Population Modeling and Monte Carlo Simulation Study of the Pharmacokinetics and Antituberculosis Pharmacodynamics of Rifampin in Lungs

Sylvain Goutelle; Laurent Bourguignon; Pascal H Maire; Michael Van Guilder; John E Conte, Jr; Roger W Jelliffe

doi:10.1128/AAC.01520-08

. 2009 Apr 20;53(7):2974–2981. doi: 10.1128/AAC.01520-08

Population Modeling and Monte Carlo Simulation Study of the Pharmacokinetics and Antituberculosis Pharmacodynamics of Rifampin in Lungs^▿

Sylvain Goutelle ^1,2,^*, Laurent Bourguignon ^1,2, Pascal H Maire ^1,2, Michael Van Guilder ³, John E Conte Jr ^4,5, Roger W Jelliffe ³

PMCID: PMC2704682 PMID: 19380594

Abstract

Little information exists on the pulmonary pharmacology of antituberculosis drugs. We used population pharmacokinetic modeling and Monte Carlo simulation to describe and explore the pulmonary pharmacokinetics and pharmacodynamics of rifampin (RIF; rifampicin). A population pharmacokinetic model that adequately described the plasma, epithelial lining fluid (ELF), and alveolar cell (AC) concentrations of RIF in a population of 34 human volunteers was made by use of the nonparametric adaptive grid (NPAG) algorithm. The estimated concentrations correlated well with the measured concentrations, and there was little bias and good precision. The results obtained with the NPAG algorithm were then imported into Matlab software to perform a 10,000-subject Monte Carlo simulation. The ability of RIF to suppress the development of drug resistance and to induce a sufficient bactericidal effect against Mycobacterium tuberculosis was evaluated by calculating the proportion of subjects achieving specific target values for the maximum concentration of drug (C_max)/MIC ratio and the area under the concentration-time curve from time zero to 24 h (AUC_0-24)/MIC ratio, respectively. At the lowest MIC (0.01 mg/liter), after the administration of one 600-mg oral dose, the rates of target attainment for C_max/MIC (≥175) were 95% in ACs, 48.8% in plasma, and 35.9% in ELF. Under the same conditions, the target attainment results for the killing effect were 100% in plasma (AUC_0-24/MIC ≥ 271) but only 54.5% in ELF (AUC_0-24/MIC ≥ 665). The use of a 1,200-mg RIF dose was associated with better results for target attainment. The overall results suggest that the pulmonary concentrations obtained with the standard RIF dose are too low in most subjects. This work supports the need to evaluate higher doses of RIF for the treatment of patients with tuberculosis.

Despite the discovery of drugs highly active against Mycobacterium tuberculosis in the second half of the 20th century, tuberculosis (TB) remains one of the biggest killers among all infectious diseases. In 2007, there were an estimated 1.75 million deaths from TB, according to the World Health Organization (37).

Treatments for TB, which use a combination of several antibiotics, are usually effective when they are well conducted, since 95% of patients who complete the directly observed therapy are cured (38). However, treatment is long and difficult, which hampers patient compliance. The emergence of multidrug-resistant and, more recently, extensively drug resistant Mycobacterium tuberculosis is another major issue in current TB treatment (1, 6).

In addition to poor patient compliance and drug resistance, low concentrations of anti-TB drugs may negatively affect the treatment outcome. Many studies have found low plasma concentrations of anti-TB drugs in both human immunodeficiency virus (HIV)-infected patients (25, 30, 35) and non-HIV-infected patients (15, 19, 34). Low plasma levels were associated with a longer time to respond to treatment, treatment failure, and the emergence of drug resistance (15, 19, 35).

Among the drugs used for the treatment of TB, rifampin (RIF; rifampicin) is often viewed as the most important drug because it is active against both extracellular and intracellular M. tuberculosis, even under slow replicative conditions (28). Previous work has shown that the antibacterial effect of RIF against M. tuberculosis is concentration dependent, based on the high correlation observed between its killing effect and the area under the concentration-time curve (AUC) of RIF (8, 9). Moreover, it has been shown that the postantibiotic effect of RIF and the suppression of resistance were also concentration dependent. Gumbo et al. have demonstrated that those effects were best correlated with the maximum concentration of drug (C_max)/MIC ratio (8).

While the lung is the major site of M. tuberculosis infection, little is known about the pulmonary pharmacokinetics (PK) and pharmacodynamics (PD) of RIF. No study has specifically addressed the concentration-dependent behavior of RIF in the lungs. In this report, we present a nonparametric PK model which describes the plasma and pulmonary concentrations of RIF in 34 human volunteers. This model was used to analyze the concentration-dependent effect profile of RIF for two dosage regimens in a 10,000-subject Monte Carlo simulation.

(This work was presented in part as a scientific poster at the 17th meeting of the Population Approach Group in Europe, 18 to 20 June 2008, Marseille, France.)

MATERIALS AND METHODS

Subject population, study design, and analytical method.

Details about the subject population, informed consent, and the design of the study have been described previously (3). Briefly, it was a prospective, nonblinded study. Forty adult subjects without TB were recruited and included 10 women without AIDS, 10 men without AIDS, 10 women with AIDS, and 10 men with AIDS.

The subjects received RIF at 600 mg orally once daily for 5 days. RIF concentrations were measured in plasma at approximately 2 h and 4 h, and RIF concentrations in epithelial lining fluid (ELF) and alveolar cells (ACs) recovered by bronchoalveolar lavage (BAL) were measured at approximately 4 h after the administration of the last dose. All sampling times were recorded precisely for each patient. RIF concentrations were measured by high-performance liquid chromatography, as described elsewhere (4). For the determination of the RIF concentration in ELF and ACs, the volume of ELF recovered by BAL was calculated by using the urea dilution method (27), while the volume of ACs was estimated from the cell count performed in BAL fluid.

Population PK modeling.

The population PK analysis was carried out with the nonparametric adaptive grid (NPAG) algorithm in the MM-USCPACK software (16). RIF concentrations in plasma, ELF, and ACs were modeled simultaneously. Oral bioavailability was arbitrarily fixed at 100%. The structural PK model was a three-compartment open model consisting of plasma, ELF, and ACs with first-order absorption and first-order processes for all transfers. The model is described by the following system of ordinary differential equations:

where X_a, X₁, X₂, and X₃ are the amounts of drug (in milligrams) in the absorption compartment, the central compartment (plasma concentrations), the ELF compartment, and the AC compartment, respectively; K_a (h⁻¹) is the absorption rate constant; K_e (h⁻¹) is the elimination rate constant from the central compartment; and K₁₂, K₂₁, K₂₃, and K₃₂ are the between-compartment transfer rate constants (all in h⁻¹). The other pharmacokinetic parameters are the three volumes of distribution (V) from the following output equations:

where Y₁, Y₂, and Y₃ are the RIF concentrations (in mg/liter) in the central compartment, the ELF compartment, and the AC compartment, respectively; V_C represents V in the central compartment (in liters), V_ELF is V in the ELF compartment (in liters), and V_CELL is V in the AC compartment (in liters).

In the NPAG modeling procedure, each drug concentration was weighted by the reciprocal of the assay variance at that concentration. For plasma RIF concentrations, the overall assay error pattern was determined from the assay validation data published by Conte and colleagues (4) by fitting a second-order polynomial to the plot of the assay standard deviation (y) versus the mean concentrations (x). The resulting equation was as follows: y = 0.1141 − 0.0106x + 0.0021x² (r² = 0.977). For the ELF and AC concentrations, as the assay variance data were not sufficient to get a similar polynomial relationship, the plasma assay error pattern was used as the weighting scheme.

The log-likelihood criterion was used to assess the parameter estimates. In addition, individual predicted concentrations were computed by using each patient's individual Bayesian posterior parameter joint density. Goodness of fit was assessed by regression over the predicted-observed concentration plots and the coefficient of determination. Bias (mean weighted error) and precision (bias-adjusted mean weighted squared error) were used to assess predictive performance. The NPAG analysis yielded a joint parameter distribution consisting of 34 discrete support points.

Monte Carlo simulation.

Matlab (version 6.5) software (MathWorks, Natick, MA) was used to perform a 10,000-subject Monte Carlo simulation. All 34 support points corresponding to the nonparametric population parameter joint density, as well as the population parameter variance-covariance matrix from the NPAG analysis, were exported into Matlab. The rand function of Matlab was used to select support points randomly according to their probability. Each support point was then treated as a mean vector, and the population variance-covariance matrix from NPAG was placed around that mean. This was done by using the mvrnd function in Matlab. The bounds for each PK parameter were set in accordance with the NPAG population analysis results.

The 24-h concentration-time profiles in plasma, ELF, and ACs were calculated for each subject, after the administration of a single oral dose of RIF. Two doses of RIF were studied: a 600-mg standard dose and a 1,200-mg dose. In order to assess the PD effects of RIF against M. tuberculosis in this simulated population, the ratio of the C_max (in mg/liter) to the MIC and the ratio of the AUC from time zero to 24 h (AUC_0-24; in mg · h/liter) to the MIC were computed for the plasma, ELF, and AC concentration profiles for all subjects. Six RIF MICs for M. tuberculosis (0.01, 0.025, 0.05, 0.1, 0.5, and 1 mg/liter) were evaluated. In order to avoid unusual and unrealistic drug exposures, the data for each subject with a plasma AUC_0-24 of less than 10 or more than 1,000 mg · h/liter were discarded and replaced by resampling data obtained by using the randomization and simulation process.

The percentage of simulated subjects who achieved a C_max/MIC of ≥175 was calculated for the plasma, ELF, and AC concentrations. This target was selected because it was shown to be correlated with the suppression of RIF resistance in vitro in a single drug experiment with a hollow-fiber system (8). In addition, the percentage of simulated subjects who achieved an AUC_0-24/MIC of ≥271 in plasma and also an AUC_0-24/MIC of ≥665 in ELF was calculated. It has been shown that the AUC/MIC PD index best correlated with the bactericidal effect of RIF against M. tuberculosis (8, 9). Those target values of plasma AUC_0-24/MIC of ≥271 and ELF AUC_0-24/MIC of ≥665 were associated with a 1-log₁₀-CFU reduction in the total lung bacterial load in mice for the plasma target value and in infected macrophages for the ELF target value after 4 to 6 days of treatment (9).

For ease of comparison with published results, plasma concentrations were expressed in two different ways, according to the index considered. The total drug concentration was used for the calculation of AUC_0-24/MIC, in accordance with the work of Jayaram et al. (9), while the free drug concentration was used to calculate C_max/MIC, as Gumbo et al. did (8). The RIF free drug concentration in plasma was determined by assuming 80% protein binding. As the level of RIF protein binding in ELF and AC is currently unknown, it was not considered in the calculation of C_max/MIC ratios in those two media. It has been reported that the concentration of protein in ELF is low, so the concentration of drug bound to protein might also be considered low or negligible in ELF (14).

RESULTS

PK modeling with NPAG algorithm.

Among the initial group of 40 patients, the data for 34 remained in the final PK analysis. One subject's data were discarded at the beginning, since he had no detectable RIF concentration in plasma, ELF, or ACs at any sampling time. Five other subjects were excluded during the early stages of the analysis. Those five patients seemed to have a much delayed RIF absorption since the mean plasma concentration 2 h after the administration of the dose was only 0.32 mg/liter (range, 0 to 0.80 mg/liter), while the mean plasma concentration 4 h after the administration of the dose was 10.95 mg/liter (range, 4.22 to 24.6 mg/liter).

The characteristics of the six excluded patients were as follows: mean age, 38.3 years; mean body weight, 68.2 kg; mean height, 169.1 cm; sex ratio (number of men/number of women), 2/4; and HIV infection status (number of HIV-infected individuals/number of HIV-noninfected individuals), 5/1. Three of six subjects were Caucasian, two were black, and one was Hispanic. It is noteworthy that five of those six patients had AIDS.

The three-compartment, nine-parameter model fit the data very well. The population parameter values estimated by use of the NPAG algorithm are summarized in Table 1. The use of lag time as a fixed (30 min) or random parameter was tested but did not improve either the fit or the estimation performance (data not shown).

TABLE 1.

Summary statistics of population PK parameters estimated by use of NPAG algorithm

Parameter	K_a (h⁻¹)	K_e (h⁻¹)	K₁₂ (h⁻¹)	K₂₁ (h⁻¹)	K₂₃ (h⁻¹)	K₃₂ (h⁻¹)	V_C (liters)	V_ELF (liters)	V_CELL (liters)
Mean	2.66	1.62	34.74	23.98	39.98	21.24	9.27	96.61	47.72
Median	2	1.1	39	19.38	46	22	5.57	80.6	26.87
SD	2.74	1.53	12.35	13.3	13.97	12.99	11.2	64.06	54.32

Open in a new tab

Predicted concentrations based on the Bayesian posterior parameter estimates correlated very well with the observed concentrations, as depicted in Fig. 1. The best-fit regression lines were close to the line of identity; and the coefficients of determination were 0.95, >0.99, and >0.99 for plasma, ELF, and AC concentrations, respectively. Bias and precision of the Bayesian posterior concentration estimates obtained by use of the NPAG algorithm are shown in Table 2.

FIG. 1. — Plots of observed versus predicted concentrations in plasma (A), ELF (B), and ACs (C). Predicted concentrations were calculated by using each patient's individual Bayesian posterior parameter joint density. The solid line represents the best regression line, and the dashed line is the line of identity (where y = x).

TABLE 2.

Predictive performance of Bayesian posterior concentration estimates

Compartment	Rifampin concn
Compartment	Bias^a (mg/liter)	Precision^b (mg²/liter²)
Plasma	−0.144	1.797
ELF	0.237	0.315
ACs	−0.0282	0.00418

Open in a new tab

Calculated as mean weighted error.

Calculted as bias-adjusted mean weighted squared error.

A profile of the RIF concentrations in plasma, ELF, and ACs over 24 h after the administration of one oral 600-mg dose of RIF in a representative subject is shown in Fig. 2.

FIG. 2. — Example of a 24-h RIF concentration-time profile in plasma, ELF, and ACs for one subject after the administration of a single oral 600-mg dose of RIF. Concentrations were calculated by using the individual Bayesian posterior PK parameter values estimated for 1 of the 34 human volunteers.

Monte Carlo simulation in Matlab.

The PK parameter values for the 10,000 simulated subjects are summarized in Table 3. While no statistical comparison was done, one can see that most mean parameter values were reasonably close to those estimated for the 34 volunteers by use of the NPAG algorithm. However, the median parameter values for some parameters (e.g., K_a, V_C, and V_CELL) were significantly greater than those estimated by use of the NPAG algorithm. This was because of the change in the shape of the parameter distributions caused by putting the overall covariance matrix around each support point and then sampling from those altered distributions. Figure 3 depicts such a change in the shape of the distribution for V_C. Parameter standard deviation values were slightly lower than those estimated for the 34 volunteers by use of the NPAG algorithm.

TABLE 3.

PK parameter values obtained from the 10,000-subject Monte Carlo simulation

Parameter	K_a (h⁻¹)	K_e (h⁻¹)	K₁₂ (h⁻¹)	K₂₁ (h⁻¹)	K₂₃ (h⁻¹)	K₃₂ (h⁻¹)	V_C (liters)	V_ELF (liters)	V_CELL (liters)
Mean	3.59	1.42	33.34	23.78	39.55	20.27	11.56	102.6	49.94
Median	3.12	1.2	35.22	23.11	40.02	19.23	10.16	102.8	43.66
SD	2.42	0.97	11.22	11.14	9.44	11.22	8.05	52.1	33.65

Open in a new tab

FIG. 3. — Example of change in the shape of the PK parameter distributions between NPAG results and Monte Carlo simulation results. (A) Histogram of the *V_C* parameter distribution, as calculated by use of the NPAG algorithm; (B) histogram of the distribution of the same parameter obtained from the Monte Carlo simulation.

The expected ability of the 600-mg standard dose to suppress RIF resistance, as well as that of a 1,200-mg dose, expressed as the percentage of subjects achieving a C_max/MIC ratio of ≥175, is displayed in Fig. 4. After the administration of a single 600-mg RIF dose, the percentage of subjects achieving C_max/MIC ratios of ≥175 for plasma and ELF was less than 50% (48.8% and 35.9%, respectively), even at the lowest MIC (0.01 mg/liter). The calculated probability of target attainment was greater in ACs at low MICs. These probabilities were calculated to be 95.0% and 62.2% at MICs of 0.01 and 0.025 mg/liter, respectively. However, the probability of RIF suppression of resistance was less than 50% for both plasma and lungs for MICs of ≥0.05 mg/liter, with very low suppression probability values for plasma and ELF (8.1% and 5.6% at a MIC of 0.05 mg/liter, respectively).

FIG. 4. — Percentage of subjects achieving C_max/MIC ratios of ≥175 in plasma (•), ELF (▪), and ACs (▴) after the administration of a single oral 600-mg dose (A) and 1,200 mg dose (B) of RIF. Calculations were done by using the free drug concentrations in plasma and the total drug concentrations in ELF and ACs. Values of less than 5% have no tag, for ease of reading.

After the administration of a 1,200-mg dose, as expected, the percentage of subjects achieving C_max/MIC ratios of ≥175 was significantly greater for all concentrations (98.7%, 59.7%, and 80.3% for plasma, ELF, and ACs, respectively, at the lowest MIC of 0.01 mg/liter). However, this percentage declined as the MICs increased. It was less than 50% for all concentrations when MICs were ≥0.05 mg/liter.

Figure 5 depicts the target attainment in plasma (AUC_0-24/MIC ≥ 271) and ELF (AUC_0-24/MIC ≥ 665) related to the killing effect of RIF against M. tuberculosis in the 10,000 simulated subjects for the two RIF doses (600 mg and 1,200 mg).

FIG. 5. — Percentage of subjects achieving AUC_0-24/MIC ratios of ≥271 in plasma (•) and AUC_0-24/MIC ratios of ≥665 in ELF (▪) after the administration of a single oral 600-mg dose (A) and 1,200 mg dose (B) of RIF. Calculations were done by using the total drug concentrations in all compartments. Values of less than 5% have no tag for ease of reading.

After administration of the 600-mg oral dose, the target attainment rates in plasma were high, with values of more than 75% at MICs of ≤0.1 mg/liter. However, those rates were significantly less at higher MICs. In ELF, the target attainment rates were significantly lower than those in plasma at all MICs. The proportion of subjects who achieved AUC_0-24/MIC ratios of ≥665 in ELF was only 54% at a MIC of 0.01 mg/liter and was less than 30% at MICs of ≥0.025 mg/liter.

Increasing the dose to 1,200 mg in the simulations resulted in higher target attainment rates in plasma as well as in ELF. In plasma, the target attainment scores were especially improved for MICs of ≥0.1 mg/liter. In ELF, the proportion of subjects who achieved the AUC_0-24/MIC target value was increased up to 68% at a MIC of 0.01 mg/liter, but it was still less than 50% at MICs of 0.025 mg/liter and higher.

DISCUSSION

In the treatment of pulmonary TB, as for other pulmonary infectious diseases, it is important to ensure the delivery of a sufficient amount of drug to the site of infection, since the intrapulmonary concentrations of antibiotics are thought to support the antibacterial effects of the drugs.

Although isolated drug measurements at the infection site and drug concentration ratios (e.g., lung concentration divided by blood concentration at a given time) are informative, they do not permit the proper evaluation of the dynamics of drug exposure over time.

We used population PK modeling and Monte Carlo simulation to explore the pulmonary PK and PD of RIF. Such methods have proven to be useful in pulmonary PK-PD studies with other antibiotics (17, 18).

By using a nonparametric method, a population model that adequately described the plasma, ELF, and AC concentrations of RIF in a population of 34 volunteers was created. The three-compartment, nine-parameter model was the simplest structure describing the three-output system. This model had no covariates. The model structure is consistent with those of previous models describing plasma RIF concentrations; those models were linear one-compartment models and included few or no covariates (24, 36).

RIF is well known to stimulate its own metabolism by autoinduction (7). This results in a progressive reduction of drug exposure (as measured by the AUC_0-24) over the first several days of treatment. This feature was not included in the modeling because all concentrations were measured on the fifth day of treatment, all during the same dosing interval. Even if the time for autoinduction is not precisely known, the concentrations measured on the fifth day of therapy might not yet reflect the steady state in the autoinduction. This should be considered in the interpretation of our results.

Data for six patients were excluded from the PK analysis: one had no detectable RIF concentration in any sample, and five had significantly delayed RIF absorption. Of note, five of the six excluded patients were HIV infected. Sahai et al. reported poor absorption of anti-TB drugs in HIV-infected subjects, as measured by low C_maxs (30). In another study with 34 HIV-infected patients with TB, the absorption of RIF was found to be much delayed and highly variable (26). However, such variability in drug absorption has also been described in healthy subjects. Using a nonparametric method and a rich data set, Peloquin et al. (24) identified two subpopulations of RIF absorbers in a group of 24 healthy volunteers: smooth (more rapid) absorbers and low (slower) absorbers. Smooth absorbers presented a mean absorption rate constant three times greater than that for the slower absorbers (24).

In the 34 subjects remaining in our final analysis, no significant difference in individual PK parameter values between HIV-infected subjects (n = 15) and HIV-noninfected subjects (n = 19) was found (data not shown). Other PK studies found similar results (2, 36). However, as the overall published data show no agreement, further studies with patients with HIV infection are needed to clarify the possible relationships between the PK of RIF and HIV infection.

The model created here can serve as a benchmark for future pulmonary PK-PD studies with RIF. In comparison with noninfected subjects, patients with TB might have altered PK of RIF because of their specific pulmonary lesions. Further studies with infected patients are required to explore this question.

The simulation method used in this work is somewhat different from the methods used in comparable studies. In two recent studies concerning the pulmonary PK of two antibiotics, after their nonparametric population analysis, Lodise et al. performed a Monte Carlo simulation using only the mean or the median of the population PK parameters, along with the covariance matrix estimated by use of the NPAG algorithm (17, 18). In comparison, our method used all the population information, as represented by all the various support points calculated by use of the NPAG algorithm, to generate new individual PK parameter vectors. We believe that this simulation method represents a logical extension of the nonparametric population approach. It is likely, however, that the use of the population covariance matrix around each support point creates too great a variance in the simulation process. This could result in the selection of very unusual values for the PK parameters in the random process. For this reason, unrealistic values for the PK parameters and drug exposures were discarded when they occurred and were replaced by other randomly sampled values. This is a clear limitation of the present approach. The simulation method needs to be refined in the future, especially in the management of the variance, in order to avoid the placement of arbitrary bounds for PK parameters and exposures. Nevertheless, the mean PK parameter values obtained for the 10,000 simulated subjects (Table 3) were reasonably close to those obtained by use of the NPAG algorithm. In addition, plasma drug exposures (C_max and AUC_0-24) after the administration of a single 600-mg or 1,200-mg oral dose of RIF were in agreement with published results (5, 22, 24, 29). The differences found between the simulation results and the results obtained with the NPAG algorithm for some of the PK parameter medians and standard deviations can be explained by the parametric assumption in the random sampling of individual PK parameters from the Gaussian multivariate distributions and by the use of the overall population covariance matrix around each population support point.

The principal aim of the simulation was to study the ability of the standard 600-mg dose of RIF to ensure a sufficient bactericidal effect and prevent the development of resistance by M. tuberculosis under pulmonary tissue conditions. Our simulation results indicate that with the 600-mg oral dose, the pulmonary concentrations of RIF are probably too low in most patients to prevent the emergence of RIF resistance, especially in the extracellular fraction (ELF). Even at a low M. tuberculosis MIC (0.01 mg/liter), the probability of target attainment for the prevention of RIF resistance was less than 36% in ELF. Intracellular concentrations appeared to be more effective for the prevention of resistance at low MICs, but the level of target attainment decreased to less than 50% when the MIC was ≥0.05 mg/liter. The level of target attainment in plasma was slightly better than that in ELF, but this may be less relevant in patients with pulmonary TB. The use of the higher 1,200-mg RIF dose in the simulations resulted in more subjects achieving the C_max/MIC target value. For example, in ELF, this proportion was increased to approximately 60% at a MIC of 0.01 mg/liter. Nevertheless, for higher MICs, the target attainment results were still clearly suboptimal for all concentrations in all three compartments.

Our simulation showed the inability of the 600-mg standard dose of RIF to prevent the emergence of resistance. Because of the concentration-dependent character of the suppression of RIF resistance shown by Gumbo et al. (8), one should expect that a higher dose of RIF might be beneficial for the prevention of resistance. However, our results suggest that the prevention of M. tuberculosis resistance to RIF by the drug itself is probably impossible to achieve in all patients, unless a much higher dose is used, but the ability of patients to tolerate that dose is not yet known.

The second effect evaluated in the simulation was the killing effect of RIF against M. tuberculosis, which has been shown to be correlated with the AUC_0-24/MIC ratio (8, 9). Our results showed large difference in the expected efficacy of RIF between extracellular and intracellular locations. Plasma concentrations, which were supposed to drive the effect against extracellular M. tuberculosis in the simulation (see the work of Jayaram et al. [9]), were effective in a very large proportion of subjects. After the administration of a single 600-mg oral dose of RIF, the rates of target attainment (AUC_0-24/MIC ≥ 271) ranged from 78 to 100% for MICs between 0.01 and 0.1 mg/liter. However, the probabilities of target attainment in ELF were much lower. In the simulation, ELF concentrations were considered equivalent to extracellular concentrations, which were correlated with the bactericidal effect of RIF against intracellular M. tuberculosis in a previous study (9). At the lowest MIC, the probability of target attainment was only 54.5% in ELF, and it was less than 20% when the MIC was ≥0.05 mg/liter.

Again, as expected, the use of the 1,200-mg dose had a favorable effect on the target attainment results, especially in ELF, while the concentrations achieved in ELF were still much less effective than the concentrations achieved in plasma.

Only single-dose simulations were done in this study. One could expect that multiple doses might result in greater and, consequently, more effective RIF concentrations. However, because of its short half-life (about 4 h) in plasma (24) and because of autoinduction, RIF does not accumulate with the administration of multiple doses in a once-daily regimen. Actually, the RIF concentrations on the first day of therapy are often higher than those in the subsequent steady state.

The simulation work performed was based on experimental published results and also on several assumptions that should be considered in the interpretation of the results.

The target values of both the C_max/MIC and the AUC_0-24/MIC ratios were determined under specific experimental conditions. The target value for the prevention of RIF resistance was established by Gumbo et al., who used a hollow-fiber system which directly exposed the bacterial suspensions to changing concentrations of RIF (8). In our simulations, we assumed that the same relationship takes place between the drug and the M. tuberculosis bacilli in plasma, ELF, and ACs. The reference study was performed with a strain of M. tuberculosis of low virulence and with a low MIC. Therefore, the C_max/MIC target value required to suppress RIF resistance may still be more difficult to attain in most clinical settings in patients infected with more virulent strains (8).

The two target values for the AUC_0-24/MIC ratio were determined by Jayaram et al. (9). The AUC_0-24/MIC target value of 271 was calculated for plasma concentrations in mice. It was associated with a 1-log₁₀ CFU reduction in the total lung bacterial load after 6 days of treatment. The AUC_0-24/MIC target value of 665 was the drug exposure in a macrophage culture medium associated with a 1-log₁₀ CFU reduction in the intracellular bacterial counts after 4 days of treatment. In our simulations, those two values have been extrapolated and used as target values for the killing effect of the plasma and the ELF concentrations. One should keep in mind that those target values were not all or none but, rather, were quantitative end-points of efficacy.

The effect associated with the plasma AUC_0-24/MIC target value, i.e., a mean 0.167-log₁₀ CFU reduction per day over the first 6 days of treatment is relevant. It is comparable to the early bactericidal activity of RIF reported in the literature. Early bactericidal activity, which is generally measured over the first 2 or 5 days of treatment, is the mean daily log₁₀ CFU reduction in patient sputum cultures. In studies done with patients with pulmonary TB receiving standard doses of RIF, early bactericidal activity values ranged from 0.17 ± 0.16 to 0.53 ± 0.46 log₁₀ CFU (13, 20, 32).

There is no clinical reference for the bactericidal effect of RIF, i.e., a mean 0.25-log₁₀ CFU reduction per day over the first 4 days of treatment, associated with the target AUC_0-24/MIC of 665 defined for intracellular conditions. However, this seems to be a fairly strong bactericidal effect, since it is close to the maximum effect (0.38-log₁₀ CFU reduction) found in the reference study (9). This means that exposures below the target value may be quantitatively less bactericidal than the target one but still reasonably effective.

Gumbo et al. (8) and Jayaram et al. (9) both performed short-term studies with replicative bacilli. The PK-PD relationships observed in those bacteria may not apply to the slowly replicative persistent or dormant subpopulations of microorganisms. There is a need for PK-PD models that predict the sterilizing activities of anti-TB drugs.

The overall results of this study strongly suggest that the efficacy of the current dosage regimen of RIF is suboptimal in most patients. A higher dose of RIF may well be needed to treat patients with TB more effectively. Of course, this would be feasible only if patients can tolerate such higher doses of RIF. Further clinical studies are necessary to evaluate this problem. However, while intermittent high doses of RIF have been associated with hypersensitivity reactions (in particular, flu-like syndrome), the literature suggests that RIF hepatotoxicity is idiosyncratic (31). High daily doses of RIF have been used for the treatment of nonmycobacterial infections, such as brucellosis, with overall good tolerance (33).

The various limitations described above may limit the generalizability of our results. However, our conclusions are in agreement with those from other studies. The suboptimal efficacy of the current standard doses of RIF has been reported many times. The need to evaluate higher doses has been described by other scientists for years (5, 8, 9, 20, 21, 23). The demonstration of the concentration-dependent character of the effect of RIF, as well as the definition of experimental effective target values for RIF exposure, has been an important breakthrough in knowledge of the PK-PD of RIF in recent years. The existence of target exposure values for both the bactericidal effect and the suppression of resistance has yet to be confirmed in the clinical setting. However, one can start to think about possible ways to hit such potential targets in routine patient care. As illustrated by the simulation in this work, because of PK-PD variability, there is no way to achieve any therapeutic target exactly in all patients by using a standard dose of RIF, whatever it might be. The use of a higher dose of RIF should achieve the therapeutic target in more subjects but would expose more patients to high and potentially toxic concentrations of RIF. To achieve a therapeutic target most precisely in each patient, individualized dosage regimens and adaptive control of drug therapy are required. Specific tools for this have been developed (10, 12). While new drugs appear to be necessary to enhance anti-TB drug therapy, the better use of current drugs should also be considered. Dosage individualization and maximally precise attainment of the desired therapeutic target may help in this way (11).

Acknowledgments

This work was supported in part by NIH grant EB005803.

We have no conflicts of interest that are relevant to the content of this study.

Footnotes

^▿

Published ahead of print on 20 April 2009.

REFERENCES

1.Centers for Disease Control and Prevention. 2006. Emergence of Mycobacterium tuberculosis with extensive resistance to second-line drugs. MMWR Morb. Mortal. Wkly. Rep. 55:301-305. [PubMed] [Google Scholar]
2.Choudhri, S. H., M. Hawken, S. Gathua, G. O. Minyiri, W. Watkins, J. Sahai, D. S. Sitar, F. Y. Aoki, and R. Long. 1997. Pharmacokinetics of antimycobacterial drugs in patients with tuberculosis, AIDS, and diarrhea. Clin. Infect. Dis. 25:104-111. [DOI] [PubMed] [Google Scholar]
3.Conte, J. E., J. A. Golden, J. E. Kipps, E. T. Lin, and E. Zurlinden. 2004. Effect of sex and AIDS status on the plasma and intrapulmonary pharmacokinetics of rifampicin. Clin. Pharmacokinet. 43:395-404. [DOI] [PubMed] [Google Scholar]
4.Conte, J. E., Jr., E. Lin, and E. Zurlinden. 2000. Liquid chromatographic determination of rifampin in human plasma, bronchoalveolar lavage fluid, and alveolar cells. J. Chromatogr. Sci. 38:72-76. [DOI] [PubMed] [Google Scholar]
5.Diacon, A. H., R. F. Patientia, A. Venter, P. D. van Helden, P. J. Smith, H. McIlleron, J. S. Maritz, and P. R. Donald. 2007. Early bactericidal activity of high-dose rifampin in patients with pulmonary tuberculosis evidenced by positive sputum smears. Antimicrob. Agents Chemother. 51:2994-2996. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Gandhi, N. R., A. Moll, A. W. Sturm, R. Pawinski, T. Govender, U. Lalloo, K. Zeller, J. Andrews, and G. Friedland. 2006. Extensively drug-resistant tuberculosis as a cause of death in patients co-infected with tuberculosis and HIV in a rural area of South Africa. Lancet 368:1575-1580. [DOI] [PubMed] [Google Scholar]
7.Gilman, A., L. S. Goodman, T. W. Rall, and F. Murad. 1985. Goodman and Gilman's the pharmacological basis of therapeutics, 7th ed. Macmillan Publishing Company, New York, NY.
8.Gumbo, T., A. Louie, M. R. Deziel, W. Liu, L. M. Parsons, M. Salfinger, and G. L. Drusano. 2007. Concentration-dependent Mycobacterium tuberculosis killing and prevention of resistance by rifampin. Antimicrob. Agents Chemother. 51:3781-3788. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Jayaram, R., S. Gaonkar, P. Kaur, B. L. Suresh, B. N. Mahesh, R. Jayashree, V. Nandi, S. Bharat, R. K. Shandil, E. Kantharaj, and V. Balasubramanian. 2003. Pharmacokinetics-pharmacodynamics of rifampin in an aerosol infection model of tuberculosis. Antimicrob. Agents Chemother. 47:2118-2124. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Jelliffe, R. 2000. Goal-oriented, model-based drug regimens: setting individualized goals for each patient. Ther. Drug Monit. 22:325-329. [DOI] [PubMed] [Google Scholar]
11.Jelliffe, R., D. Bayard, M. Milman, M. Van Guilder, and A. Schumitzky. 2000. Achieving target goals most precisely using nonparametric compartmental models and “multiple model” design of dosage regimens. Ther. Drug Monit. 22:346-353. [DOI] [PubMed] [Google Scholar]
12.Jelliffe, R. W., A. Schumitzky, D. Bayard, M. Milman, M. Van Guilder, X. Wang, F. Jiang, X. Barbaut, and P. Maire. 1998. Model-based, goal-oriented, individualised drug therapy. Linkage of population modelling, new ‘multiple model’ dosage design, Bayesian feedback and individualised target goals. Clin. Pharmacokinet. 34:57-77. [DOI] [PubMed] [Google Scholar]
13.Jindani, A., C. J. Dore, and D. A. Mitchison. 2003. Bactericidal and sterilizing activities of antituberculosis drugs during the first 14 days. Am. J. Respir. Crit. Care Med. 167:1348-1354. [DOI] [PubMed] [Google Scholar]
14.Kiem, S., and J. J. Schentag. 2008. Interpretation of antibiotic concentration ratios measured in epithelial lining fluid. Antimicrob. Agents Chemother. 52:24-36. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Kimerling, M. E., P. Phillips, P. Patterson, M. Hall, C. A. Robinson, and N. E. Dunlap. 1998. Low serum antimycobacterial drug levels in non-HIV-infected tuberculosis patients. Chest 113:1178-1183. [DOI] [PubMed] [Google Scholar]
16.Leary, R., R. Jelliffe, A. Schumitzky, and M. Van Guilder. 2001. Abstr. 14th IEEE Symposium on Computer-Based Medical Systems, p. 389-394, Bethesda, MD, 26 to 27 July 2001.
17.Lodise, T. P., M. Gotfried, S. Barriere, and G. L. Drusano. 2008. Telavancin penetration into human epithelial lining fluid as determined by population pharmacokinetic modeling and Monte Carlo simulation. Antimicrob. Agents Chemother. 52:2300-2304. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Lodise, T. P., M. Kinzig-Schippers, G. L. Drusano, U. Loos, F. Vogel, J. Bulitta, M. Hinder, and F. Sorgel. 2008. Use of population pharmacokinetic modeling and Monte Carlo simulation to describe the pharmacodynamic profile of cefditoren in plasma and epithelial lining fluid. Antimicrob. Agents Chemother. 52:1945-1951. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Mehta, J. B., H. Shantaveerapa, R. P. Byrd, Jr., S. E. Morton, F. Fountain, and T. M. Roy. 2001. Utility of rifampin blood levels in the treatment and follow-up of active pulmonary tuberculosis in patients who were slow to respond to routine directly observed therapy. Chest 120:1520-1524. [DOI] [PubMed] [Google Scholar]
20.Mitchison, D. A. 2000. Role of individual drugs in the chemotherapy of tuberculosis. Int. J. Tuberc. Lung Dis. 4:796-806. [PubMed] [Google Scholar]
21.Nuermberger, E., and J. Grosset. 2004. Pharmacokinetic and pharmacodynamic issues in the treatment of mycobacterial infections. Eur. J. Clin. Microbiol. Infect. Dis. 23:243-255. [DOI] [PubMed] [Google Scholar]
22.Pargal, A., and S. Rani. 2001. Non-linear pharmacokinetics of rifampicin in healthy Asian Indian volunteers. Int. J. Tuberc. Lung Dis. 5:70-79. [PubMed] [Google Scholar]
23.Peloquin, C. 2003. What is the ‘right’ dose of rifampin? Int. J. Tuberc. Lung Dis. 7:3-5. [PubMed] [Google Scholar]
24.Peloquin, C. A., G. S. Jaresko, C. L. Yong, A. C. Keung, A. E. Bulpitt, and R. W. Jelliffe. 1997. Population pharmacokinetic modeling of isoniazid, rifampin, and pyrazinamide. Antimicrob. Agents Chemother. 41:2670-2679. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Peloquin, C. A., A. T. Nitta, W. J. Burman, K. F. Brudney, J. R. Miranda-Massari, M. E. McGuinness, S. E. Berning, and G. T. Gerena. 1996. Low antituberculosis drug concentrations in patients with AIDS. Ann. Pharmacother. 30:919-925. [DOI] [PubMed] [Google Scholar]
26.Perlman, D. C., Y. Segal, S. Rosenkranz, P. M. Rainey, R. P. Remmel, N. Salomon, R. Hafner, and C. A. Peloquin. 2005. The clinical pharmacokinetics of rifampin and ethambutol in HIV-infected persons with tuberculosis. Clin. Infect. Dis. 41:1638-1647. [DOI] [PubMed] [Google Scholar]
27.Rennard, S. I., G. Basset, D. Lecossier, K. M. O'Donnell, P. Pinkston, P. G. Martin, and R. G. Crystal. 1986. Estimation of volume of epithelial lining fluid recovered by lavage using urea as a marker of dilution. J. Appl. Physiol. 60:532-538. [DOI] [PubMed] [Google Scholar]
28.Rom, W. N., and S. M. Garay. 2004. Tuberculosis, 2nd ed. Lippincott Williams & Wilkins, Philadelphia, PA.
29.Ruslami, R., H. M. Nijland, B. Alisjahbana, I. Parwati, R. van Crevel, and R. E. Aarnoutse. 2007. Pharmacokinetics and tolerability of a higher rifampin dose versus the standard dose in pulmonary tuberculosis patients. Antimicrob. Agents Chemother. 51:2546-2551. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Sahai, J., K. Gallicano, L. Swick, S. Tailor, G. Garber, I. Seguin, L. Oliveras, S. Walker, A. Rachlis, and D. W. Cameron. 1997. Reduced plasma concentrations of antituberculosis drugs in patients with HIV infection. Ann. Intern. Med. 127:289-293. [DOI] [PubMed] [Google Scholar]
31.Saukkonen, J. J., D. L. Cohn, R. M. Jasmer, S. Schenker, J. A. Jereb, C. M. Nolan, C. A. Peloquin, F. M. Gordin, D. Nunes, D. B. Strader, J. Bernardo, R. Venkataramanan, and T. R. Sterling. 2006. An official ATS statement: hepatotoxicity of antituberculosis therapy. Am. J. Respir. Crit. Care Med. 174:935-952. [DOI] [PubMed] [Google Scholar]
32.Sirgel, F. A., P. B. Fourie, P. R. Donald, N. Padayatchi, R. Rustomjee, J. Levin, G. Roscigno, J. Norman, H. McIlleron, and D. A. Mitchison. 2005. The early bactericidal activities of rifampin and rifapentine in pulmonary tuberculosis. Am. J. Respir. Crit. Care Med. 172:128-135. [DOI] [PubMed] [Google Scholar]
33.Solera, J., M. Rodriguez-Zapata, P. Geijo, J. Largo, J. Paulino, L. Saez, E. Martinez-Alfaro, L. Sanchez, M. A. Sepulveda, M. D. Ruiz-Ribo, et al. 1995. Doxycycline-rifampin versus doxycycline-streptomycin in treatment of human brucellosis due to Brucella melitensis. Antimicrob. Agents Chemother. 39:2061-2067. [DOI] [PMC free article] [PubMed] [Google Scholar]
34.van Crevel, R., B. Alisjahbana, W. C. de Lange, F. Borst, H. Danusantoso, J. W. van der Meer, D. Burger, and R. H. Nelwan. 2002. Low plasma concentrations of rifampicin in tuberculosis patients in Indonesia. Int. J. Tuberc. Lung Dis. 6:497-502. [DOI] [PubMed] [Google Scholar]
35.Weiner, M., W. Burman, A. Vernon, D. Benator, C. A. Peloquin, A. Khan, S. Weis, B. King, N. Shah, T. Hodge, and the Tuberculosis Trials Consortium. 2003. Low isoniazid concentrations and outcome of tuberculosis treatment with once-weekly isoniazid and rifapentine. Am. J. Respir. Crit. Care Med. 167:1341-1347. [DOI] [PubMed] [Google Scholar]
36.Wilkins, J. J., R. M. Savic, M. O. Karlsson, G. Langdon, H. McIlleron, G. Pillai, P. J. Smith, and U. S. Simonsson. 2008. Population pharmacokinetics of rifampin in pulmonary tuberculosis patients, including a semimechanistic model to describe variable absorption. Antimicrob. Agents Chemother. 52:2138-2148. [DOI] [PMC free article] [PubMed] [Google Scholar]
37.World Health Organization. 2009. Global tuberculosis control: epidemiology, strategy, financing. Report WHO/HTM/TB/2009.411. World Health Organization, Geneva, Switzerland. http://www.who.int/tb/publications/en/index.html.
38.World Health Organization. 2003. Treatment of tuberculosis: guidelines for national programmes. Report WHO/CDS/TB/2003.313. World Health Organization, Geneva, Switzerland. http://www.who.int/tb/publications/en/index.html.

[r1] 1.Centers for Disease Control and Prevention. 2006. Emergence of Mycobacterium tuberculosis with extensive resistance to second-line drugs. MMWR Morb. Mortal. Wkly. Rep. 55:301-305. [PubMed] [Google Scholar]

[r2] 2.Choudhri, S. H., M. Hawken, S. Gathua, G. O. Minyiri, W. Watkins, J. Sahai, D. S. Sitar, F. Y. Aoki, and R. Long. 1997. Pharmacokinetics of antimycobacterial drugs in patients with tuberculosis, AIDS, and diarrhea. Clin. Infect. Dis. 25:104-111. [DOI] [PubMed] [Google Scholar]

[r3] 3.Conte, J. E., J. A. Golden, J. E. Kipps, E. T. Lin, and E. Zurlinden. 2004. Effect of sex and AIDS status on the plasma and intrapulmonary pharmacokinetics of rifampicin. Clin. Pharmacokinet. 43:395-404. [DOI] [PubMed] [Google Scholar]

[r4] 4.Conte, J. E., Jr., E. Lin, and E. Zurlinden. 2000. Liquid chromatographic determination of rifampin in human plasma, bronchoalveolar lavage fluid, and alveolar cells. J. Chromatogr. Sci. 38:72-76. [DOI] [PubMed] [Google Scholar]

[r5] 5.Diacon, A. H., R. F. Patientia, A. Venter, P. D. van Helden, P. J. Smith, H. McIlleron, J. S. Maritz, and P. R. Donald. 2007. Early bactericidal activity of high-dose rifampin in patients with pulmonary tuberculosis evidenced by positive sputum smears. Antimicrob. Agents Chemother. 51:2994-2996. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r6] 6.Gandhi, N. R., A. Moll, A. W. Sturm, R. Pawinski, T. Govender, U. Lalloo, K. Zeller, J. Andrews, and G. Friedland. 2006. Extensively drug-resistant tuberculosis as a cause of death in patients co-infected with tuberculosis and HIV in a rural area of South Africa. Lancet 368:1575-1580. [DOI] [PubMed] [Google Scholar]

[r7] 7.Gilman, A., L. S. Goodman, T. W. Rall, and F. Murad. 1985. Goodman and Gilman's the pharmacological basis of therapeutics, 7th ed. Macmillan Publishing Company, New York, NY.

[r8] 8.Gumbo, T., A. Louie, M. R. Deziel, W. Liu, L. M. Parsons, M. Salfinger, and G. L. Drusano. 2007. Concentration-dependent Mycobacterium tuberculosis killing and prevention of resistance by rifampin. Antimicrob. Agents Chemother. 51:3781-3788. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r9] 9.Jayaram, R., S. Gaonkar, P. Kaur, B. L. Suresh, B. N. Mahesh, R. Jayashree, V. Nandi, S. Bharat, R. K. Shandil, E. Kantharaj, and V. Balasubramanian. 2003. Pharmacokinetics-pharmacodynamics of rifampin in an aerosol infection model of tuberculosis. Antimicrob. Agents Chemother. 47:2118-2124. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r10] 10.Jelliffe, R. 2000. Goal-oriented, model-based drug regimens: setting individualized goals for each patient. Ther. Drug Monit. 22:325-329. [DOI] [PubMed] [Google Scholar]

[r11] 11.Jelliffe, R., D. Bayard, M. Milman, M. Van Guilder, and A. Schumitzky. 2000. Achieving target goals most precisely using nonparametric compartmental models and “multiple model” design of dosage regimens. Ther. Drug Monit. 22:346-353. [DOI] [PubMed] [Google Scholar]

[r12] 12.Jelliffe, R. W., A. Schumitzky, D. Bayard, M. Milman, M. Van Guilder, X. Wang, F. Jiang, X. Barbaut, and P. Maire. 1998. Model-based, goal-oriented, individualised drug therapy. Linkage of population modelling, new ‘multiple model’ dosage design, Bayesian feedback and individualised target goals. Clin. Pharmacokinet. 34:57-77. [DOI] [PubMed] [Google Scholar]

[r13] 13.Jindani, A., C. J. Dore, and D. A. Mitchison. 2003. Bactericidal and sterilizing activities of antituberculosis drugs during the first 14 days. Am. J. Respir. Crit. Care Med. 167:1348-1354. [DOI] [PubMed] [Google Scholar]

[r14] 14.Kiem, S., and J. J. Schentag. 2008. Interpretation of antibiotic concentration ratios measured in epithelial lining fluid. Antimicrob. Agents Chemother. 52:24-36. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r15] 15.Kimerling, M. E., P. Phillips, P. Patterson, M. Hall, C. A. Robinson, and N. E. Dunlap. 1998. Low serum antimycobacterial drug levels in non-HIV-infected tuberculosis patients. Chest 113:1178-1183. [DOI] [PubMed] [Google Scholar]

[r16] 16.Leary, R., R. Jelliffe, A. Schumitzky, and M. Van Guilder. 2001. Abstr. 14th IEEE Symposium on Computer-Based Medical Systems, p. 389-394, Bethesda, MD, 26 to 27 July 2001.

[r17] 17.Lodise, T. P., M. Gotfried, S. Barriere, and G. L. Drusano. 2008. Telavancin penetration into human epithelial lining fluid as determined by population pharmacokinetic modeling and Monte Carlo simulation. Antimicrob. Agents Chemother. 52:2300-2304. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r18] 18.Lodise, T. P., M. Kinzig-Schippers, G. L. Drusano, U. Loos, F. Vogel, J. Bulitta, M. Hinder, and F. Sorgel. 2008. Use of population pharmacokinetic modeling and Monte Carlo simulation to describe the pharmacodynamic profile of cefditoren in plasma and epithelial lining fluid. Antimicrob. Agents Chemother. 52:1945-1951. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r19] 19.Mehta, J. B., H. Shantaveerapa, R. P. Byrd, Jr., S. E. Morton, F. Fountain, and T. M. Roy. 2001. Utility of rifampin blood levels in the treatment and follow-up of active pulmonary tuberculosis in patients who were slow to respond to routine directly observed therapy. Chest 120:1520-1524. [DOI] [PubMed] [Google Scholar]

[r20] 20.Mitchison, D. A. 2000. Role of individual drugs in the chemotherapy of tuberculosis. Int. J. Tuberc. Lung Dis. 4:796-806. [PubMed] [Google Scholar]

[r21] 21.Nuermberger, E., and J. Grosset. 2004. Pharmacokinetic and pharmacodynamic issues in the treatment of mycobacterial infections. Eur. J. Clin. Microbiol. Infect. Dis. 23:243-255. [DOI] [PubMed] [Google Scholar]

[r22] 22.Pargal, A., and S. Rani. 2001. Non-linear pharmacokinetics of rifampicin in healthy Asian Indian volunteers. Int. J. Tuberc. Lung Dis. 5:70-79. [PubMed] [Google Scholar]

[r23] 23.Peloquin, C. 2003. What is the ‘right’ dose of rifampin? Int. J. Tuberc. Lung Dis. 7:3-5. [PubMed] [Google Scholar]

[r24] 24.Peloquin, C. A., G. S. Jaresko, C. L. Yong, A. C. Keung, A. E. Bulpitt, and R. W. Jelliffe. 1997. Population pharmacokinetic modeling of isoniazid, rifampin, and pyrazinamide. Antimicrob. Agents Chemother. 41:2670-2679. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r25] 25.Peloquin, C. A., A. T. Nitta, W. J. Burman, K. F. Brudney, J. R. Miranda-Massari, M. E. McGuinness, S. E. Berning, and G. T. Gerena. 1996. Low antituberculosis drug concentrations in patients with AIDS. Ann. Pharmacother. 30:919-925. [DOI] [PubMed] [Google Scholar]

[r26] 26.Perlman, D. C., Y. Segal, S. Rosenkranz, P. M. Rainey, R. P. Remmel, N. Salomon, R. Hafner, and C. A. Peloquin. 2005. The clinical pharmacokinetics of rifampin and ethambutol in HIV-infected persons with tuberculosis. Clin. Infect. Dis. 41:1638-1647. [DOI] [PubMed] [Google Scholar]

[r27] 27.Rennard, S. I., G. Basset, D. Lecossier, K. M. O'Donnell, P. Pinkston, P. G. Martin, and R. G. Crystal. 1986. Estimation of volume of epithelial lining fluid recovered by lavage using urea as a marker of dilution. J. Appl. Physiol. 60:532-538. [DOI] [PubMed] [Google Scholar]

[r28] 28.Rom, W. N., and S. M. Garay. 2004. Tuberculosis, 2nd ed. Lippincott Williams & Wilkins, Philadelphia, PA.

[r29] 29.Ruslami, R., H. M. Nijland, B. Alisjahbana, I. Parwati, R. van Crevel, and R. E. Aarnoutse. 2007. Pharmacokinetics and tolerability of a higher rifampin dose versus the standard dose in pulmonary tuberculosis patients. Antimicrob. Agents Chemother. 51:2546-2551. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r30] 30.Sahai, J., K. Gallicano, L. Swick, S. Tailor, G. Garber, I. Seguin, L. Oliveras, S. Walker, A. Rachlis, and D. W. Cameron. 1997. Reduced plasma concentrations of antituberculosis drugs in patients with HIV infection. Ann. Intern. Med. 127:289-293. [DOI] [PubMed] [Google Scholar]

[r31] 31.Saukkonen, J. J., D. L. Cohn, R. M. Jasmer, S. Schenker, J. A. Jereb, C. M. Nolan, C. A. Peloquin, F. M. Gordin, D. Nunes, D. B. Strader, J. Bernardo, R. Venkataramanan, and T. R. Sterling. 2006. An official ATS statement: hepatotoxicity of antituberculosis therapy. Am. J. Respir. Crit. Care Med. 174:935-952. [DOI] [PubMed] [Google Scholar]

[r32] 32.Sirgel, F. A., P. B. Fourie, P. R. Donald, N. Padayatchi, R. Rustomjee, J. Levin, G. Roscigno, J. Norman, H. McIlleron, and D. A. Mitchison. 2005. The early bactericidal activities of rifampin and rifapentine in pulmonary tuberculosis. Am. J. Respir. Crit. Care Med. 172:128-135. [DOI] [PubMed] [Google Scholar]

[r33] 33.Solera, J., M. Rodriguez-Zapata, P. Geijo, J. Largo, J. Paulino, L. Saez, E. Martinez-Alfaro, L. Sanchez, M. A. Sepulveda, M. D. Ruiz-Ribo, et al. 1995. Doxycycline-rifampin versus doxycycline-streptomycin in treatment of human brucellosis due to Brucella melitensis. Antimicrob. Agents Chemother. 39:2061-2067. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r34] 34.van Crevel, R., B. Alisjahbana, W. C. de Lange, F. Borst, H. Danusantoso, J. W. van der Meer, D. Burger, and R. H. Nelwan. 2002. Low plasma concentrations of rifampicin in tuberculosis patients in Indonesia. Int. J. Tuberc. Lung Dis. 6:497-502. [DOI] [PubMed] [Google Scholar]

[r35] 35.Weiner, M., W. Burman, A. Vernon, D. Benator, C. A. Peloquin, A. Khan, S. Weis, B. King, N. Shah, T. Hodge, and the Tuberculosis Trials Consortium. 2003. Low isoniazid concentrations and outcome of tuberculosis treatment with once-weekly isoniazid and rifapentine. Am. J. Respir. Crit. Care Med. 167:1341-1347. [DOI] [PubMed] [Google Scholar]

[r36] 36.Wilkins, J. J., R. M. Savic, M. O. Karlsson, G. Langdon, H. McIlleron, G. Pillai, P. J. Smith, and U. S. Simonsson. 2008. Population pharmacokinetics of rifampin in pulmonary tuberculosis patients, including a semimechanistic model to describe variable absorption. Antimicrob. Agents Chemother. 52:2138-2148. [DOI] [PMC free article] [PubMed] [Google Scholar]

[r37] 37.World Health Organization. 2009. Global tuberculosis control: epidemiology, strategy, financing. Report WHO/HTM/TB/2009.411. World Health Organization, Geneva, Switzerland. http://www.who.int/tb/publications/en/index.html.

[r38] 38.World Health Organization. 2003. Treatment of tuberculosis: guidelines for national programmes. Report WHO/CDS/TB/2003.313. World Health Organization, Geneva, Switzerland. http://www.who.int/tb/publications/en/index.html.

PERMALINK

Population Modeling and Monte Carlo Simulation Study of the Pharmacokinetics and Antituberculosis Pharmacodynamics of Rifampin in Lungs^▿

Sylvain Goutelle

Laurent Bourguignon

Pascal H Maire

Michael Van Guilder

John E Conte Jr

Roger W Jelliffe

Abstract

MATERIALS AND METHODS