Metronidazole and hydroxy-metronidazole central nervous system distribution: な 2. Cerebrospinal fluid concentrations measurements in patients with external に ventricular drain ぬ Denis Frasca
1,2,3, Claire Dahyot-Fizelier
1,2,3 , Christophe Adier
1,4, Olivier Mimoz
1,2,3, Bertrand ね
Debaene1,2,3
, William Couet1,2,4
, Sandrine Marchand1,2,4#
の は 1Inserm U1070, Pole Biologie Santé, 1 rue Georges Bonnet, 86000 Poitiers, France ば
2Université de Poitiers, UFR Médecine-Pharmacie, 6 rue de la Milétrie, 86000 Poitiers, ぱ
France ひ 3 CHU Poitiers, Service d’Anesthésiologie-Réanimation Chirurgicale, 2 rue de la Milétrie, など
86000 Poitiers, France なな 4 CHU Poitiers, Service de Toxicologie-Pharmacocinétique, 2 rue de la Milétrie, 86000 なに
Abstract にに This study explored CSF metronidazole and hydroxy-metronidazole distribution in brain-にぬ injured patients. にね Four brain-injured patients with external ventricular drain received 500 mg of metronidazole にの over 0.5 h every 8h. CSF and blood samples were collected at steady-state over 8 h and には metronidazole and hydroxy-metronidazole concentrations were assayed by HPLC. A non-にば compartmental analysis was performed. にぱ Metronidazole distributes extensively within CSF with a mean CSF over unbound plasma にひ AUC0-τ ratio equal to 86 ± 16 %. Yet concentrations profiles in CSF were mostly flat ぬど compared with plasma profiles. Hydroxy-metronidazole concentrations were much lower than ぬな those of metronidazole both in plasma and CSF with a corresponding CSF over unbound ぬに plasma AUC0-τ ratio equal to 79 ± 16 %. ぬぬ This study was the first one to describe in details the pharmacokinetics of metronidazole and ぬね hydroxy -metronidazole in CSF. ぬの ぬは ぬば
Introduction ぬひ Blood brain barrier (BBB) and blood-CSF barrier (BCSFB) protect brain from neurotoxic ねど substances and control drugs distribution. CNS distribution studies are essential to predict ねな drug’s efficacy or/and side effects. Antibiotics may be used for the treatment of CNS ねに infections but may also induce undesirable excitatory side effects. CNS drugs distribution ねぬ has for long been exclusively assessed from cerebrospinal fluid (CSFょ concentrations after ねね lumbar puncture or external ventricular drainage┻ (owever these studies present a ねの number of limits that have recently been reviewed ゅなょ┻ More recently microdialysis was ねは developed to measure brain extracellular fluid (ECF) concentrations of drugs as a better way ねば to characterize CNS distribution (2). Yet brain microdialysis studies in human are relatively ねぱ complex to handle and limited data have been published (3, 4). Interestingly recent ねひ experimental studies have shown differences between CSF and ECF concentrations for のど acetaminophen and quinidine in rats (5, 6). In fact, despite similarities, BBB and BSCFB のな present structural differences (2), that may lead to distinct brain ECF and CSF concentrations のに versus time profiles. Comparisons between ECF and CSF concentrations have been reported のぬ in animals (1, 5, 6) and to our knowledge in one previous study on carbamazepine in humans のね (7). Metronidazole, a nitroimidazole antibiotic, is indicated in cerebromeningeal infections (8) のの due to anaerobic bacteria but it is also known to induce peripheral and central side effects (9-のは 12). Our goal was to perform a CSF distribution study in patients and to compare CSF のば concentrations with ECF concentrations previously determined by brain microdialysis (13), のぱ using metronidazole and OH-metronidazole, an active metabolite of metronidazole, as a のひ representative antibiotics with extensive CNS distribution はど はな はに
Material and methods はぬ Patients はね This study was performed at University Hospital of Poitiers (France) and was approved by the はの local ethics committee (CPP OUEST III, protocol N°2008-003311-12). Written informed はは consent was obtained from a legal representative of each patient. Four brain-injured male はば patients (3 with subarachnoid haemorrhage and 1 with ventricular haemorrhage) were sedated はぱ with midazolam and fentanyl and were mechanically ventilated. The demographic はひ characteristics of patients were detailed in Table 1. Patients have undergone external ばど ventricular drainage (EVD) (EDS 3™ CSF External Drainage System, CODMAN & ばな SHURTLEFF Inc. Raynham, USA) due to non-inflammatory occlusive hydrocephalus. ばに ばぬ Drug administration and sampling ばね Patients received metronidazole (B BRAUN, Boulogne Billancourt, France) and cefotaxime ばの (PANPHARMA, Boulogne Billancourt, France) to treat lung infection at respective dosing ばは regimens of 500mg three times daily (tid) and 2g tid. Metronidazole CSF pharmacokinetic ばば study was conducted at steady state after 6 to 16 metronidazole administrations corresponding ばぱ to 2 to 5 days after starting treatment. On day of experiment, a 30 min intravenous infusion ばひ containing 500 mg of metronidazole was started after collection of residual CSF and blood ぱど samples. Eight to twelve CSF samples and seven to ten blood samples were collected over 8 ぱな h. Blood samples were centrifuged and plasma was collected. Two extra plasma samples were ぱに collected, one early and one later post dosing, to determine metronidazole and OH-ぱぬ metronidazole unbound concentrations by ultrafiltration at 2500 g for 15 min at 4 °C ぱね (Centrifree
®, MILLIPORE Corporation, Billerica, USA). The unbound fractions were ぱの
determined for each compound and each patient as the mean value of the unbound to total ぱは concentrations ratios estimated in these two extra plasma samples. ぱば
ぱぱ Metronidazole and OH-metronidazole assay ぱひ Metronidazole and OH-metronidazole concentrations were determined by high performance ひど liquid performance (HPLC) with UV detection (310nm) as previously described (13). Briefly, ひな the mobile phase was a solution of KH2PO4 0.01M mixed with acetonitrile (86/14, vol/vol) at ひに a flow rate of 0.4mL/min. Brain dialysates and UF samples were injected directly onto an ひぬ Xterra C18 column after dilution with an internal standard solution of dimetridazole. The ひね dialysates and UF metronidazole and OH-metronidazole intra and inter days variability were ひの respectively characterized at four (0.25, 0.5, 2 and 20 µg.mL
-1) and three (0.5, 2 and 15 ひは
µg.mL-1
) levels of concentrations and were always lower than 15 %. Plasma samples were ひば precipitated by acetonitrile containing internal standard. The plasma intra and inter days ひぱ variability were respectively characterized at four (1, 2, 10 and 40 µg.mL
-1) and three (1, 5 ひひ
and 30 µg.mL-1
) levels of concentrations and were always lower than 15 %. などど などな Non compartmental analysis などに Unbound plasma concentrations were used for pharmacokinetic analysis whereas binding was などぬ considered to be negligible in CSF (14). Individual non-compartmental pharmacokinetic などね analysis was performed (Phoenix WinNonlin (version 6.2, Pharsight, US)) as previously などの described (13). The areas under the plasma and brain ECF unbound concentrations-time などは curves from dosing time to dosing interval (τ) (AUC0→τ) were calculated using the linear などば trapezoidal rule. The elimination rate constant ke and corresponding half-lives (t1/2) were などぱ determined by least squares fit of data points (log concentration time) in the terminal phase of などひ the decline. Steady state unbound body clearance of cefotaxime (CLss,u) and volume of ななど distribution (Vss,u) was calculated according to standard procedures. Results are expressed as ななな mean ± standard deviation (sd). ななに
ななぬ Results ななね Estimated mean unbound fractions (fu) of metronidazole and OH-metronidazole in plasma ななの were respectively equal to 89.3 ± 5.6 % and 88.0 ± 6.9 % and were independent of ななは concentrations. Individual unbound plasma and CSF metronidazole and OH-metronidazole ななば profiles are presented on Figure 1. Metronidazole unbound concentrations at steady-state ななぱ fluctuated between mean maximal and minimal concentrations equal to 21.7 ± 7.1 and 7.7 ± ななひ 4.4 µg.mL
-1 with an average concentration (Caverage= AUC0-τ/τ) at 12.7 ± 4.5 µg.mL
-1 in なにど
plasma, whereas in CSF concentrations profiles did not exhibit a clear peak and were rather なにな mostly flat with an average steady-state unbound concentration equal to 11.0 ± 5.0 µg.mL
-1 なにに
(Figure 1). The mean unbound metronidazole CSF to unbound plasma AUC0-τ ratio was equal なにぬ to 86 ± 16 % (Table 1). Concerning OH-metronidazole, unbound plasma concentrations なにね represent only 20 % of corresponding metronidazole values, which is unlikely to contribute なにの significantly to antimicrobial effect. Both unbound plasma and CSF versus time なには concentrations profiles were virtually flat (Figure 1) and the mean unbound OH-なにば metronidazole CSF to unbound plasma AUC0-τ ratio was equal to 79 ± 16 % (Table 1). All なにぱ pharmacokinetic parameters are presented in Table 1. なにひ なぬど Discussion なぬな なぬに Metronidazole systemic pharmacokinetic parameters values (Table 1) are consistent with なぬぬ those estimated during the microdialysis study (13). Although CNS distribution of なぬね metronidazole is generally considered to be extensive (15, 16), this is the first study to explore なぬの CSF distribution of metronidazole and OH-metronidazole in brain-injured patients with EVD なぬは
and un-inflamed meninges. Previously published studies on metronidazole CSF penetration in なぬば human (17-22) have mostly been performed with non-specific microbiological assay なぬぱ procedure, not allowing separation between metronidazole and OH-metronidazole (17-20). なぬひ Nevertheless, a first case report in adult with meningitis, using HPLC as analytical method, なねど described metronidazole CSF concentration after administration of 500 mg/6 h, but なねな concomitant plasma concentrations were not reported (22). A second case report in a なねに premature infant with bacterial meningitis reported results consistent with ours, in particular なねぬ with a CSF over plasma concentrations ratio close to one (21). なねね なねの In the present study, metronidazole demonstrates an extensive penetration through BCSFB as なねは it was suggested through BBB using brain microdialysis, when a mean metronidazole なねば unbound brain to plasma AUC0-τ ratio of 102 ± 19 % was found (13). These results are なねぱ consistent with the fact that metronidazole is not known to be substrate of efflux transport なねひ system at the CNS level. Concerning OH-metronidazole, mean CSF to plasma AUC0-τ ratio なのど was equal to 79 ± 16 % (Table 1). However, metronidazole concentrations profiles in ECF なのな and CSF were different. Indeed, concentration-time curves in brain ECF were delayed by なのに comparison with unbound plasma concentration-time curves, but a peak was still visible なのぬ within ECF with a mean maximal concentration equal to 14.5 ± 1.2 µg.mL
-1 (13), whereas なのね
CSF versus time profiles were essentially flat with a mean steady state concentration close to なのの 2.5 µg.mL
-1. Consequently, it was possible to estimate a metronidazole half-life in brain ECF なのは
by microdialysis (4.5 ± 0.2 h) not significantly different from that in plasma (6.3 ± 2.2 h) (13), なのば whereas metronidazole half-life in CSF could not be determined. Moreover, the absence of なのぱ peak is likely to have consequences on antimicrobial efficacy of concentration dependent なのひ antibiotics such as metronidazole where peak level and AUC parameters determine なはど antimicrobial in vivo efficacy (23), but also possibly on excitatory side effects. なはな
なはに It should also be mentioned that differences observed between CSF and ECF metronidazole なはぬ profiles must be interpreted with caution. First, although inter-patient variability was not なはね particularly large, this study was conducted with a small number of patients (n=4). But more なはの importantly, differences observed between ECF and CSF profiles during these two studies なはは could be due not only to differences between BBB and BCSFB, since respective patients also なはば differed by their physiopathology. Only patients recruited in this CSF distribution study なはぱ required EVD for intracranial hypertension and these hemodynamic alterations may influence なはひ BCSFB permeability, CSF turnover and eventually metronidazole CSF distribution. なばど Furthermore, EVD itself constitutes an extra route of drug elimination. Physiologically based なばな pharmacokinetic approaches (PB-PK) will be used in order to clarify these issues. なばに なばぬ なばね Conclusion なばの This was the first study describing the pharmacokinetics of metronidazole and OH-なばは metronidazole in CSF. Metronidazole and OH-metronidazole distribute extensively within なばば CSF but concentrations versus time profiles are relatively flat compared to brain ECF なばぱ concentrations estimated by microdialysis (13). Metronidazole penetrates extensively within なばひ brain consistent with passive diffusion. Yet CSF and ECF concentrations versus time profiles なぱど are not superimposable, and therefore CSF distribution studies should not be used to predict なぱな concentrations of this type of drugs within brain tissue. なぱに なぱぬ なぱね Acknowledgments なぱの
References なひば な┻ de Lange EC. にどなぬ┻ Utility of CSF in translational neuroscience┻ J Pharmacokinet なひぱ Pharmacodyn 40:ぬなの┽ぬには┻ なひひ に┻ Westerhout J, Danhof M, De Lange EC. にどなな┻ Preclinical prediction of human にどど brain target site concentrations┺ considerations in extrapolating to the clinical にどな setting┻ J Pharm Sci 100:ぬのばば┽ぬのひぬ┻ にどに ぬ┻ Dahyot-Fizelier C, Timofeev I, Marchand S, Hutchinson P, Debaene B, Menon にどぬ D, Mimoz O, Gupta A, Couet W. にどなど┻ Brain microdialysis study of meropenem にどね in two patients with acute brain injury┻ Antimicrobial agents and chemotherapy にどの 54:ぬのどに┽ぬのどね┻ にどは ね┻ Dahyot-Fizelier C, Frasca D, Gregoire N, Adier C, Mimoz O, Debaene B, Couet にどば W, Marchand S. にどなぬ┻ Microdialysis study of cefotaxime cerebral distribution in にどぱ patients with acute brain injury┻ Antimicrob Agents Chemother 57:にばぬぱ┽にばねに┻ にどひ の┻ Westerhout J, Ploeger B, Smeets J, Danhof M, de Lange EC. にどなに┻ になど Physiologically based pharmacokinetic modeling to investigate regional brain になな distribution kinetics in rats┻ AAPS J 14:のねぬ┽ののぬ┻ になに は┻ Westerhout J, Smeets J, Danhof M, de Lange EC. にどなぬ┻ The impact of P┽gp になぬ functionality on non┽steady state relationships between CSF and brain になね extracellular fluid┻ J Pharmacokinet Pharmacodyn 40:ぬにば┽ぬねに┻ になの ば┻ Rambeck B, Jurgens UH, May TW, Pannek HW, Behne F, Ebner A, Gorji A, になは Straub H, Speckmann EJ, Pohlmann-Eden B, Loscher W. にどどは┻ Comparison of になば brain extracellular fluid┸ brain tissue┸ cerebrospinal fluid┸ and serum になぱ concentrations of antiepileptic drugs measured intraoperatively in patients with になひ intractable epilepsy┻ Epilepsia 47:はぱな┽はひね┻ ににど ぱ┻ Shin SH, Kim KS. にどなに┻ Treatment of bacterial meningitis┺ an update┻ Expert ににな opinion on pharmacotherapy 13:になぱひ┽ににどは┻ ににに ひ┻ Bernstein LH, Frank MS, Brandt LJ, Boley SJ. なひぱど┻ (ealing of perineal Crohnｆs ににぬ disease with metronidazole┻ Gastroenterology 79:のひひ┻ ににね など┻ Kim E, Na DG, Kim EY, Kim JH, Son KR, Chang KH. にどどば┻ MR imaging of ににの metronidazole┽induced encephalopathy┺ lesion distribution and diffusion┽にには weighted imaging findings┻ AJNR Am J Neuroradiol 28:なはのに┽なはのぱ┻ ににば なな┻ Patel K, Green-Hopkins I, Lu S, Tunkel AR. にどどぱ┻ Cerebellar ataxia following ににぱ prolonged use of metronidazole┺ case report and literature review┻ )nt J )nfect Dis ににひ 12:eななな┽ななね┻ にぬど なに┻ Seok JI, Yi H, Song YM, Lee WY. にどどぬ┻ Metronidazole┽induced encephalopathy にぬな and inferior olivary hypertrophy┺ lesion analysis with diffusion┽weighted imaging にぬに and apparent diffusion coefficient maps┻ Arch┻ Neurol┻ 60:なばひは┽なぱどど┻ にぬぬ なぬ┻ Frasca D, Dahyot-Fizelier C, Gregoire N, Adier C, Mimoz O, Debaene B, Couet にぬね W, Marchand S. accepted┻ Metronidazole and hydroxy┽metronidazole central にぬの nervous system distribution┺ な┻ Microdialysis assessment of brain extracellular にぬは fluid concentrations in patients with acute brain injury┻ Journal of Antimicrobial にぬば Chemotherapy┻ にぬぱ なね┻ Regeniter A, Kuhle J, Mehling M, Moller H, Wurster U, Freidank H, Siede WH. にぬひ にどどひ┻ A modern approach to CSF analysis┺ pathophysiology┸ clinical application┸ にねど proof of concept and laboratory reporting┻ Clin Neurol Neurosurg 111:ぬなぬ┽ぬなぱ┻ にねな なの┻ Lau AH, Lam NP, Piscitelli SC, Wilkes L, Danziger LH. なひひに┻ Clinical にねに pharmacokinetics of metronidazole and other nitroimidazole anti┽infectives┻ にねぬ Clinical pharmacokinetics 23:ぬにぱ┽ぬはね┻ にねね
なは┻ Lamp KC, Freeman CD, Klutman NE, Lacy MK. なひひひ┻ Pharmacokinetics and にねの pharmacodynamics of the nitroimidazole antimicrobials┻ Clinical にねは pharmacokinetics 36:ぬのぬ┽ぬばぬ┻ にねば なば┻ O'Grady LR, Ralph ED. なひばは┻ Anaerobic meningitis and bacteremia caused by にねぱ Fusobacterium species┻ American journal of diseases of children 130:ぱばな┽ぱばぬ┻ にねひ なぱ┻ Feldman WE. なひばは┻ Bacteroides fragilis ventriculitis and meningitis┻ Report of にのど two cases┻ American journal of diseases of children 130:ぱぱど┽ぱぱぬ┻ にのな なひ┻ Jokipii AM, Myllyla VV, Hokkanen E, Jokipii L. なひばば┻ Penetration of the blood にのに brain barrier by metronidazole and tinidazole┻ J Antimicrob Chemother 3:にぬひ┽にのぬ にねの┻ にのね にど┻ Davies AH. なひはば┻ Metronidazole in human infections with syphilis┻ The British にのの journal of venereal diseases 43:なひば┽にどど┻ にのは にな┻ Warner JF, Perkins RL, Cordero L. なひばひ┻ Metronidazole therapy of anaerobic にのば bacteremia┸ meningitis┸ and brain abscess┻ Archives of internal medicine にのぱ 139:なはば┽なはひ┻ にのひ にに┻ Soriano F, Aguado JM, Tornero J, Fernandez-Guerrero ML, Gomez-Garces JL. にはど なひぱは┻ Bacteroides fragilis meningitis successfully treated with metronidazole にはな after a previous failure with thiamphenicol┻ J Clin Microbiol 24:ねばに┽ねばぬ┻ にはに にぬ┻ Craig WA. にどどば┻ Pharmacodynamics of Antimicrobials┺ General Concepts and にはぬ Applications┻ )n Nightingale C(┸ Ambrose PG┸ Drusano GL┸ Murakawa T┻ ゅEdょ┻ にはね Antimicrobial Pharmacodynamics in Theory and Clinical Practice┻ Second Edition┻ にはの )nforma (ealthcare┸ New York ゅNYょ┻ にはは にはば
にはひ Figure 1. Individual steady state unbound plasma concentrations of metronidazole (Ɣ, full にばど line), OH-metronidazole (Ÿ, dashed line), CSF metronidazole (ż, full line) and hydroxy-にばな metronidazole (, dashed line) concentrations after 500 mg metronidazole infusion over 30 にばに min every 8 h in critical care patients. In patient 3, unbound plasma concentrations of OH-にばぬ metronidazole were not determined due to analytical interference. にばね にばの
