1 13 September 2018 2 EMA/CVMP/AWP/842786/2015 3 Committee for Medicinal Products for Veterinary Use (CVMP) Reflection paper on the use of aminopenicillins and their 4 beta-lactamase inhibitor combinations in animals in the 5 European Union: development of resistance and impact 6 on human and animal health 7 8 Draft Draft agreed by Antimicrobials Working Party (AWP) 9 July 2018 Adopted by CVMP for release for consultation 13 September 2018 Start of public consultation 21 September 2018 End of consultation (deadline for comments) 21 December 2018 9 Comments should be provided using this template. The completed comments form should be sent to [email protected] 10 Keywords aminopenicillins, antimicrobial resistance 11 30 Churchill Place ● Canary Wharf ● London E14 5EU ● United Kingdom Telephone +44 (0)20 3660 6000 Facsimile +44 (0)20 3660 5555 Send a question via our website www.ema.europa.eu/contact An agency of the European Union © European Medicines Agency, 2018. Reproduction is authorised provided the source is acknowledged. 12 Table of contents 13 14 Executive summary ..................................................................................... 4  15 CVMP Recommendations for action ............................................................. 6  16 1. Background ............................................................................................. 7  17 2. General drug characteristics .................................................................... 8  18 2.1. Structure and mechanism of action ....................................................................... 8  19 2.2. Antimicrobial spectrum ....................................................................................... 8  20 2.3. Pharmacodynamics ............................................................................................. 9  21 2.4. Pharmacokinetics ............................................................................................. 10  22 3. Resistance mechanisms and susceptibility testing ................................ 12  23 3.1. Resistance mechanisms .................................................................................... 12  24 3.1.1. Enzymatic degradation of beta-lactams by beta-lactamases ................................. 12  25 3.1.2. Modification of the target site .......................................................................... 17  26 3.1.3. Other resistance mechanisms .......................................................................... 17  27 3.2. Susceptibility testing ........................................................................................ 17  28 4. Sales and use of aminopenicillins and their inhibitor combinations in 29 veterinary medicine ................................................................................... 19  30 4.1. Sales .............................................................................................................. 19  31 4.2. Use and indications in food-producing animals ...................................................... 19  32 4.3. Use and indications in horses ............................................................................. 20  33 4.4. Use and indications in companion animals ........................................................... 21  34 5. The use of aminopenicillins and their inhibitor combinations in human 35 medicine .................................................................................................... 25  36 5.1. Indications in human medicine ........................................................................... 25  37 5.2. Consumption of aminopenicillins in humans in the EU ............................................ 26  38 6. Occurrence of resistance ....................................................................... 29  39 6.1. Resistance in bacteria of animal origin covered by EU surveillance ........................... 30  40 6.2. Resistance in animal target pathogens ................................................................ 31  41 6.3. Resistance in human pathogens ......................................................................... 34  42 7. Possible links between the use of aminopenicillins and their inhibitor 43 combinations in animals and resistance in bacteria of animal origin ......... 36  44 8. Impact of resistance on animal and human health ................................ 37  45 8.1. Animal health .................................................................................................. 37  46 8.2. Human health .................................................................................................. 38  47 9. Transmission of resistance or resistance determinants between animals 48 and humans ............................................................................................... 39  49 9.1. Transmission of resistant bacteria ....................................................................... 39  50 9.2. Transmission of resistance determinants ............................................................. 41  Reflection paper on the use of aminopenicillins and their beta-lactamase inhibitor combinations in animals in the European Union: development of resistance and impact on human and animal health EMA/CVMP/AWP/842786/2015 Page 2/65 51 10. Discussion ........................................................................................... 44  52 11. Conclusions ......................................................................................... 46  53 12. References .......................................................................................... 53  54 Appendix ................................................................................................... 65  55 56 Reflection paper on the use of aminopenicillins and their beta-lactamase inhibitor combinations in animals in the European Union: development of resistance and impact on human and animal health EMA/CVMP/AWP/842786/2015 Page 3/65 Executive summary 57 58 The objective of this document is to review available information on the use of aminopenicillins and 59 their beta-lactamase inhibitor combinations in veterinary medicine in the EU, their effect on the 60 emergence of antimicrobial resistance (AMR) and the potential impact of resistance on human and 61 animal health. The document provides information for the risk profiling, as recommended by the 62 Antimicrobial Advice ad hoc Expert Group (AMEG) of the EMA, to assist with placing these substances 63 within the AMEG’s categorisation (EMA/AMEG, 2014). The focus of this paper is on veterinary 64 aminopenicillins authorised in the EU, which are ampicillin (ATC J01CA01), amoxicillin (ATC J01CA04), 65 and their beta-lactamase inhibitor combination amoxicillin-clavulanic acid (J01CR02). 66 The WHO classifies penicillins (natural, aminopenicillins and antipseudomonal) as critically important 67 antimicrobials (CIA) for humans. According to the WHO, the CIA status is justified due to limited 68 therapy options for listeriosis and infections caused by Enterococcus spp., and the likelihood of 69 transmission of resistant Enterococcus spp. and Enterobacteriaceae, including both Salmonella spp. 70 and Escherichia coli, from non-human sources to humans. 71 Although aminopenicillins are seldom among the sole treatment options, with the exception of for 72 Listeria and enterococci, they are often used as first line antimicrobials for many infections in animals 73 and humans. In animals aminopenicillins are used for infections caused by species belonging to 74 Pasteurellaceae, Streptococcus spp., Staphylococcus spp., Erysipelothrix rhusiopathiae, Listeria 75 monocytogenes, Clostridium spp. and other anaerobic species, Bordetella bronchiseptica and species 76 belonging to the Enterobacteriaceae. Aminopenicillins and their inhibitor combinations are very 77 valuable drugs for treating respiratory infections in humans caused by Streptococcus pneumoniae, 78 Haemophilus influenzae, and Branhamella catarrhalis. Due to the abundant presence of beta- 79 lactamases in E. coli and in many other Enterobacteriaceae, aminopenicillins are combined with beta- 80 lactamase inhibitors for the treatment of infections caused by these bacteria. Inhibitor combinations 81 can also be useful in certain infections caused by ESBL-producing E.coli provided that an isolate is 82 susceptible to the combination in vitro. The combination is ineffective against AmpC-mediated 83 resistance. 84 Ampicillin, amoxicillin, and to a lesser extent amoxicillin-clavulanic acid combinations have been widely 85 used for decades for the treatment of infections in several animal species in European countries. 86 Measured in mg/PCU (population correction unit), penicillins were the second most used antimicrobial 87 class in food-producing animals in the EU in 2015 and accounted for 25% of the total sales. 88 Aminopenicillins (amoxicillin) made up the major proportion (88%) of the total penicillin use, while 89 their inhibitor combinations formed a very limited fraction of the total penicillin use. There are 90 substantial differences between the uses of different beta-lactam drug classes in animals in Nordic 91 countries, where benzyl penicillin and its pro-drugs dominates, vs. in other European countries, where 92 aminopenicillins are the prevailing beta-lactams used. This may be due to differences in treatment 93 guidelines, availability of authorised products, production systems (including dominant animal species), 94 herd sizes, disease occurrences, and production facilities, or even manners and habits of antimicrobial 95 usage (e.g. whether mass medication is favoured instead of individual treatment). 96 According to the European Surveillance of Antimicrobial Consumption Network (ESAC-Net) summary 97 report on 2016 data, the most commonly used antimicrobials in human medicine were penicillins (ATC 98 J01C), however data that specify the human use of those aminopenicillins (J01CA01, J01CA04) and 99 inhibitor combinations (J01CR02) that also have authorisation for animals, are not readily available. If Reflection paper on the use of aminopenicillins and their beta-lactamase inhibitor combinations in animals in the European Union: development of resistance and impact on human and animal health EMA/CVMP/AWP/842786/2015 Page 4/65 100 human and animal beta-lactam use are compared as mg/kg of estimated biomass, human use is 101 approximately twice that for animals (80 vs. 40 mg/kg of estimated biomass). 102 Aminopenicillin (or penicillin) resistance has not yet been described in group A, B, C or G beta- 103 hemolytic streptococci, regardless of origin (animal/human). Aminopenicillin resistance in clinical 104 Listeria monocytogenes is very rare. Regarding other streptococci, enterococci (mainly E. feacalis) and 105 Pasteurellaceae, penicillin/aminopenicillin non-susceptibility levels are generally low but vary by 106 country, production system, and animal and bacterial species. More than 75% of human E. faecium 107 isolates show resistance to ampicillin while less resistance has been detected in isolates of animal 108 origin. 109 Aminopenicillins are able to select not only for aminopenicillin resistance, but also co-select for other 110 resistances, including to extended spectrum cephalosporins. It is clear that resistant organisms, such 111 as MRSA and those producing ESBL/AmpC, are transferred between animals and humans but both the 112 direction and magnitude of transfer are often difficult to prove or quantify. The pathway from animals 113 to humans is obvious for zoonotic organisms, such as salmonellae and campylobacters, which cause 114 illness in humans. Also the origin of certain LA-MRSA clones is proven to be in livestock, but for 115 commensals that are part of the normal microbiota, the role of animals as the source of resistance is 116 unclear. Although identical clones, the same resistance genes and mobile genetic elements have been 117 detected in many bacteria of animal and human origin, the effect of veterinary antimicrobial use on 118 their presence or emergence in the human population is equivocal. For example, studies utilising new 119 sequencing methods have revealed high genetic diversity between the isolates from different sources 120 indicating that veterinary antimicrobial use might not have a major impact on selection of ESBL/AmpCs 121 detected in humans. Resistance to aminopenicillins is common in E. coli of animal and human origin, 122 but resistance levels to the inhibitor combinations in bacteria of animal origin are lower. 123 Considering that aminopenicillin resistance is at a very high level in some organisms and that 124 aminopenicillins have been extensively used for decades both in animals and humans, it is currently 125 impossible to estimate to what extent the use of these substances in animals, could create negative 126 health consequences to humans at the population level. There are studies that have attempted to 127 address these challenges. In general, risk estimates range from a few additional infections per million 128 at risk to thousands, depending on antimicrobial substance and pathogen in question. Individual risk 129 estimates following assessments of aminopenicillin resistance exposure via the food might be low, 130 especially if good food hygiene practices are followed. However other routes of exposure should be 131 taken into consideration (such as direct contact). 132 Although the direct AMR risk to humans from the veterinary use of aminopenicillins would be lower 133 compared to the risk from their use in human medicine, it is evident that veterinary aminopenicillin use 134 increases the selection pressure towards AMR and jeopardizes at least animal health and welfare. 135 Based on an assessment of current use and resistance profiling, it may be possible to make 136 recommendations to limit the further development of resistance to both aminopenicillins and related 137 classes of antimicrobials and to maintain the efficacy of these valuable drugs in the future. Tools 138 include improvements in hygiene in animal husbandry, use of vaccinations, proper diagnostics and 139 avoidance of use of antimicrobials prophylactically to animals having no signs of infection. Also, the 140 route of administration should be considered to reduce the selection pressure in the gut microbiota. For 141 example group medication of food-producing animal flocks by the oral route facilitates the selection 142 and spread of resistance and attempts to reduce such use are needed. 143 Reflection paper on the use of aminopenicillins and their beta-lactamase inhibitor combinations in animals in the European Union: development of resistance and impact on human and animal health EMA/CVMP/AWP/842786/2015 Page 5/65 CVMP Recommendations for action 144 145 Proposal on categorisation for consideration by AMEG 146  The AMEG categorisation considers the risk to public health from AMR due to the use of 147 antimicrobials in veterinary medicine. The categorisation is based primarily on the need for the 148 antimicrobial in human medicine, and the risk for spread of resistance from animals to 149 humans. Aminopenicillins are important in human medicine in terms of their high extent of use 150 to treat a variety of important infections, although there are alternatives of last resort. 151 Aminopenicillins have potential to select LA-MRSA and resistance in foodborne zoonotic 152 pathogens, including Salmonella spp., which can be transferred to humans from livestock. In 153 addition, resistance to aminopenicillins is very frequent in commensal Enterobacteriaceae from 154 food-producing animals in the EU, which could act as a reservoir for resistance genes that may 155 be transferred to pathogenic bacteria in humans. However, the high extent of aminopenicillin 156 use in humans itself provides a selection pressure for resistance in the human microbiota and 157 the significance to public health of additional aminopenicillin resistance transferred from 158 animals is considered to be low. Although amoxicillin beta-lactamase inhibitor combinations 159 have very low use in food-producing animals, AmpC/ESBL resistance mechanisms, which also 160 confer resistance to 3rd- and 4th-generation cephalosporins, have emerged in 161 Enterobacteriaceae from animals in recent years and the combination has the potential to 162 select further these types of resistance than amniopenicillins alone. 163  It should also be considered that aminopenicillins have been widely used for decades in 164 veterinary medicine in the EU, and that they are categorised as veterinary CIAs by the OIE on 165 the grounds that they are very important in the treatment of many diseases in a broad range 166 of animal species. 167  All these factors should be taken into account for the AMEG’s categorisation, which is currently 168 under review. It is suggested that the AMEG could give consideration to a further stratification 169 of the categorisation to allow a distinction in the ranking between those substances currently in 170 Category 2 (fluoroquinolones, 3rd- and 4th-generation cephalosporins and colistin, for which 171 there are fewer alternatives) and the amoxicillin-clavulanate combinations, and between the 172 latter and the straight aminopenicillins. Amoxicillin-clavulanate has wider spectrum and thus it 173 is likely that it has higher chance to select multidrug resistant organisms compared to 174 aminopenicillin alone. In case accumulating evidence from future scientific research indicates 175 that veterinary use of aminopenicillins poses an added threat to public health due to animal-to- 176 human resistance transfer, it could then be considered if a distinction in the categorisation 177 should be made between straight aminopenicillins and narrow-spectrum penicillins. 178 179 Considerations for Marketing Authorisations and summary of product 180 characteristics (SPCs) 181  Current indications should be reviewed in relation to authorised dosing regimens in order to 182 ensure achievement of sufficient pharmacokinetic/pharmacodynamic (PK/PD) targets and 183 subsequently to minimise the risk for resistance selection, especially concerning inherently less 184 susceptible organisms such as Enterobacteriaceae and Bordetella bronchiseptica. 185  Since there is great variation in dosing regimens between similar products authorised in the EU, 186 these should be reviewed to harmonise schemes and ensure effective dosing. Reflection paper on the use of aminopenicillins and their beta-lactamase inhibitor combinations in animals in the European Union: development of resistance and impact on human and animal health EMA/CVMP/AWP/842786/2015 Page 6/65 187  In reference to the above recommendations and the scope of any referral procedures for 188 aminopenicillins and their combinations, review of groups of products would be prioritised 189 according to their relative risk to animal and public health. 190  Based on high levels of resistance in Enterobacteriaceae, it is recommended that the use of 191 aminopenicillins for the treatment of infections caused by such pathogens should be based on 192 susceptibility testing. 193 Responsible parties: CVMP, Regulatory Agencies, Marketing Authorisation Holders (MAHs) 194 195 Need for research 196  Susceptibility testing should be standardised and veterinary clinical breakpoints should be 197 established for aminopenicillins to enable proper interpretation of susceptibility tests. 198  There is need for a harmonised European wide surveillance scheme to encompass target 199 pathogens from food-producing and companion animals. 200  The same resistance genes carried by the same mobile genetic elements have been found in 201 isolates from animals and humans and there is potential for transmission of resistance from 202 animals to humans. Further research is needed to elaborate on the link between the use of 203 antimicrobials in animals and the impact on public health. 204 Responsible parties: European Commission, EURL-AMR, EFSA, VetCAST 1. Background 205 206 As part of the EC Action plan against antimicrobial resistance (AMR), the European Commission (EC) 207 requested advice from the European Medicines Agency (EMA) on the impact of the use of 208 antimicrobials in animals on public and animal health and measures to manage the possible risks it 209 may cause to humans. This is because aminopenicillins, especially those combined with beta-lactamase 210 inhibitors, have a spectrum of activity which overlaps with 2nd- and to lesser extent 3rd-generation 211 cephalosporins. Thus they might have the ability to select and facilitate the spread of bacteria carrying 212 extended spectrum beta-lactamases (ESBLs), similarly to 3rd- and 4th-generation cephalosporins and 213 fluoroquinolones (EMA/AMEG, 2014). WHO classifies penicillins (natural, aminopenicillins and 214 antipseudomonal) as critically important antimicrobials (CIA) for human medicine (WHO, 2017). 215 As in the concept paper published by the CVMP (EMA/CVMP, 2015a), the focus of this paper is on 216 veterinary authorised extended-spectrum penicillins in the EU, which are the aminopenicillins ampicillin 217 (ATC J01CA01) and amoxicillin (ATC J01CA04), and the beta-lactamase inhibitor combination 218 amoxicillin-clavulanic acid (J01CR02). The objective of this document is to review available information 219 on the use of these substances in veterinary and in human medicine in the EU, the influence that 220 veterinary use in particular has on the emergence of AMR and its potential impact on human and 221 animal health. The document provides information for risk profiling, as recommended by the 222 Antimicrobial Advice ad hoc Expert Group (AMEG), which will allow these substances to be placed 223 within the AMEG’s categorisation. The AMEG is currently reviewing the criteria for its categorization and 224 could give consideration to its further stratification. Reflection paper on the use of aminopenicillins and their beta-lactamase inhibitor combinations in animals in the European Union: development of resistance and impact on human and animal health EMA/CVMP/AWP/842786/2015 Page 7/65 2. General drug characteristics 225 226 2.1. Structure and mechanism of action 227 Ampicillin, amino-p-hydroxy-benzyl penicillin, was the first semisynthetic penicillin introduced into 228 clinical use in 1961 by Beecham Laboratories. It was followed by amoxicillin in the early 1970’s 229 (Rolinson, 1998). Amoxicillin has an otherwise identical structure to ampicillin, except for an additional 230 hydroxyl group attached to a phenyl ring of the side chain. Discovery of the active moiety, 6- 231 aminopenicillanic acid nucleus (6-APA), from the penicillin molecule enabled the development of 232 semisynthetic penicillins with enhanced spectrum of activity for Gram-negative bacteria. In 6-APA, a 233 beta-lactam ring is attached to a thiazole ring. The structure of the side chain linked to the amino 234 group of the 6-APA determines the pharmacokinetic properties and antimicrobial activity of the drug 235 (Rolinson, 1998). 236 Aminopenicillins inhibit the activity of the transpeptidase and other peptidoglycan-active enzymes that 237 catalyse the cross-linking of the glycopeptide units in the bacterial cell wall. Target enzymes are called 238 penicillin binding proteins (PBPs). Aminopenicillins bind to PBPs by mimicking the structure of the 239 natural substrate (D-alanyl-D-alanine) of the enzymes. This leads to incomplete cross-linking of 240 peptidoglycan building blocks and induces osmotic lysis of the bacterial cell due to loss of rigidity of the 241 peptidoglycan layer. The action is bactericidal, but affects only actively dividing bacterial cells (Giguère 242 et al., 2013). The composition of PBPs in the bacterial species in question partly explains the spectrum 243 of different beta-lactams, for example, enterococci are naturally susceptible to aminopenicillins but not 244 to cephalosporins (Kristich and Little, 2012). 245 Due to the emergence of beta-lactamase mediated resistance that impaired the efficacy of 246 aminopenicillins, the search for beta-lactamase inhibitors started in the late 1960s (Rolinson, 1998). 247 Clavulanic acid is a beta-lactamase inhibitor with a beta-lactam-like structure. It is produced by 248 Streptomyces clavuligerous (Brown et al., 1976). Clavulanic acid and other beta-lactamase inhibitors 249 with a beta-lactam core, such as sulbactam and tazobactam, have only a weak antimicrobial activity of 250 their own. In combination products, a beta-lactamase inhibitor binds irreversibly to bacterial beta- 251 lactamases blocking their activity, while the actual beta-lactam component maintains its activity 252 against bacteria. Amoxicillin-clavulanic acid was the first beta-lactam - beta-lactamase inhibitor 253 combination coming into the market in 1981 (Bush, 1988). Clavulanic acid binds covalently to several 254 bacterial beta-lactamases including type II, III, IV and V beta-lactamases, as well as staphylococcal 255 penicillinases, but it is ineffective against class I cephalosporinases (AmpC type) and carbapenemases 256 (Drawz and Bonomo, 2010). In veterinary therapeutic products amoxicillin is combined with clavulanic 257 acid usually in a 4:1 ratio. There are no other beta-lactam beta-lactamase inhibitor combinations 258 authorized in veterinary medicine in the EU. 259 2.2. Antimicrobial spectrum 260 The antimicrobial spectrum of ampicillin and amoxicillin against Gram-positive bacteria covers, among 261 others, the following Gram-positive genera: Staphylococcus, Streptococcus, Enterococcus, Listeria, 262 Actinomyces, Trueperella, Corynebacterium, and Erysipelothrix. Compared to natural penicillins, 263 aminopenicillins are more hydrophilic and thus are able to diffuse better through the outer membrane 264 of the Gram-negative bacteria. Of Gram-negative genera, Haemophilus, Histophilus, Pasteurella, 265 Mannheimia, Actinobacillus, Neisseria, Moraxella, Borrelia, and Leptospira are usually susceptible. Of 266 the Enterobacteriaceae, Escherichia coli, Proteus mirabilis, and Salmonella species are susceptible Reflection paper on the use of aminopenicillins and their beta-lactamase inhibitor combinations in animals in the European Union: development of resistance and impact on human and animal health EMA/CVMP/AWP/842786/2015 Page 8/65 267 unless they have acquired resistance mechanisms. Susceptible anaerobes include, among others, 268 anaerobic Gram-positive cocci, Clostridium spp., Fusobacterium spp., Prevotella spp. and 269 Porphyromonas spp. 270 Ampicillin and amoxicillin are ineffective against Klebsiella spp., Enterobacter spp., Citrobacter spp., 271 Serratia spp., indole-positive Proteus spp., Acinetobacter spp. and Pseudomonas spp. due to intrinsic 272 resistance mechanisms in these species. Also Bordetella ssp., rickettsia, mycoplasma and mycobacteria 273 are resistant (Giguère et al., 2013). 274 Staphylococcal penicillinases and beta-lactamases produced by Gram-negative bacteria inactivate 275 ampicillin and amoxicillin. Thus aminopenicillins are often combined with a beta-lactam inhibitor or 276 replaced by cephalosporin group antimicrobials. In the EU, the only veterinary authorized inhibitor 277 combination is amoxicillin clavulanic-acid. It has a spectrum of activity corresponding to that of 2nd- 278 generation cephalosporins and covers also Klebsiella spp., Bordetella spp., Bacteroides spp. and indole 279 positive Proteus spp. (Giguère et al., 2013). 280 There is great variation in relative susceptibility to aminopenicillins between bacterial genera. The wild 281 type Streptococcus spp., Actinomyces spp., Clostridium perfringens, Listeria spp., Haemophilus spp., 282 Histophilus spp., Moraxella spp., and Pasteurella spp, have the lowest minimal inhibitory 283 concentrations (MICs), ≤ 1 mg/L. MICs for the wild type Enterococcus spp. range from 0.25 to 4 mg/L, 284 while the wild type E. coli isolates have relatively high MICs, 1 - 8 mg/L, both for ampicillin and 285 amoxicillin clavulanic acid. The same applies to Salmonella Enteritidis, while other salmonellae are 286 slightly more susceptible (https://mic.eucast.org/Eucast2/). Klebsiella species are intrinsically resistant 287 to ampicillin or amoxicillin (MICs ≥ 4 mg/L), but when amoxicillin is combined with clavulanic acid, the 288 MICs of the main population range from 1 - 8 mg/L (www.eucast.org). 289 2.3. Pharmacodynamics 290 Ampicillin and amoxicillin are bactericidal and their effect is time-dependent. Optimal killing occurs if 291 bacteria are exposed to an antimicrobial concentration exceeding 1 - 4 x the MIC for sufficient time 292 between the dosing intervals. Thus, for time dependent drugs, a time above the MIC (T>MIC) is the 293 best pharmacokinetic/pharmacodynamic (PK/PD) parameter predicting microbiological and clinical 294 efficacy. For beta-lactams, the target T>MIC is 50 – 80% of the dosing interval (Toutain et al., 2002). 295 Beta-lactams possess significant post-antibiotic effect (PAE) against Staphylococcus aureus, 296 Streptococcus pneumoniae and Enterococcus faecalis, although the length of the PAE ranges widely, 297 between 0.5 - 6 hrs (Preston and Drusano, 1999). Gram-negative bacteria show no considerable PAE 298 effect after exposure to ampicillin or amoxicillin (Brown et al., 1976). Therefore, for infections caused 299 by Gram-negative bacteria, a shorter dosing interval is recommended compared to infections caused 300 by Gram-positive bacteria (Toutain et al., 2002). Aminopenicillins penetrate poorly into phagocytes and 301 hence have limited ability to kill intracellular pathogens like Salmonella spp. (Mandel and Petri Jr, 302 1996). 303 Although the antimicrobial spectrums of ampicillin and amoxicillin are nearly identical, an early study 304 proved that at concentrations close to the MIC, ampicillin shows a slower killing rate in-vitro than 305 amoxicillin against E. coli and Salmonella Typhi due to slower lysis of bacterial cells (Basker et al., 306 1979). The same has been observed in vivo in mice an experimental intra-peritoneal infection model in 307 which amoxicillin was observed to be more effective than ampicillin in protecting the mice from the 308 lethal effects of the E. coli infection - regardless that concentrations of both compounds in the body 309 fluids were equal (Comber et al., 1977). Amoxicillin induced the formation of rapidly lysing spheroplast Reflection paper on the use of aminopenicillins and their beta-lactamase inhibitor combinations in animals in the European Union: development of resistance and impact on human and animal health EMA/CVMP/AWP/842786/2015 Page 9/65 310 forms of the bacterial cell while ampicillin resulted in slowly lysing long bacterial filaments (Comber et 311 al., 1977). 312 Paradoxically, increasing the concentration of beta-lactam antimicrobials above the optimal killing 313 concentration can lead to impaired killing of bacteria. This is known as the Eagle effect, and is 314 sometimes observed in vitro with beta-lactams against Gram-positive cocci and rods (Grandière-Pérez 315 et al., 2005; SHAH, 1982). The effect is probably due to binding of a beta-lactam to other than primary 316 target PBPs, so preventing bacterial cell wall synthesis and multiplication, while beta-lactams are active 317 only against actively dividing cells. The clinical impact of this phenomenon is unclear (Lamb et al., 318 2015). 319 2.4. Pharmacokinetics 320 Although amoxicillin and ampicillin are closely related in their structure as well as in chemical and 321 physical properties, the extent of absorption after oral dosing differs markedly between these 322 molecules. Generally speaking, the amoxicillin serum drug concentration is twice that of ampicillin with 323 the same dose. The speed of bactericidal action of amoxicillin is more rapid and complete compared to 324 ampicillin when administered at the same dose (Prescott, 2013). In monogastric animals 33 – 92% of 325 the dose is absorbed after oral administration of amoxicillin. The comparative figure for ampicillin is 326 30-55%. The absorption of amoxicillin is unaffected by feeding in pigs (Agersø and Friis, 1998a), dogs 327 and humans, unlike ampicillin (Watson and Egerton, 1977). Aminopenicillins cannot be administered 328 orally for adult ruminants, horses or animal species [such as rabbits] that are prone to severe 329 disturbance of their gut microbiota. The volume of distribution is 0.2 - 0.3 L/kg depending on species. 330 The drug is distributed widely in the extracellular fluids of many tissues including lungs, muscle, bile, 331 peritoneal and pleural fluid, and synovial fluid. If the meninges are inflamed, therapeutic drug 332 concentrations may be achieved in the cerebrospinal fluid. In milk the concentration is low, 333 approximately one fifth of that in serum. Protein binding varies between 8 – 20% depending on animal 334 species. The elimination half-life is 45 - 90 min, being longest in cattle, although it can be prolonged by 335 the use of sustained release drug formulations. Elimination occurs through renal excretion mainly as 336 active drug (Prescott, 2013). The pharmacokinetics of clavulanic acid resembles that of amoxicillin. 337 Clavulanic acid is readily absorbed after oral administration. It is widely distributed into extracellular 338 fluids, but poorly into milk or inflamed cerebrospinal fluid. Its half-life is approximately 1.25 hrs and it 339 is excreted primarily in urine as unchanged drug (Prescott, 2013). Achievable drug concentrations after 340 various formulations and dosages in different animal species are summarised below and are presented 341 as mg/L (instead of μg/ml) in order to facilitate comparison of bacterial susceptibilities in relation to 342 achievable drug concentrations in-vivo. 343 Pigs: An intra-muscular (i.m.) dose of 10 mg/kg ampicillin sodium resulted C of 12 mg/L and 14 max 344 mg/L in plasma of healthy and Streptococcus suum [Streptococcus suis] infected 2-month-old pigs, 345 respectively. The half-life was shorter in the latter group (0.76 vs. 0.57 h) (Yuan et al., 1997). In 346 three-week old piglets a peak plasma concentration of 7 mg/L was observed after i.m. injection of 17.6 347 mg ampicillin trihydrate /kg (Apley et al., 2007). With the oral dose of 20mg/kg ampicillin, a high drug 348 concentration (720 mg/L) in caecal fluid was achieved, while twice that dose intramuscularly resulted 349 in a concentration of only 15 mg/ml (Escoula et al., 1982). A conventional amoxicillin-trihydrate 350 formulation, dosed at 14.7 mg/kg i.m. produced a peak concentration of 5.1 mg/L but with a 351 sustainable release (LA) formulation (dose 14.1 mg/kg), the peak concentration was only 1.7 mg/L. 352 Oral administration of amoxicillin produced very low peak plasma concentrations in pigs, ranging from 353 0.2 to 3.1 mg/L depending on dose (10-23 mg/kg), and on whether the drug was given as oral bolus, Reflection paper on the use of aminopenicillins and their beta-lactamase inhibitor combinations in animals in the European Union: development of resistance and impact on human and animal health EMA/CVMP/AWP/842786/2015 Page 10/65