Université catholique de Louvain Louvain Drug Research Institute Centre hospitalo-universitaire de Mont-Godinne Continuous Infusion of Vancomycin in non-ICU Patients: Why, How and What's the Benefit? Els Ampe, Pharm. Thesis submitted in view to obtain the degree of Doctor in Biomedical and Pharmaceutical Sciences (orientation "Clinical Pharmacy") Co-promotors: Professor Youri Glupczynski Professor Paul M. Tulkens 2013 Table of Contents page Foreword and Acknowledgements 1 General Introduction 5 Chapter 1: Continuous infusion of -lactams and vancomycin: a review 15 Aims and objectives 103 Chapter 2: Overcoming Insufficiencies in Therapeutic Drug Monitoring of 105 Vancomycin by Switching from Intermittent Administration to Continuous Infusion: a Combined Observational and Qualitative Study [submitted] Chapter 3: Implementation of a Protocol for Administration of Vancomycin 140 by Continuous Infusion: Pharmacokinetic, Pharmacodynamic and Toxicological aspects [published in International Journal of Antimicrobial Agents: 41 (2013) 439– 446] General Discussion and perspectives 162 Additional Material Correlation between Free and Total Vancomycin Serum 182 Concentrations in Patients treated for Gram-positive Infections [published in International Journal of Antimicrobial Agents: 34 (2009) 555–560] Stability and Compatibility of Vancomycin for Administration by 188 Continuous Infusion [published in Journal of Antimicrobial Chemotherapy: 68 (2013) 1179–1182] Foreword and Acknowledgments When I finished my Hospital Pharmacy education at the Katholieke Universiteit Leuven (KU-Leuven), clinical pharmacy was only beginning its developments in Belgium. However, I got the unique opportunity to join the Université catholique de Louvain (UCL) to be one of the first Belgian pharmacist making a PhD thesis in Infectious Disease based on a Clinical Pharmacy approach. I had the opportunity to work with the Infectious Disease Management Team of the Cliniques Universitaires de Mont-Godinne (now CHU Mont-Godinne) where I learned a lot about Infectious Diseases during the tours in the Microbiology Laboratory and in different wards. Soon, we were able to determine what my research project would focus on. It was a difficult but also challenging task for a pharmacist because many aspects of antibiotic therapy were already well covered by the infectious disease physicians. Pharmacists interested in Infectious Diseases often specialize in subareas by focusing on pharmacovigilance, pharmaco-economics or clinical pharmacology and pharmacokinetics. The latter seemed a good starting point as (i) there were perceived but not fully analyzed issues about the quality of therapeutic drug monitoring (TDM) of antibiotics at that time, and (ii) the Cellular and Molecular Pharmacology group to which I was member through my nomination as "Assistant universitaire" was deeply involved in basic research on pharmacology, pharmacodynamics, and pharmacokinetics of antibiotics. So, the basic starting point of my thesis was now set, and I started to analyze the performance of TDM of vancomycin and amikacin, which would eventually lead me to what is now the topic of this Thesis. I want to emphasize that it would have been impossible to accomplish this Thesis project without the help and support of many people. I want to address special thanks to: pFaogreew 1ord and Acknowledgments  Mon promoteur, le Professeur Paul M. Tulkens pour m’avoir donné la possibilité d’entamer un des premiers projets de recherche dans ce domaine. Merci, Mr. Tulkens, d’avoir cru en ce projet, de m’avoir fait confiance et d’avoir investi tant d’énergie à le rendre réalisable. Je vous remercie également pour l’encadrement scientifique tout au long de ce projet. J’ai vraiment été encouragée par votre enthousiasme éternel et votre passion pour la science, plus particulièrement pour la pharmacologie et la pharmacie clinique.  Mon co-promoteur, le Professor Youri Glypczynski, pour m'avoir accueilli dans son laboratoire et de m’avoir permis d’apprendre le rôle essentiel que joue le laboratoire de chimie et de microbiologie clinique dans l’adaptation des traitements anti-infectieux au niveau du patient. Merci Monsieur Glupczynski d’avoir partagé avec moi votre connaissance de la Microbiologie Clinique et d’avoir eu beaucoup d’attention pour certains aspects pratiques importants pour la réalisation de ce projet. Merci également pour l’analyse de tous les isolats cliniques pertinents et pour vos suggestions utiles quant aux publications réalisées.  Le Professeur Bénédicte Delaere, pour avoir aiguisé mon esprit critique quant au traitement des maladies infectieuses. Je suis convaincue que l’implémentation de l’infusion continue de la vancomycine aux Cliniques Universitaires de Mont-Godinne a eu le succès qu'on lui reconnait aujourd'hui grâce à votre engagement permanent dans ce projet. Merci également d’avoir pris le temps de discuter ensemble malgré un agenda déjà trop rempli et de m’avoir aidé à valider les données cliniques.  Le Docteur Sc. Pharm. Jean-Daniel Hecq, pour votre accueil au Département de Pharmacie du CHU Mont-Godinne. Votre suggestion d’utiliser une préparation standardisée prêt à l’emploi de vancomycine a considérablement amélioré la qualité de ce projet. Votre collaboration amicale, et votre disponibilité ont été particulièrement précieux pour moi. Grace à vous j’ai pu faire connaissance de l’énorme expérience de pFaogreew 2ord and Acknowledgments nos collègues Américains dans ce domaine. Merci de votre investissement dans la qualité des services de pharmacie hospitalière et clinique. Je remercie également les membres du comité d’encadrement (les Professeurs Véronique Préat, Pierre Wallemacq et Emmanuel Hermans) pour leurs conseils avisés. I want also to thank the external members of the jury for having accepted to participate to the evaluation of Thesis: Professor Hartmut Derendorf, PhD. (Distinguished Professor of Pharmaceutics at the University of Florida) and Professor Frédérique Jacobs (Chef de service de maladies Infectieuses de l’Hôpital Erasme de l’Université Libre de Bruxelles). Your well known and very appreciated activities in the field of anti-infective pharmacology are the best guarantee of the quality of my work and I look forward hearing from you how it could be improved. J’ai eu la chance de travailler avec de nombreuses autres personnes venant d’horizons divers et d’avoir bénéficié de leur compétence et leur aide dans la collecte et l’analyse des données. Je remercie particulièrement Vincent Lorant pour son aide dans l’analyse des données qualitatives. Les collègues actifs en Pharmacie clinique et en Pharmacologie des antibiotiques m'ont aussi aidé à valider les documents utilisés. J'adresse des remerciements particuliers à Anne Spinewine, Jean-Marc Feron, Catherine Bouland, Stéphane Carryn, Séverine Noirhomme, Karine Berthoin, Laurence Galanti, Patricia Gillet, Catherine Berhin, Martial Vergauwen, Cédric Baude et Joseph Mathieu. Je remercie particulièrement le Professeur Françoise Van Bambeke. Merci, Françoise de tout ce que tu m’as appris pendent les séminaires de pharmacothérapie, pour ta présence et ton soutien. pFaogreew 3ord and Acknowledgments Je veux également exprimer toute ma reconnaissance aux patients et aux professionnels de santé qui ont acceptés de participer à ce projet, parfois dans des situations difficiles. Si ce projet a réussi, c’est grâce à eux, merci. Merci à tous ceux qui ont montré de l’intérêt dans ce projet et que je n’ai pas cités personnellement. Bedankt aan de collega’s van de ziekenhuisapotheek en het Centrum Klinische Farmacologie van UZ Leuven voor de vele gesprekken en om mij op weg te helpen met weer eens een nieuw statistiekprogramma. Jullie gedrevenheid en passie voor dit vak hebben mij de moed gegeven om door te gaan tot de eindstreep. Tenslotte wil ik mijn familie en vrienden bedanken voor hun liefde, hun eindeloze begrip en hun onvoorwaardelijke steun. Een bijzonder woordje van dank aan mijn ouders, mijn schoonfamilie, Pieter en Dany, Gust, Winand, Koen en Elena. Jullie morele en fysieke bijdrage in het realiseren van deze thesis waren van onschatbare waarde. Raf en Samuel, jullie hebben mij de voorbije tijd wat meer moeten missen. Bedankt om er voor mij te zijn. Jullie slaagden er iedere keer weer in om mij de thesis-stress van de voorbije maanden te doen vergeten. pFaogreew 4ord and Acknowledgments General Introduction In light of the increasing resistance of micro-organisms towards antibiotics and the limited number of new antimicrobial agents in clinical development, the optimal use of available drugs is more then ever important [1]. New insights have emerged from the study of the pharmacokinetics/pharmacodynamics (PK/PD) of antibiotics [2;3], several of which led to successful implementation in new clinical practices, such as the once daily dosing of aminoglycosides [4-7] or dose adaptation based on MIC for fluoroquinolones [8;9]). Infectious disease management teams in which medical doctors and clinical pharmacists collaborate have delivered important work in this context (see [7;10] typical examples). We focused our attention on vancomycin. This antibiotic was originally isolated by Eli-Lilly from a soil sample coming from Borneo and containing the actinomycete Nocardia Orientalis [11]. The molecule was found to have in vitro activity against Staphylococci [12] and animal studies showed a low level of toxicity [13]. The drug was eventually approved and used in clinical practice in 1958 for the treatment of Gram-positive infections. Methicillin, licensed in 1960, resulted in the decline of vancomycin use but the appearance of methicillin resistance in the 1980s renewed interest in vancomycin [14;15]. Today, vancomycin is at the forefront of clinical use for the treatment of infections caused by Gram-positive organisms resistant to -lactams [16;17], although its restricted use has been largely advocated [18;19]. The chemical structure of vancomycin (Figure 1.1.) was confirmed in 1978 (CAS registry number 1404-93-9, molecular formula C66H75C12N9O24; molecular weight 1449 g/mol). The main properties of vancomycin related to its chemistry and mode of action have been summarized in 4 key review articles [15;20-22] from which we have extracted the following information. pGaegnee r5al Introduction OH H3C NH2 CH3 O OH H HO O HO O O Cl O O HO OH Cl O O O H H H O N N N N N N CH3 H H H HN COOH O O O H3C NH2 CH3 OH HO OH vancomycin Molecular Weight: 1449.25 Molecular Formula: C66H75Cl2N9O24 Derivative Types: Monohydrochloride Figure 1: structural formula of vancomycin The molecule consists of a heptapeptide core (in which two peptides bear a chloride) substituted with vancosamine and glucose sugars. The heptapeptide core is responsible for the pharmacological activity of the molecule, whereas the sugars are thought to modulate its hydrophylicity and its propensity to form dimers. As a result of its large size, vancomycin is unable to cross the outer membrane of Gram-negative bacteria explaining inactivity against these organisms. Inability to penetrate inside bacteria limits vancomycin activity to extracellular targets. Vancomycin inhibits the late stage of cell wall peptidoglycan synthesis by binding to the D-Ala-D-Ala termini of the pentapeptide ending precursors localized at the outer surface of the cytoplasmatic membrane. It forms a high affinity complex with the D-Ala- D-Ala by forming hydrogen bonds with the heptapeptide core. The strength of this bond is enhanced by vancomycin dimerization. The steric hindrance around the pentapeptide termini blocks the reticulation of peptidoglycan by inhibiting the activity of transglycolsylases responsible for attaching the new disaccharide-pentapeptide subunit to the nascent peptidoglycan and of transpeptidases catalysing the formation of interpeptide bridges. pGaegnee r6al Introduction Vancomycin shows a volume of distribution (Vd) of about 0.7 L/kg, an half-live about 6-12 h, about 50 % protein binding, and a moderate post antibiotic effect. Because of lack of intestinal resorption, vancomycin is exclusively administered by the intravenous route for systemic infections. The most pertinent PK/PD index predicting vancomycin efficacy is the ratio between the 24 h Area under the serum concentration curve (AUC24h) and the MIC of the offending organism [2], with a value of 350-400 for severe lung infections caused by Staphylococcus aureus [23]. Vancomycin is commercialized as a hydrochloride salt and is most soluble at pH 3 to 5. Solubility and stability decreases at increasing pH. Reconstitution of commercially available vancomycin is made in water with further dilution in Glucose 5% or saline up to maximally recommended concentrations up to 10 mg/L. Since its introduction in clinical practice, vancomycin has been subject of extensive debate [24]. Initial concerns rose for reasons of renal toxicity which seemed to be caused to some extent by impurities in the first commercial preparations leading to the nickname ‘Mississippi mud’ [25]. Renal toxicity greatly improved after marketing more pure preparations. In the meantime monitoring of blood levels for reason of toxicity had become routine clinical practice [26]. The other important part of the debate was about concerns of efficacy in deep seated infections with high inocula due to its slow killing rate [2] associated with inter-individual differences in pharmacokinetics [27] and limited tissue penetration [28-30]. The raising MICs of Gram-positive organisms for vancomycin (commonly referred to as "MIC creep" [31;32]) has only further fuelled this debate and several publications have questioned whether vancomycin remains a viable treatment option today [33-37]. Despite all these limitations, no clinical studies have proven global and clear superiority for alternative agents and vancomycin still remains the drug of choice for the majority of infections caused by methicillin resistant Gram-positive infections [38;39]. Measurement of serum vancomycin concentrations by therapeutic drug monitoring is widely recommended in routine practice as it is supposed to allow for dose readjustment on an individual patient level in order to optimize efficacy and avoid toxicity [40;41]. pGaegnee r7al Introduction Extensive pharmacokinetic studies in a variety of patient populations have been conducted [42-48]. Commercial drug assays have allowed clinicians to target serum vancomycin concentrations in routine practice. There is some controversy that has resulted from conflicting evidence regarding the use of serum vancomycin concentrations to predict and prevent drug induced toxicity and as a measure of effectiveness in treating infections [49]. Further, data derived from more recent studies appear to suggest that vancomycin has little potential for nephrotoxicity or ototoxicity when used at conventional dosages unless it is used concomitantly with known nephrotoxic drugs or at very high dosages [50]. Several studies did not find a clear correlation between vancomycin levels and toxicity. The use of a nomogram is an alternative method for dosage adjustments, with the original one and still widely used proposed by Moellering et al. [51]. Several others have been proposed but not fully clinically validated and often point to too low trough levels (10–15 mg/l), which, as we shall see, is not consistent with current recommendations. Recent North American guidelines recommend conventional twice daily dosing (BID) for vancomycin with through levels around 15-20 mg/L [17;40] because of recent pharmacokinetic insights about the risk of subtherapeutic doses in face of less susceptible organisms. This practice, however, has some important limitations. Firstly, higher trough levels have been associated with significantly higher rates of nephrotoxicity [52-54] and it is therefore important to detect them. Secondly, the omission of peak levels withdraws information about the exact AUC obtained, although, as stated above, it is AUC24h/MIC that governs the overall activity of the drug. BID was the common practice in our institution but with therapeutic monitoring of both peak and through levels to meet the criticisms raised against "trough level only" determinations. The following recommendations were in place: standard doses of 1g every 12 h (to be modulated according to the renal function and TDM results); target peak and trough levels of 30-40 mg/L and 5-10 mg/L respectively. Based on its PK/PD index, continuous infusion (CI) of vancomycin should be equally effective compared to its BID schedule. The AUC24h of an intravenously administered drug depends, indeed, only on the ratio between the total drug daily dose and drug creatinine clearance, irrespective of is schedule of administration. pGaegnee r8al Introduction