strong correlation for both VSS and CL, suggesting that it may be possible to predict either VSS or CL using Cmax value in the morbidly obese subjects (Figure 6). The 2 regression equations that defined VSS and CL of oseltamivir carboxylate were:

FIGURE 6

FIGURE 6

Interestingly, the metabolite pharmacokinetics seemed to show differences between the 2 doses (75 vs. 150 mg) evaluated in this study. As is evident, the regression lines of the Cmax versus VSS and Cmax versus CL for the 75-mg group in nonobese subjects and obese subjects had a steeper slope value as compared with the similar regression lines generated for the corresponding Cmax versus Vss and Cmax versus CL analysis performed for the 150-mg group (Figure 7). However, it seemed that data from the morbidly obese and nonobese subjects blended well within the respective dose group suggesting a possible utility of linear regression models for the prediction of either VSS or CL in morbidly obese subjects.

FIGURE 7

FIGURE 7

The 2 regression equations that defined VSS and CL of oseltamivir at the 2 dose levels (75 and 150 mg) were:

Back to Top | Article Outline

DISCUSSION

The work of Brill et al53 has suggested influence of obesity on the various pathways of drug metabolism and elimination; these have consequences on the pharmacokinetics of drugs, both in adult and children obese patients. The data compiled in this review showed that generally CYP3A4-involved oxidative metabolism was relatively slower in obese subjects as compared with nonobese subjects. Hence, this resulted in a lower CL of the CYP3A4 substrates in obese subjects relative to nonobese subjects. In sharp contrast, other Phase I CYP enzymes such as CYP1A2, CYP2C9, CYP2C19, CYP2D6, and CYP2E1 showed an opposite effect in obese patients. The CL of substrates for this group of enzymes was higher in obese patients relative to nonobese patients. The examination of other pathways of drug elimination outside CYP oxidation was collated in this report. Accordingly pathways such as uridine diphosphate glucuronosyltransferase, xanthine oxidase, N-acetyltransferase seemed to be more functional in obese patients. Furthermore; renal mechanism of excretion involving either passive glomerular filtration process and/or tubular-mediated process was found to be faster in obese as compared with nonobese subjects.

The correlation of peak concentration or trough concentration with respect to either VSS or CL was conducted with the purpose to check whether or not the predictability of the parameters would occur when the concentration data from the obese subjects were combined with nonobese subjects as a single pool of data. It was interesting to note that with the exception of moxifloxacin, where the correlation was poor, the remaining drugs, namely, vancomycin, voriconazole, oseltamivir carboxylate, and oseltamivir showed strong correlations between the respective circulatory levels versus VSS or CL. However, given the sample size that evaluated such correlations in this work was very limited; one needs to view the findings as possible trends which may need a larger pool of data for reconfirmatory analysis. Despite this caveat, there may be an opportunity to develop a single time point strategy for the prediction of both VSS and CL from drugs such as vancomycin, voriconazole, and oseltamivir. As a casing point, a 3-point strategy comprising of peak concentration, trough concentration, and another time point in between was used to obtain population pharmacokinetics for vancomycin.6 Also, single time point strategy has been applied in other therapeutic areas and was shown to reasonably predict the exposure of several drugs namely, cyclosporine,54 tacrolimus,55,56 pravastatin,57 simvastatin,57 indinavir,58 etc.

The present review of 12 drugs clearly showed that development of dosing strategy to treat morbidly obese patients was less of an issue for most of the examined drugs with the exception of drugs such as vancomycin, ethambutol, and fluconazole, where a dosing strategy needs to be developed and appropriately followed up during the therapy (Table 2). Interestingly, all 3 drugs exhibited at least 2–3-fold greater VD in morbidly obese subjects relative to nonobese subjects. Hence, to compensate for the reduced circulatory levels, standard dosing of these drugs may not be sufficient. Therefore, an appropriate strategy of using TBW, ABW, or LBW needs to be applied as the case may be to ensure adequacy of blood levels of these drugs in morbidly obese subjects. Although it is well known that obesity leads to increased distribution of lipophilic drugs such as vancomycin, fluconazole, and ethambutol, the extent of distribution would depend on number of other variables such as protein binding, tissue permeability, and transporter dependency. On this topic, Cheynol (2000) noted that whereas morbid obesity is expected to increase the body deposition of fat, one should be equally cognizant that simultaneously an increase in LBW should be expected.59 In this context, Martin et al60 proposed the need for therapeutic drug monitoring in morbidly obese patients and have suggested consideration of recalibration of the therapeutic dose during the therapy in such patients to balance the risk of therapy failure versus overt toxicity occurrence. It is important to note that in today's regulatory scenario at the time of new drug approvals or market authorization, there is a big lacuna on the possible dosing strategies in obese subjects or patients.

Table 2

Table 2

In an interesting review, Morrish et al61 have explored the relationship between obesity-related physiological changes in relation to the body composition. Although the scaling metric of allometry (ie, relationship of BW such as TBW, LBW, ABW, etc. vs. pharmacokinetic parameter) is well established, the authors have provided a greater depth of the role of physiological changes in obesity, which may provide valuable guidance in the traditional scaling approaches and may lead to the identification of the dosing scalar to appropriately define the dosage adjustment of the drug pertaining to the patient population.60

Although the debate on dosing strategy for morbidly obese patients continues, one need to strongly emphasize the need for the generation of data by performing pharmacological, pharmacodynamics, and pharmacokinetic studies in morbidly obese/obese patients in a routine fashion to be included in new drug applications.

Back to Top | Article Outline

CONCLUSIONS

Based on the appraisal of the recently published pharmacokinetic data, dosing strategy for many of the anti-infective drugs is not expected to pose a challenge in morbidly obese subjects. However, for drugs such as vancomycin, ethambutol, and fluconazole a dosing strategy based on the appropriate body size descriptor should be developed and implemented with adequate frequency of drug monitoring during the therapy. Using limited data, strong correlations between Cmax/Ctrough versus VSS and CL were observed for drugs such as vancomycin, voriconazole, oseltamivir carboxylate, oseltamivir; there may be an opportunity for the prediction of parameters such as VSS or CL using a limited sampling strategy in morbidly obese patients.

Back to Top | Article Outline

REFERENCES

1. Janson B, Thursky K. Dosing of antibiotics in obesity. Curr Opin Infect Dis. 2012;25:634–649.
2. Flegal KM, Carroll MD, Kuczmarski RJ, et al Overweight and obesity in the United States: prevalence and trends, 1960–1994. Int J Obes Relat Metab Disord. 1998;22:39–47.
3. Flegal KM, Carroll MD, Kit BK, et al Prevalence of obesity and trends in the distribution of body mass index among US adults, 2012 1999–2010. JAMA. 2012;307:491–497.
4. Knibbe CA, Brill MJ, van Rongen A, et al Drug disposition in obesity: toward evidence-based dosing. Annu Rev Pharmacol Toxicol. 2015;55:149–167.
5. Sankaralingam S, Kim RB, Padwal RS. The impact of obesity on the pharmacology of medications used for cardiovascular risk factor control. Can J Cardiol. 2015;31:167–176.
6. Adane ED, Herald M, Koura F. Pharmacokinetics of vancomycin in extremely obese patients with suspected or confirmed Staphylococcus aureus infections. Pharmacotherapy. 2015;35:127–139.
7. Blouin RA, Bauer LA, Miller DD, et al Vancomycin pharmacokinetics in normal and morbidly obese subjects. Antimicrob Agents Chemother. 1982;21:575–580.
8. Mahmoud MA, Ebid AH, Shouman SA, et al Pharmacokinetics of vancomycin in oncology egyptian paediatrics: a dosage adjustment trial. Indian J Pharm Sci. 2014;76:82–86.
9. Bhalodi AA, Papasavas PK, Tishler DS, et al Pharmacokinetics of intravenous linezolid in moderately to morbidly obese adults. Antimicrob Agents Chemother. 2013;57:1144–1149.
10. Meagher AK, Forrest A, Rayner CR, et al Population pharmacokinetics of linezolid in patients treated in a compassionate-use program. Antimicrob Agents Chemother. 2013;47:548–553.
11. Stalker DJ, Jungbluth GL, Hopkins NK, et al Pharmacokinetics and tolerance of single- and multiple-dose oral or intravenous linezolid, an oxazolidinone antibiotic, in healthy volunteers. J Antimicrob Chemother. 2003;51:239–246.
12. Stein GE, Schooley SL, Peloquin CA, et al Pharmacokinetics and pharmacodynamics of linezolid in obese patients with cellulitis. Ann Pharmacother. 2005;39:427–432.
13. Mersfelder TL, Smith CL. Linezolid pharmacokinetics in an obese patient. Am J Health Syst Pharm. 2005;53:464–467.
14. Tsuji Y, Hiraki Y, Matsumoto K, et al Evaluation of the pharmacokinetics of linezolid in an obese Japanese patient. Scand J Infect Dis. 2012;44:626–629.
15. Hall RG II, Swancutt MA, Meek C, et al Ethambutol pharmacokinetic variability is linked to body mass in overweight, obese, and extremely obese people. Antimicrob Agents Chemother. 2012;56:1502–1507.
16. Gumbo T. Chemotherapy of tuberculosis, mycobacterium avium complex disease, and leprosy. In: Brunton LL, Chabner B, Knollmann B, eds. Goodman & Gilman's the Pharmacological Basis of Therapeutics. 12th ed. New York, NY: McGraw Hill Medical; 2010:1549–1570.
17. Baker SS, Baker RD, Liu W, et al Role of alcohol metabolism in non-alcoholic steatohepatitis. PLoS One. 2010;5:e9570.
18. Pasipanodya JG, Weis S, Gumbo T. 2010. Weight not diabetes mellitus predicts failure of anti-tuberculosis therapy in treated tuberculosis patients with concurrent diabetes mellitus. Abstr. 50th Intersci. Conf. Antimicrob. Agents Chemother., abstr A1-059.
19. World Health Organization. Treatment of Tuberculosis: Guidelines. WHO/HTM/TB/2009.420. 2010. 4th ed. Geneva, Switzerland: World Health Organization; 2010.
20. Hasenbosch RE, Alffenaar JW, Koopmans SA, et al Ethambutol-induced optical neuropathy: risk of overdosing in obese subjects. Int J Tuberc Lung Dis. 2008;12:967–971.
21. Peloquin CA, Bulpitt AE, Jaresko GS, et al Pharmacokinetics of ethambutol under fasting conditions, with food, and with antacids. Antimicrob Agents Chemother. 1999;43:568–572.
22. Lee CS, Brater DC, Gambertoglio JG, et al Disposition kinetics of ethambutol in man. J Pharmacokinet Biopharm. 1980;8:335–346.
23. Cheatham SC, Fleming MR, Healy DP, et al Steady-state pharmacokinetics and pharmacodynamics of meropenem in morbidly obese patients hospitalized in an intensive care unit. J Clin Pharmacol. 2014;54:324–330.
24. Li C, Du X, Kuti JL, et al Clinical pharmacodynamics of meropenem in patients with lower respiratory tract infections. Antimicrob Agents Chemother. 2007;51:1725–1730.
25. Cheatham SC, Kays MB, Smith DW, et al Steady‐state pharmacokinetics and pharmacodynamics of meropenem in hospitalized patients. Pharmacotherapy. 2008;28:691–698.
26. Kitzes‐Cohen R, Farin D, Piva G, et al Pharmacokinetics and pharmacodynamics of meropenem in critically ill patients. Int J Antimicrob Agents. 2002;19:105–110.
27. Leroy A, Fillastre JP, Borsa-Lebas F, et al Pharmacokinetics of meropenem (ICI 194,660) and its metabolite (ICI 213,689) in healthy subjects and in patients with renal impairment. Antimicrob Agents Chemother. 1992;36:2794–2798.
28. Mouton JW, van den Anker JN. Meropenem clinical pharmacokinetics. Clin Pharmacokinet. 1995;28:275–286.
29. Pai MP. Serum and urine pharmacokinetics of tigecycline in obese class III and normal weight adults. J Antimicrob Chemother. 2014;69:190–199.
30. Muralidharan G, Micalizzi M, Speth J, et al Pharmacokinetics of tigecycline after single and multiple doses in healthy subjects. Antimicrob Agents Chemother. 2005;49:220–229.
31. Meagher AK, Ambrose PG, Grasela TH, et al The pharmacokinetic and pharmacodynamic profile of tigecycline. Clin Infect Dis. 2005;41(Suppl):S333–S5340.
32. Cook AM, Martin C, Adams VR, et al Pharmacokinetics of intravenous levofloxacin administered at 750 milligrams in obese adults. Antimicrob Agents Chemother. 2011;55:3240–3243.
33. Chien SC, Chow AT, Natarajan J, et al Absence of age and gender effects on the pharmacokinetics of a single 500-milligram oral dose of levofloxacin in healthy subjects. Antimicrob Agents Chemother. 1997;41:1562–1565.
34. Chien SC, Rogge MC, Gisclon LG, et al Pharmacokinetic profile of levofloxacin following once-daily 500-milligram oral or intravenous doses. Antimicrob Agents Chemother. 1997;41:2256–2260.
35. Chow AT, Fowler C, Williams RR, et al Safety and pharmacokinetics of multiple 750-milligram doses of intravenous levofloxacin in healthy volunteers. Antimicrob Agents Chemother. 2001;45:2122–2125.
36. Luque S, Grau S, Valle M, et al Levofloxacin weight-adjusted dosing and pharmacokinetic disposition in a morbidly obese patient. J Antimicrob Chemother. 2011;66:1653–1654.
37. Kays MB, Fleming MR, Cheatham SC, et al Comparative pharmacokinetics and pharmacodynamics of doripenem and meropenem in obese patients. Ann Pharmacother. 2014;48:178–186.
38. Cirillo I, Vaccaro N, Turner K, et al Pharmacokinetics, safety, and tolerability of doripenem after 0.5-, 1-, and 4-hour infusions in healthy volunteers. J Clin Pharmacol. 2009;49:798–806.
39. Nandy P, Samtani MN, Lin R. Population pharmacokinetics of doripenem based on data from phase 1 studies with healthy volunteers and phase 2 and 3 studies in critically ill patients. Antimicrob Agents Chemother. 2010;54:2354–2359.
40. Lopez ND, Phillips KM. Fluconazole pharmacokinetics in a morbidly obese, critically ill patient receiving continuous venovenous hemofiltration. Pharmacotherapy. 2014;34:e162–168.
41. Ripa S, Ferrante L, Prenna M. Pharmacokinetics of fluconazole in normal volunteers. Chemotherapy. 1993;39:6–12.
42. Yeates RA, Ruhnke M, Pfaff G, et al The pharmacokinetics of fluconazole after a single intravenous dose in AIDS patients. Br J Clin Pharmacol. 1994;38:77–79.
43. Kees MG, Weber S, Kees F, et al Pharmacokinetics of moxifloxacin in plasma and tissue of morbidly obese patients. J Antimicrob Chemother. 2011;66:2330–2335.
44. Stass H, Dalhoff A, Kubitza D, et al Pharmacokinetics, safety, and tolerability of ascending single doses of moxifloxacin, a new 8-methoxy quinolone, administered to healthy subjects. Antimicrob Agents Chemother. 1998;42:2060–2065.
45. Jittamala P, Pukrittayakamee S, Tarning J, et al Pharmacokinetics of orally administered oseltamivir in healthy obese and nonobese Thai subjects. Antimicrob Agents Chemother. 2014;58:1615–1621.
46. Pai MP, Lodise TP Jr. Oseltamivir and oseltamivir carboxylate pharmacokinetics in obese adults: dose modification for weight is not necessary. Antimicrob Agents Chemother. 2011;55:5640–5645.
47. Thorne-Humphrey LM, Goralski KB, Slayter KL, et al Oseltamivir pharmacokinetics in morbid obesity (OPTIMO trial). J Antimicrob Chemother. 2011;66:2083–2091.
48. Pai MP, Lodise TP. Steady-state plasma pharmacokinetics of oral voriconazole in obese adults. Antimicrob Agents Chemother. 2011;55:2601–2605.
49. Dickmeyer NJ, Kiel PJ. Dosing voriconazole in an obese patient. Clin Infect Dis. 2011;53:745.
50. Moriyama B, Jarosinski PF, Figg WD, et al Pharmacokinetics of intravenous voriconazole in obese patients: implications of CYP2C19 homozygous poor metabolizer genotype. Pharmacotherapy. 2013;33:e19–22.
51. Purkins L, Wood N, Greenhalgh K, et al Voriconazole, a novel wide-spectrum triazole: oral pharmacokinetics and safety. Br J Clin Pharmacol. 2003;56(Suppl):10–16.
52. Leong JV, Boro MS, Winter M. Determining vancomycin clearance in an overweight and obese population. Am J Health Syst Pharm. 2011;68:599–603.
53. Brill MJ, Diepstraten J, van Rongen A, et al Impact of obesity on drug metabolism and elimination in adults and children. Clin Pharmacokinet. 2012;51:277–304.
54. Hangler HB, Ruttmann E, Geltner C, et al Single time point measurement by C2 or C3 is highly predictive in cyclosporine area under the curve estimation immediately after lung transplantation. Clin Transpl. 2008;22:35–40.
55. Takeuchi H, Matsuno N, Hirano T, et al Evidence of different pharmacokinetics between cyclosporine and tacrolimus in renal transplant recipients: why cyclosporine is monitored by C2 level and tacrolimus by trough level. Transpl Proc. 2008;40:2240–2242.
56. Niioka T, Miura M, Kagaya H, et al A limited sampling strategy to estimate the area under the concentration-time curve of tacrolimus modified-release once-daily preparation in renal transplant recipients. Ther Drug Monit. 2013;35:228–232.
57. Srinivas NR. Limited sampling strategy for the prediction of area under the curve (AUC) of statins: reliability of a single time point for AUC prediction for pravastatin and simvastatin. Drug Res. 2015. Epub ahead of print.
58. Srinivas NR. Differences in the prediction of area under the curve for a protease inhibitor using trough versus peak concentration: feasibility assessment using published pharmacokinetic data for indinavir. Am J Ther. 2015 Aug 18. Epub ahead of print.
59. Cheymol G. Effects of obesity on pharmacokinetics: implications for drug therapy. Clin Pharmacokinet. 2000;39:215–231.
60. Martin JH, Saleem M, Looke D. Therapeutic drug monitoring to adjust dosing in morbid obesity—a new use for an old methodology. Br J Clin Pharmacol. 2012;73:685–690.
61. Morrish GA, Pai MP, Green B. The effects of obesity on drug pharmacokinetics in humans. Expert Opin Drug Metab Toxicol. 2011;7:697–706.
Keywords:

Pharmacokinetics; anti-infective drugs; morbid obesity; vancomycin; oseltamivir; voriconazole; doripenem; meropenem; fluconazole; linezolid; ethambutol; tigecycline; moxifloxacin; levofloxacin

Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.