Bacterial Pneumonia and Emerging Antibiotic Resistance ECCMID 2012 Curetis Symposia Proceedings April 1, 2012 London, UK

Pneumonia is a major problem Jean-Louis Vincent Pneumonia-causing pathogens and their resistances David Livermore Challenges and opportunities in testing respiratory tract infections Ingo Autenrieth, Berit Schulte Would faster molecular testing make a difference in the current standard of care? A Microbiologist’s point of view Christine Ginocchio Would faster molecular testing make a difference in the current standard of care? A Clinician’s point of view Antoni Torres Health Economic Modeling of the impact of fast pneumonia testing Anne Thews

Preface

Curetis Symposium ECCMID London 2012

Preface At ECCMID 2012 held in London April 2012 Curetis AG sponsored an Integrated Symposium focusing on Bacterial Pneumonia and Emerging Antibiotic Resistance The symposium was shared by two leading scientists in the infectious disease area: Prof. Dr. Christine Ginocchio, New York and Prof. Dr. Keith Klugman, Atlanta. The workshop was intended to support Curetis AG`s first product release – the Unyvero™ P50 Pneumonia Application using the Unyvero™ System. Five internationally re-owned experts: • • • • •

Prof. Dr. Jean-Louis Vincent, Brussels Prof. David Livermore, Norwich Prof. Ingo Autenrieth, Tübingen Prof. Christine Ginocchio, New York Prof. Antoni Torres, Barcelona

gave presentations and discussed the latest information on pneumonia. Their talks regarding this severe acute infection raised concerns about growing antibiotic resistance as a major burden for today‘s health care systems. In addition, they discussed whether such lung infections would benefit from faster and more comprehensive diagnostics.

All experts agreed about the high demand for faster results in pneumonia testing. Such quick results are a prerequisite for giving adequate antibiotic treatment as early as possible in order to improve the standard of care. These experts have agreed to contribute to this ECCMID Symposium Proceeding in order to capture their high profile presentations. The management of Curetis would like to express our deep thanks to all of the authors. We have added an additional chapter about an economic model Curetis AG has developed in cooperation with Halteres Associates, an highly recognized IVD industry consulting firm, to demonstrate the potential medical and economic benefits of fast molecular testing.

Sincerely yours

Anne Thews Head of Medical Affairs Curetis AG

The talks clearly indicated the need to balance antibiotic treatment to cure patients but to limit treatment with regard to antibiotic steward ship as illustrated in Figure 1.

Figure 1 The balance of antibiotic treatment

03

Content

Curetis Symposium ECCMID London 2012

Content

UNYVERO™ SYSTEM -2-

-1-

-3-

-4-

The CE-marked Unyvero™ System is a versatile hardware platform for the detection of a broad panel of bacteria and antibiotic resistances from a single sample in one run. It processes a disposable cartridge providing the necessary reagents to complete the analysis from sample to result. The platform enables the DNA-based testing of all clinically relevant samples in a fully automated, unsupervised analysis process requiring only few, quick manual preparation steps. The analysis can be performed with 1. Unyvero™ A50 Analyzer: Universal Analyzer

minimal operator time and without the need of skilled staff or special infrastructure. As a result, clinically relevant information is available within

2. Unyvero™ C8 Cockpit:

about four hours to support an informed therapy decision as early as pos-

Intuitive User Cockpit

sible. The first CE-marked Unyvero™ Cartridge, Unyvero™ P50, focu-

3. Unyvero™ L4 Lysator:

ses on pneumonia testing and simultaneously analyzes 39 DNA targets.

Universal Lysator 4. Unyvero™ Cartridge:

Caution – Investigational Device, Limited by Federal (or US) Law to

Disposable for different

Investigational Use. 5

clinical applications

Not available for sale in the United States.

04

Pneumonia is a major problem 06 Jean-Louis Vincent There is an increasing incidence of gram-negative and mixed lung infections in acutely ill patients as well as a growing percentage of multidrug resistant bacteria. Results were obtained from two very large international studies (SOAP, EPIC II) that collected data from critically ill patients in the intensive care unit. Data collected both on the incidence as well as the type of infections show a similar set of findings and outcomes, namely that unsuccessful management of severe infections is largely caused by the long delay in determining the identity of the pathogen and its antibiotic sensitivity. Pneumonia-causing pathogens 12 and their resistances David Livermore Levels of antibiotic resistance are rising around the world and threatening the future availability of useful antibiotics for treating serious infections such as pneumonia. Antibiotic resistance occurs by different mechanisms in different pathogens. The presence of resistance with poorly defined mechanisms presents a huge detection challenge. In this article, the temporal spread of antibiotic resistance throughout Europe and the rest of the world will be discussed.

Challenges in testing respiratory 18 tract infections Ingo Autenrieth, Berit Schulte Due to the increasing prevalence of antibiotic resistant bacteria, selection of appropriate antibiotic therapy for different types of pneumonia has become much more challenging. Today, pathogen identification and antibiotic sensitivity are based on time-consuming culture-based microbiology technology. New molecular diagnostic technologies are being developed that can rapidly identify the pathogen and determine antibiotic sensitivity in hours. The Curetis pneumonia panel is highly specific and has excellent sensitivity for detecting pathogens. The correlation between genotype and phenotype is reproducibly between eighty and one hundred percent. Moreover, genotype resistance testing may provide valuable results to optimize early antibiotic therapy for the treatment of pneumonia.

Would faster molecular testing make a 26 difference in the current standard of care? A Microbiologist’s point of view Christine Ginocchio Microbiology culture technology is currently the standard of care for determining treatment of pneumonia. But traditional microbiology methods are slow and can delay treatment of patients with the appropriate targeted antibiotic therapy. With rising antibiotic resistance, there is a need to develop new technologies that can carry out pathogen identification and antibiotic resistance testing within hours, not days. With new techniques currently being developed, this goal can be realized.

Would faster molecular testing make a 32 difference in the current standard of care? A Clinician’s point of view Antoni Torres Under the current paradigm for treatment of pneumonia, the clinician must wait up to forty-eight to seventytwo hours to identify the causative microorganisms using standard microbiological culture techniques. However, treatment is required immediately and so the clinician must make potentially life-saving decisions without the benefit of knowing the responsible pathogen(s). Routine availability of rapid accurate, molecular diagnostic tests might result in fewer inappropriate and inadequate antibiotic treatments.

Health Economic Modeling of the 38 impact of fast pneumonia testing Anne Thews Unyvero™ Solution is a new multiplexed molecular assay targeting pneumonia patients. A Health Economic (HE) model for application of this assay is discussed. Based on data published in peer-reviewed journals, the model studies the impact of this test in ventilatorassociated pneumonia (VAP). The model compares the current standard of care, microbiological culture, to the new multiplexed assay. It shows that Unyvero™ Solution offers significant cost savings as well as a gain in quality-adjusted life years (QALY) for a typical VAP patient.

05

Pneumonia is a major problem

Curetis Symposium ECCMID London 2012

Jean-Louis Vincent Université Libre de Bruxelles, Belgium

Introduction

Abstract Lung infections are a grave problem in acutely ill patients. Moreover, there is an increasing incidence of gram-negative and mixed infections as well as a growing percentage of multidrug resistant bacteria in acutely ill patients. The impact of these trends will be discussed as they relate to the management of severe sepsis and different types of pneumonia such as hospitaland community-acquired pneumonia. Results from two very large international studies that collected data from critically ill patients in the intensive care unit on the incidence and type of lung infections will be summarized here. The results from both studies show a similar set of findings and outcomes.

The respiratory tract is the most common source of infection in acutely ill patients and is one of the leading causes of death in these patients. Pneumonia can occur inside the hospital or outside in the community. Greater incidence of both types of pneumonia is associated with increasing antimicrobial resistance, which can both result in higher rates of morbidity and mortality as well as raise the economic burden of treating pneumonia. An antibiotic therapy should be given as early as possible after an initial pneumonia diagnosis (hit hard and early3). The therapeutic regimen chosen needs to take into account local antibiotic resistance patterns as well as a host of other factors and variables. Data from two large international studies conducted in patients in the intensive care unit have greatly contributed to our understanding of the growing incidence of gram-negative and mixed infections as well as the growing threat of multi-drug resistant organisms around the world.

cent, respectively, in the SOAP and EPIC II studies.

Pneumonia most frequent infection

Keywords SOAP, EPIC II, sepsis, hospital acquired pneumonia, ventilator acquired pneumonia

Two large international studies were conducted to collect data on the incidence of infection and characteristics of critically ill patients in the intensive care unit (ICU). The first of these studies, “Sepsis Occurrence in Acutely ill Patients” (SOAP) was a multiple-center observational study of over three thousand patients admitted to ICUs across twenty-four European countries.1 The study was conducted over a two-week period from May 1 to May 15, 2002. The second study entitled, “The Extended Prevalence of Infection in Intensive Care,” (EPIC II) was designed as a one-day prospective, point prevalent study with follow-up conducted on a single day, May 8, 2007.2 It was a follow-on studyfrom the original EPIC study published in 1995.4 The EPIC II study included over fourteen thousand ICU patients treated worldwide across seventy-five participating countries. Table 1 summarizes the major clinical and microbiological findings in the SOAP and EPIC II studies. As Table 1 shows, in the SOAP study, twenty-five percent of patients were initially diagnosed with sepsis upon admission to the ICU and this number increased to thirty-seven percent over the two week study period due to patients succumbing to sepsis while in the ICU. In the EPIC II study, over seven thousand patients or fifty-one percent of the total number in the study were diagnosed with an infection. Nearly eighty-five percent of the infected patients were diagnosed with sepsis. The frequency of lung infections ranged from sixty-eight to sixty-four per

Table 1 Clinical and Microbiology Findings SOAP and EPICS II Studies

The next most frequent infections were infections in the abdomen and bloodstream. Lung infection was the biggest problem in patients from all countries who participated in EPIC II. The mean value for lung infection globally was sixty-three percent with the highest and lowest rates seen in Russia and Africa, respectively (seventy-one and forty-seven percent). The frequency of infections in the abdomen was fairly constant with an average of nineteen percent and a range from seventeen to twenty-one percent. See centerfold.

Figure 1 Infection in ICU

07

Curetis Symposium ECCMID London 2012

Pneumonia is a major problem

Infection rate correlates with length of time in ICU The length of time patients were in the ICU prior to the study day was shown to correlate directly with increasing rates of infection.5 As illustrated in Figure 1 there is a linear correlation between the infection rate, which increased from thirty to over seventy percent, with length of time in the ICU from one day to up to seven to ten days. Infection rates reached a plateau between seventy and eighty percent in patients cared for in the ICU longer than seven to ten days.

Gram-negative infections are increasing worldwide Characterizing the microorganisms in infected ICU patients was a vital component of the EPIC II study.6 With over seven thousand infected ICU patients in seventy-five countries, these data provided valuable information about the geographical distribution of pathogens as well as their patterns of microbial resistance on a worldwide scale. The global distribution of gram-positive and gram-negative organisms is shown in the graphic in the centerfold. In all continents of the world, with the exception of North America, the mean frequency of infections with gram-negative organisms was higher than for gram-positive organisms, with a worldwide mean value of forty-seven percent gram-positive and sixty two percent gramnegative organisms. The remainder of the infections included mixed infections, which are quite common (twenty-three percent in SOAP) and fungal infections (seventeen to nineteen percent in both studies).

patients (twenty percent), followed by Klebsiella (thirteen percent) and Acinetobacter (nine percent). The global distribution of these major gram-negative pathogens can be seen in the graphic in the centerfold. The frequency of gram-negative organisms was particularly high in Central and South America as well as Eastern Europe, as both regions reported elevated frequencies of gram-negative organisms compared with the mean values worldwide. The high values for Pseudomonas infections are troubling because Pseudomonas is difficult to treat and is associated with increased mortality in ICU patients. When length of time in the ICU was compared with the infection rate and associated pathogens, Pseudomonas infections accounted for the greatest percentage of infections in patients treated in the ICU for longer than four to six days.2

08

•

•

Give the appropriate antibiotic or antibiotic combination, which is effective against the infectious pathogens. Start adequate antibiotic therapy as early as possible at the right dose, via the correct route of administration to maximize exposure of the antibiotic at the infection site.

The importance of providing appropriate and adequate antibiotic therapy is a vital component of successful treatment for severe infections as numerous published studies have demonstrated that inadequate antibiotic therapy increases the mortality rate.4-16 These studies clearly demonstrate that if inappropriate and/or inadequate antibiotic therapy is administered to patients with a variety of severe infections, the mortality rate and mean duration of hospital stay is significantly increased. For example, in patients with community acquired blood stream infections, giving an inappropriate empiric therapy is a predictor of mortality in the patients. Patients in septic shock who received inappropriate treatment have a survival rate of less than twenty percent. In a study published by Fraser in 2006, in 920 patients with sepsis, thirty-six percent of patients given inappropriate therapy had greater than twenty percent mortality, whereas sixty-four percent of patients receiving appropriate therapy had only a twelve percent mortality rate.

Challenges in CAP Treatment Figure 2 Rules for therapy

Another important finding from the SOAP and EPIC II studies is that gram-negative organisms are increasing in frequency. While the frequency of gram-positive organisms was similar in the two studies (forty percent in SOAP, forty-seven percent in EPIC II), the frequency of immune organisms rose from thirty-eight percent in SOAP to sixty-two percent in EPIC II. This is a concerning trend because the frequency of multi-drug resistant (MDR) gram-negative organisms is also increasing. Staphylococcus species including S.aureus and S. epidermidis were the most abundant gram-positive organisms in infected ICU patients. For S. aureus, the frequency of MRSA ranged from a high of twenty-one percent in Africa to a low of nine percent in Western Europe. These data are shown in the graphic in the centerfold. In the case of gram-negative infections, Pseudomonas was the most frequent organism found in infected ICU

The rules of choosing an antibiotic therapy are severalfold and include:

Rules for antimicrobial therapy Several approaches are typically applied in parallel when treating patients with severe infections, e.g. sepsis. These could include infection control, hemodynamic stabilization and modulation of the immune response as shown is Figure 2. In the case of infection control, early and effective antibiotic therapy is the key to successful management. Without the knowledge of the causative pathogens, initial antibiotic therapy is empiric and consists of a broad-spectrum antibiotic or a combination of antibiotics. The knowledge of important factors such as prior antibiotic therapy and local antibiotic resistance patterns are important for establishing sufficient empiric treatment schemes.

Community Acquired Pneumonia (CAP) is pneumonia that originates outside of the hospital in the community. Treatment guidelines for CAP are based on what we know about antibiotic resistance come from studies of hospital acquired infections and thus we might underestimate or miss the presence of MDR in the community. It is hard to define treatment for MDR in the community, as treatments are hospital-specific. This means that up to thirty percent of empiric therapies are wrong. Drug resistant organisms occur in CAP, albeit at a lower level than in nosocomial pneumonia infections. Similar to the treatment of other severe infections, early antibiotic therapy is associated with a better outcome in treating CAP. There have been numerous publications that demonstrate giving adequate and appropriate antibiotic therapy early in the course of treatment is beneficial.17-22 However, besides the growing

threat of multi-drug resistant bacteria, CAP is also characterized by the presence of mixed or polymicrobial infections, which is a risk factor for inappropriate initial antibiotic therapy and is associated with increased mortality in this patient population.23

VAP – poor medical outcome Ventilator Acquired Pneumonia (VAP) is a lung infection that occurs in patients who are infected in the hospital ICU during mechanical ventilation for breathing difficulties. The data from numerous studies confirm a similar picture described above for patients with CAP, namely that inadequate antibiotic therapy increases mortality rates.24-35 In several studies, where inappropriate antibiotic therapy was given, the mortality rate was greater than fifty percent.24, 29-31 This number dropped to thirty three percent when appropriate antibiotic therapy was administered, which was also associated with a shorter duration of mechanical ventilation and fewer days in the ICU.28,33 Inappropriate therapy is linked to the presence of resistant organisms such as Pseudomonas aeruginosa, Staphylococcus aureus and Acinetobacter. Ventilator acquired pneumonia can be further characterized by the length of invasive mechanical ventilation. Early-onset VAP is defined as onset of pneumonia within four days or less after mechanical ventilation. Late-onset VAP is characterized by onset of symptoms after five or more days of mechanical ventilation. The etiology of early- and late-onset VAP differs. Typical community organisms are more frequent in early-onset VAP, whereas opportunistic and antibioticresistant pathogens, such as MRSA and Pseudomonas aeruginosa are more common in late-onset VAP. Other risk factors for VAP that increase the likelihood of being infected with drug resistant pathogens include:4,6 • • • • • • • •

Hospitalization for at least five days Hospitalization in an acute care facility or a nursing home within the last three months Close contact with someone known to be infected or colonized with an MDR pathogen Antibiotic therapy within the last thirty days Presence of an IV catheter for several days Chronic hemodialysis Prolonged wound care treatment Immunosuppression

09

Curetis Symposium ECCMID London 2012

Pneumonia is a major problem

ted based on the microbiological findings. Since microbiology culture techniques are very slow, in most cases these data is provided too late in the decision process.

Jean-Louis Vincent is Professor of intensive care at the Université Libre de Bruxelles and Head of the Department of Intensive Care at the Erasme University Hospital in Brussels. He is presently Secretary General of the World Federation of Societies of Intensive and Critical Care Medicine, a Past-President of the European Society of Intensive Care Medicine, the European Shock Society, and the International Sepsis Forum. He received several awards (Society of Critical Care Medicine, American College of Chest Physicians, European Society of Intensive Care Medicine and Belgian FRS-FNRS).

The additional use of molecular diagnostic technologies used in parallel to identify the pathogen rapidly could assist in the decision making process to select the appropriate antibiotic therapy and thus may improve medical outcome.

Correspondence Prof. Dr. Jean-Louis Vincent Professor of Intensive Care Medicine ULB Hôpital Erasme Route de Lennik 808 1070 Bruxelles, Belgium

Conclusions

Figure 3 Survival in Pneumonia

Figure 3 is an illustration showing how overall survival in community and hospital acquired pneumonia is affected by the type of pathogen (sensitive or resistant to antibiotics), the site of infection (community or hospital acquired) and concomitant risk factors that vary according to other diseases the patients may have. Not surprisingly, the chances of survival are the lowest in HAP patients infected with drug resistant bacteria and patients with CAP have the highest probability of survival.

Guiding antibiotic therapy Typical clinical measurements used to diagnose a patient with severe infection include: • •

•

Evidence of signs and symptoms such as increased fever, increased white blood cell count etc. Evidence of lung infection as determined by a clinical examination, a positive X-ray and the presence of purulent sputum Microbiological findings such as tracheal aspirate and bronchiolavage samples.

Pneumonia is a real medical challenge. The results of large multicenter studies delivered a clearer picture of the severity of the problem on a worldwide scale. A confounding issue in the successful management of pneumonia is the long delay in determining the identity of the pathogen and its antibiotic sensitivity. Current guidelines for the treatment of pneumonia are based on the standard of care, which delays the identification of the pathogen by two to three days due to the time it takes to culture the organism. One way to reverse the current upward trend in the spread of MDR and to control the growing incidence of gram negative- and mixed infections is to implement new diagnostic technologies to speed up the process of pathogen and resistance identification. Rapid, accurate tests are needed which enable clinicians to identify microorganisms and determine antibiotic sensitivity quickly. The additional use of molecular diagnostic technologies used in parallel to identify the pathogen rapidly could assist in the decision making process to select the appropriate antibiotic therapy and thus may improve medical outcome.

1

Vincent J-L, Sakr Y, Sprung CL, et al. 2006. Crit. Care Med.

29

Rello et al., Am J Respir Crit Care Med 156: 196-200, 1997

34:344-353.

30

Luna et al., Chest 111: 676-85, 1997

Vincent J-L, Rello J, Marshall J, et al. 2009. JAMA. 302:2323-2329.

31

Kollef & Ward, Chest 113: 412-20, 1998

Quotation Paul Ehrlich, 1913.

32

Heyland et al., Am J Respir Crit Care Med 159:1249-56, 1999

Vincent J-L, Bihart D, Suter PM. et al. 1995. JAMA. 274:639-644.

33

Dupont et al., Intensive Care Med 27: 355-60,2001

Celis et al., Chest 93: 318-24, 1988

34

Iregui et al., Chest 122: 262-8, 2002

Torres et al., Am Rev Respir Dis 142: 523-8, 1990

35

Rello et al., Crit Care Med 32: 2183-90, 2004

Kollef, JAMA 270: 1965-9, 1993

36

Vincent JL et al., Drugs 70: 1927-1944, 2010

Kollef et al., Ann Intern Med 122: 743-9, 1995

37

Micek et al., Antimicrob. Agents Chemoth. 54:1742, 2010

Alvarez-Lerma et al, Intensive Care Med 22: 387 94, 1996

38

Kumar et al., Crit. Care Med. 38: 1651 and 1773,2010

2 3 4 4 5 6 7 8

Rello et al., Am J Respir Crit Care Med 156: 196-200, 1997

9

Luna et al, Chest 111: 676-85, 1997

10

Kollef & Ward, Chest 113: 412-20, 1998

11

Heyland et al., Am J Respir Crit Care Med 159:1249-56, 1999

12

Ibrahim et al., Chest 118: 146-55, 2000

13

Dupont et al., Intensive Care Med 27: 355-60, 2001

14

Iregui et al., Chest 122: 262-8, 2002

15

The use of antibiotics is based on the presence and severity of the symptoms listed above. If a patient is diagnosed with severe infections, combination antibiotic therapy is the preferred treatment strategy.37, 38 Combination antibiotic therapies increases the chances of being effective and when given early, avoids the delay in adding another antibiotic later. This empiric antibiotic therapy with wide coverage is typically given in generous doses at the outset, however with the growing incidence of drug-resistance it might fail and support the development of new resistances. Thus, once the bacteriological data are available after two to three days, the antibiotic spectrum should be narrowed and adjus10

Rello et al., Crit Care Med 32: 2183-90, 2004

16

Kahn et al., JAMA 264: 1969-73, 1990

17

McGarvey and Harper, QRB 19: 124-30, 1993

18

Meehan et al, JAMA 278: 2080-4, 1997

19

Gleason et al., Arch Intern Med 159: 2562-7, 1999

20

Rello et al., Intensive Care Med 28: 1030-5, 2002

21

Houck et al., Arch Intern Med 164: 637-44, 2004

22

Cilloniz C et al., Crit. Care 15:R209, 2011

23

Celis et al., Chest 93: 318-24, 1988

24

Torres et al., Am Rev Respir Dis 142: 523-8, 1990

25

Kollef, JAMA 270: 1965-9, 1993

26

Kollef et al., Ann Intern Med 122: 743-9, 1995

27

Alvarez-Lerma et al., Intensive Care Med 22: 387-94, 1996

28

11

Pneumonia-causing pathogens and their resistances

Curetis Symposium ECCMID London 2012

David M Livermore Norwich Medical School, University of East Anglia, Norwich, UK

Introduction

Abstract The prevalence of antibiotic-resistant bacteria is rising around the world. Given the lack of new antibiotics, this development is threatening our capacity to treat serious infections such as pneumonia. In this article, the temporal trends of resistance among pneumonia pathogens throughout Europe and the rest of the world are discussed, and the major resistance mechanisms of these pathogens are outlined. Keywords Antibiotic resistance, resistance mechanisms, pneumonia

Pneumonia is categorised according to the location of onset. Community-acquired pneumonia (CAP) and hospital-acquired pneumonia (HAP) are the major, well-defined types of pneumonia and HAP can be further divided into ventilator–acquired (VAP) and non-ventilator acquired pneumonia. The major pathogens of CAP are Streptococcus pneumoniae and Haemophilus influenzae, whilst staphylococci and Enterobacteriaceae are more important in HAP and VAP, with Pseudomonas aeruginosa and Acinetobacter baumannii becoming more likely where the patient has been hospitalized or ventilated for a prolonged period.

non-susceptible to penicillin or resistant to erythromycin increased from 2000 to 2010 in several European countries (see maps in the centerfold).1 Moreover, despite the growing recent availability of modern conjugate vaccines directed against the pneumococcal serotypes where resistance is most prevalent, there has been no dramatic, Europe-wide reduction in resistance from 2005 to 2010. Reductions nevertheless have been seen in particular regions or countries, for example in Scandinavia and the United Kingdom, where a previously prevalent serotype 14 macrolide-resistant strain has been substantially displaced since the deployment of the vaccine.4

Some authors recognise a third type of pneumonia, termed ‘Healthcare-associated pneumonia HCAP’, occurring in healthcare settings outside of the hospital, such as nursing homes and other long-term care facilities and with a microbial aetiology that overlaps both CAP and HAP.1 Others argue that the classical pathogens of CAP still dominate in long-term care facilities and that most ‘HCAP’ patients really have underlying problems such as chronic obstructive pulmonary disease or bronchiectasis, which anyway have a bacteriology closer to that of HAP.

Data from the Health Protection Agency (courtesy AP Johnson), covering the UK except Scotland, confirm the EARS-net finding that erythromycin resistance among bloodstream pneumococci decreased between 2004 and 2011 from approximately fourteen percent to four percent. But show that resistance increased in respiratory pneumococci from fourteen percent in 2007 (when surveillance began) to nearly eighteen percent in 2011. Such data suggest that vaccine deployment is affecting only invasive disease. Non-susceptibility to penicillin among pneumococci from blood and respiratory infections in the United Kingdom remained low and steady at around three- and six percent, respectively.

Antibiotic resistance is increasingly prevalent among important pneumonia pathogens in many parts of the world, and is a growing public health concern. Infections by resistant strains result in greater morbidity and mortality, as empirical therapy is more likely to prove ineffective, and increase the cost of treatment. As strains become more resistant to first- and secondline therapies, there is a growing need to replenish the supply of antibiotics, but antibiotic development has slowed in recent years, confounded both by low discovery rates and regulatory hurdles as well as by the financial reality that antibiotics are less profitable to the pharmaceutical industry than many long-term treatments.2

Antibiotic resistance trends in Europe: community pneumonia The main pathogens of community acquired pneumonia are S. pneumoniae and H. influenzae; other important agents are ‘atypicals’ such as Mycoplasma, Chlamydophilia and Legionella.

International data on resistance prevalence in H. influenzae are scantier than for S. pneumonia following the termination of previous international surveys such as the Alexander Project. Nevertheless local trends may be striking, and a multicenter surveillance in Spain showed that the prevalence of ampicillin resistance in H. influenzae dropped from approximately thirty-eight percent of isolates in 1996 to sixteen percent in 2007.5 The proportion of isolates with ß-lactamase decreased from over twenty-five percent to fifteen percent over the period whilst those with non ß-lactamase resistance almost disappeared, falling from thirteen percent to one percent.5 Such benign trends can, however, hide concerning shifts. In Spain Garcia-Cobos,6 found great diversity among circulating isolates with non-ß-lactamase-mediated ampicillin resistance, with some showing markedly reduced cefotaxime susceptibility. This diversity implies that H. influenzae has great potential to evolve and to adapt rapidly to confront ß-lactam challenges.

Results from the European Antimicrobial Resistance Surveillance network (EARS-Net)3 show that the proportion of S. pneumoniae bacteraemias due to strains 13

Pneumonia-causing pathogens and their resistances

Hospital pneumonia

Resistance mechanisms among CAP pathogens

The pathogens of hospital pneumonia are more diverse than those of community pneumonia, prominently including S. aureus, Enterobacteriaceae and – particularly in late onset disease - P. aeruginosa and A. baumannii.

If pathogen identification and resistance detection is to move from classical culture-based microbiology to molecular methods, as with the Unyvero™ system, it is necessary for these new methods to recognize prevalent resistance mechanisms. These are complex, variable, and can entail either mutation or acquisition of foreign DNA, with the importance of their mechanisms contingent on the species.

In the case of S. aureus, methicillin resistance has long been prevalent in most countries outside Scandinavia and the Netherlands. Although improved infection control has recently reduced the MRSA rate across much of Europe, these reductions are not universal and, even where they have been achieved, MRSA still remains a frequent pathogen. In the case of Enterobacteriaceae there have marked increases in resistance to cephalosporins and quinolones in the past decade,7 and some countries (including Greece, Cyprus, Hungary and Italy) are now also reporting substantial numbers Klebsiella (and few of other Enterobacteriaceae) isolates resistant even to carbapenems; elsewhere in Europe, Enterobacteriaceae with carbapenemases are increasingly becoming established, at least locally, including in pneumonia.8 Among relevant non-fermenters, P. aeruginosa shows very variable resistance rates across Europe, low in the north and west, and much higher in the south and east, independently of antibiotic class. This is illustrated on the maps in the centrefold, showing resistance rates in 2010 to aminoglycosides, carbapenems, ceftazidime and fluoroquinolones.1 A. baumannii has long been widely resistant to antibiotics except carbapenems and, since around year 2000, this pathogen has become substantially more resistant even to carbapenems,9 leaving only colistin and tigecycline as widely active in vitro – and neither of these is an ideal antibiotic for therapy in pneumonia.

Figure 1 Resistance mechanism among CAP pathogens

14

Mechanisms of antibiotic resistance in the pathogens of CAP are summarized In Table 1. In S. pneumoniae, resistance to macrolides and tetracycline depends largely on acquired genes, as does tetracycline and trimethoprim resistance in H. influenzae whereas, in both species, quinolone resistance is mutational. Resistance to ß-lactams involves a complex mixture of mutation and mosaic gene formation in S. pneumoniae,10, 11 as does low-level ampicillin resistance in H. influenzae. Highlevel ampicillin resistance in H. influenzae depends on acquisition of the blaTEM gene. In pneumonia (unlike meningitis) low-level resistance to ß-lactams in these species in can be overcome by increased dosage.

Resistance mechanisms among HAP pathogens The varied pathogens of HAP and VAP demonstrate a wide diversity of resistance mechanisms. In the case of MRSA, methicillin resistance is determined by the acquired (though rarely transferred) mecA gene, which is easily detected by PCR. On the other hand ‘resistance’ to vancomycin – the drug most often used in MRSA pneumonia – is far more complex. True vancomycin resistance, with MIC >2 mg/L, is extremely rare, but there are assertions, not universally confirmed, that MICs of 2 mg/L are associated with poorer outcomes than MICs of