The WHO has recognised antibiotic resistance as one of the greatest threats to human health. As a microbiologist, antibiotic resistance is a problem that keeps me awake at night and inevitably the impact of antibiotic resistance on paediatricians is a matter of when, and not if. I fear for the future of paediatric services such as neonatology, oncology and elective surgery. A recent US study found that 26.8% of post chemotherapy infections and 38.7–50.9% of post-operative infections were caused by bacteria resistant to standard antibiotic prophylaxis. The authors predicted that this will lead to an additional 6300 infection-related deaths in the USA each year. Closer to home, David Cameron commissioned a review into antimicrobial resistance in 2014 and the findings were extremely worrying. The report predicted that by 2050, 10 million annual worldwide deaths will be attributable to antimicrobial resistance. More than annual predicted cancer-associated and diabetes-associated mortality combined. The golden antibiotic era is certainly over. Selecting the most appropriate antibiotic to treat an infection depends on many factors, including route of administration, penetration to site of infection and pathogen susceptibility. Most clinicians do not need an in-depth understanding of bacterial resistance mechanisms as local microbiologists can provide expertise and advice. However, in an era of increasing antibiotic resistance, an insight into the organism factors that affect antibiotic selection can prove useful.
- Infectious Diseases
- Basic Science
Statistics from Altmetric.com
The WHO has recognised antibiotic resistance as one of the greatest threats to human health.1 As a microbiologist, antibiotic resistance is a problem that keeps me awake at night and inevitably the impact of antibiotic resistance on paediatricians is a matter of when, and not if.
I fear for the future of paediatric services such as neonatology, oncology and elective surgery. A recent US study2 found that 26.8% of post chemotherapy infections and 38.7–50.9% of postoperative infections were caused by bacteria resistant to standard antibiotic prophylaxis. The authors predicted that this will lead to an additional 6300 infection-related deaths in the USA each year. Closer to home, David Cameron commissioned a review into antimicrobial resistance in 2014 and the findings were extremely worrying. The report predicted that by 2050, 10 million annual worldwide deaths will be attributable to antimicrobial resistance.3 More than current annual predicted cancer-associated and diabetes-associated mortality combined. The golden antibiotic era is certainly over.
Selecting the most appropriate antibiotic to treat an infection depends on many factors, including route of administration, penetration to site of infection and pathogen susceptibility. Most clinicians do not need an in-depth understanding of bacterial resistance mechanisms as local microbiologists provide expertise and advice. However, in an era of increasing antibiotic resistance, an insight into the organism factors that affect antibiotic selection can prove useful. To highlight some important issues, we will use a fictional case study.
Imagine a 4-year-old boy with type 1 diabetes mellitus, involved in a road traffic accident while in Cyprus. After spending 2 weeks in a Cypriot hospital (undergoing orthopaedic procedures), he was transferred to a paediatric ward in the UK.
This boy received inpatient care in a country with high prevalence of healthcare-associated carbapenemase-producing Enterobacteriaceae (CPE), so he is at risk of CPE colonisation and infection. CPEs are Enterobacteriaceae (ie, Escherichia coli, Klebsiella spp, Enterobacter spp) that produce carbapenemase enzymes; these hydrolyse carbapenem antibiotics and confer resistance to β-lactam antibiotics (ie, penicillins, cephalosporins, monobactams and carbapenems).
Public Health England have published a national toolkit,4 and many hospital trusts produced their own local guidelines. Following guidance, he was admitted into a side room, appropriate personal protective equipment (PPE) was used and a stool sample was taken to screen for multidrug-resistant organisms. The toolkit provides guidance on risk assessment and management of patients colonised or infected with CPE to prevent spread of such bacteria within UK healthcare settings.
Within the laboratory, the stool was cultured to screen for the detection of CPE (figure 1).
Figure 1 shows growth on a selective agar; in this instance, the agar contains a modified carbapenem, ensuring reliable growth of only carbapenem-resistant organisms. Colonies appear blue due to the presence of a chromogenic indicator in the agar; they are identified as Klebsiella pneumoniae. Further testing was required to confirm if this was a CPE.
Later that day, a sample was taken from his surgical wound, as it was erythematous and discharging pus. The sample was inoculated onto non-selective agar plates and K. pneumoniae was cultured. Sensitivity testing was performed.
Figures 2 and 3 show the antibiotic susceptibility testing of this patient's K. pneumoniae wound isolate. This is a phenotypic method and is used as a proxy for the minimum inhibitory concentration (MIC) of antibiotics; MIC can be defined as ‘the lowest concentration of an antimicrobial that will inhibit the visible growth of a microorganism after overnight incubation’.5 The method allows an organism to be categorised as sensitive, intermediate or resistant. There is a defined concentration of antibiotic in each disc; this diffuses into the underlying agar and prevents growth of the test isolate if it is sensitive (as shown by red arrow). If the isolate is resistant, it can grow in the presence of the antibiotic (blue arrow). The diameter of the zone of organism inhibition (green arrow) is measured and compared with standard published zone sizes to determine whether sensitive or resistant.6
Sensitivity testing allows selection of an appropriate antibiotic for our patient and also screens for the presence of antimicrobial resistance mechanisms. Interpretative reading is an analysis of an organism's overall susceptibility pattern, allowing microbiologists to predict the mechanisms of resistance—not just the results for individual antibiotics.
Interpretation of sensitivity testing
The K. pneumoniae isolate from this patient was extremely resistant. It grew up to the amoxicillin (AMP), ciprofloxacin (CIP), gentamicin (CN), co-amoxiclav (AMC) and cefuroxime (CXM) discs. Interestingly, it was also resistant to cefpodoxime (CPD). Cefpodoxime is an indicator cephalosporin, used in laboratories to detect bacteria that produce extended spectrum β-lactamases (ESBLs). ESBLs are β-lactamases that confer resistance to penicillins, monobactams and the first-generation, second-generation and third-generation cephalosporins. The zone sizes for meropenem (MEM) and ertapenem (ETP) also appear reduced (figure 3).
To summarise, at this point, a child transferred from a high prevalence CPE country was colonised and had a wound infection, caused by a K. pneumoniae that appeared resistant to commonly used antibiotics. It may be an ESBL-producing organism and, furthermore, we are concerned about carbapenem susceptibility. Further testing is required to identify appropriate antibiotic treatment, but immediate action should be taken.
There are two preliminary concerns: optimal management of the index patient and protection of other children within this healthcare environment (infection control). Empiric antibiotic choice for potential CPE relies on national guidance4 and regional epidemiology until further sensitivity testing can be performed. The multi-resistant nature of CPE means that empiric antibiotic options are few and include colistin, tigecycline (not advised in those <8 years old due to permanent discolouration of the teeth), fosfomycin and sometimes aminoglycosides. A combination therapy is recommended in severe infections after discussion with a local microbiologist.
Prompt recognition and screening of at-risk children is vital to prevent hospital transmission of CPE. Infection control precautions are multidisciplinary and include patient isolation, single-use equipment where possible, appropriate cleaning/decontamination of reusable equipment and the environment, appropriate PPE, staff and visitor education and robust hospital surveillance to promptly detect an outbreak.4 CPE readily cause hospital outbreaks and many have been reported in the literature.7 Evidence also suggests that once entrenched in local healthcare environment, it is very difficult to eradicate these organisms.
Confirmation of ESBL production
To advise on appropriate antibiotic choice, it is vital to establish the mechanism(s) of antibiotic resistance.
Figure 4 shows phenotypic ESBL confirmatory testing. Once again, the isolate appears resistant to cefpodoxime (red arrow); however, it appears sensitive to the disc containing cefpodoxime and clavulanic acid (blue arrow). There is at least 5 mm difference between the two zone sizes, confirming the presence of an ESBL. The clavulanic acid inhibits the ESBL enzyme, preventing growth around that disc. However, the action of clavulanic acid against ESBL enzymes in the laboratory cannot be applied to patients with ESBL infection. In vivo, there is an unknown bacterial burden, and high levels of ESBL production can overwhelm the clavulanic acid causing treatment failure. Therefore, despite inhibition of ESBLs by clavulanic acid in vitro, co-amoxiclav is not recommended as treatment for serious infections with ESBL-producing Enterobacteriaceae.
To establish the meropenem MIC, an antibiotic concentration gradient strip was used (figure 5). At the top, there is a high concentration of meropenem, meaning that no sensitive organism could grow near the strip. It is impregnated with a decreasing concentration of meropenem and as the concentration decreases, the organism can grow closer and closer. Therefore, a sensitive organism forms a teardrop shape of organism inhibition (shown in blue). However, this K. pneumoniae appears to be highly resistant and also had a high MIC to ertapenem.
Carbapenem resistance can be caused by many different acquired mechanisms, not only carbapenemase enzymes. For example, a Pseudomonas aeruginosa isolate grown from a child with cystic fibrosis (who has received multiple courses of meropenem) may be resistant to meropenem, because the organism expresses fewer specific outer membrane porins.
Therefore, an organism that appears carbapenem resistant is not necessarily carrying a carbapenemase enzyme. Is this clinically important? Yes—most acquired carbapenemase enzymes are plasmid mediated. Plasmids are small, circular, DNA strands that are distinct from chromosomal DNA, and they can be transferred between organisms by a process called conjugation. Plasmid transfer occurs readily and therefore carbapenemase genes can quickly be transmitted from organism to organism and patient to patient. This has huge implications for infection control teams.
Consequently, it is important to establish how this organism has become resistant to carbapenems. Does it possess a carbapenemase enzyme? There are multiple ways to confirm this:8
Phenotypic methods—the modified Hodge test (MHT), discussed below
Molecular methods—using PCR to detect genes responsible for carbapenemase production
Other—for example, mass spectrometry
Some of these confirmatory tests are offered at reference laboratories.
Figure 6 shows a positive MHT. A lawn of sensitive E. coli is spread across an agar plate. Meropenem, ertapenem and imipenem discs are placed on the lawn. The patient's K. pneumoniae isolate is streaked in three lines from the centre of the plate to each disc (red arrow) and the plate is incubated. The sensitive E. coli lawn cannot grow around the discs. However, the K. pneumoniae can grow up to the discs because it is carbapenem resistant. Furthermore, if the K. pneumoniae is producing a carbapenemase enzyme, this diffuses into the agar around the test streak and allows the sensitive E. coli to grow closer to the disc along the streak. This clover leaf shape is a positive MHT result and tells you that there is carbapenemase production.
Our patient isolate was sent to the regional reference laboratory for molecular testing and was confirmed as a CPE (New Delhi metallo-β lactamase-producing K. pneumoniae). It was found to be sensitive to colistin, tigecycline and amikacin only. Tigecycline is contraindicated in a child of this age, leaving only two antibiotic choices. Systemic colistin exhibits dose-related neurotoxicity and nephrotoxicity, and therapeutic drug monitoring may be required. Antibiotic treatment of CPE is limited and the future Gram-negative antibiotic pipeline may fail to keep up with the development of antimicrobial resistance.
In 2014, 1648 CPE isolates were identified by the national reference laboratory in England; 67 (4%) of these were in children ≤16 years old.9 There has been a year-on-year increase in the incidence of CPE in England, and paediatricians need to be aware of the risk factors. Prompt recognition and screening can ensure that children receive appropriate antibiotics in a timely manner. This is crucial; studies have demonstrated a 28-day all-cause mortality of 40% associated with carbapenemase-producing Klebsiella spp bacteraemia.10 Additionally, prompt recognition can reduce harm to other children by rapid introduction of relevant infection prevention and control precautions to prevent transmission.
Multidrug-resistant organisms threaten to change the management of paediatric infection forever. Paediatricians will face difficulties using traditional empiric antibiotics for community-acquired infections, rigorous patient screening and infection control practices and heightened local and national surveillance of such organisms.
The authors thank Shona Braybrook, Phil Milner (Birmingham Children's Hospital NHS Foundation Trust) and Biomedical Scientists at Heart of England NHS Foundation Trust, Public Health England Laboratory for assistance with the figures 1–5. Figure 6 referenced. They also thank Professor Peter Hawkey (Professor of Public Health and Clinical Bacteriology) and Katie Hopkins (Head of Resistance Mechanisms Section, Antimicrobial Resistance and Healthcare-Associated Infections Reference Unit) for supplying data on CPE in children.
Competing interests None declared.
Provenance and peer review Commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.