Descriptive analysis of infections due to New Delhi metallo-β-lactamase-producing Enterobacterales in children at Red Cross War Memorial Children’s Hospital

The increased incidence and absence of antibiotic treatment options for New Delhi metallo-β-lactamase (NDM)-producing carbapenem-resistant Enterobacterales (CRE) infection are concerning. Recent reports have highlighted NDM-producing Serratia marcescens, as a specific concern, as it is an organism which is intrinsically resistant to colistin. In this study, a descriptive analysis of NDM-producing CRE infections was performed at the Red Cross War Memorial Children’s Hospital.


Introduction
Carbapenem resistance in Enterobacterales is most commonly mediated by the production of carbapenemases that hydrolyse almost all β-lactams (BLs), including the carbapenems that are considered the last-line and therapeutic choice for severe infections caused by extended spectrum β-lactamase (ESBL)-producing Enterobacterales. 1,2 Carbapenemases can be classified into those with a serine residue (class A and D) and those with zinc (class B or metallo-βlactamase [MBL] enzymes) at their active site. New Delhi metallo-β-lactamase (NDM) is a class B carbapenemase. 3 A recent publication on the epidemiology of carbapenem-resistant Enterobacterales (CRE) bacteraemia in South African tertiary hospitals included a significant number of children and adolescents (485/1293, 38%) and predominantly identified the class D Oxacillinase-48 (OXA-48) carbapenemase and its variants (52%), followed by the class B NDM (34%) carbapenemase. The most common bacteria identified in this study were Klebsiella pneumoniae, Enterobacter cloacae and Serratia marcescens. 4 Furthermore, in 2017, Moodley et al. raised concern about the emergence of CRE S. marcescens harbouring the NDM-1 gene, where 10/63 (16%) S. marcescens isolates from blood cultures across the KwaZulu-Natal region carried this gene. 5 This high prevalence of NDM-producing CRE, and specifically those caused by S. marcescens, is a significant public health concern as S. marcescens is intrinsically resistant to colistin, the antibiotic of choice for CRE infections in low-income settings, with no alternative directed treatment available. 6 Aztreonam-avibactam is being investigated in clinical trials for the treatment of NDM-producing CRE. 3 In the interim, cefiderocol or a combination therapy with ceftazidime-avibactam (CA) and aztreonam (AT) has shown promising results for the treatment of NDM CRE infection. Cost, registration status with the South African Health Products Regulatory Authority, limited access to the antibiotics and susceptibility testing, and uncertainty of dosage and duration of therapy in children are some of the challenges. 2,7,8 Similar issues will be expected with aztreonam-avibactam when it becomes available for clinical use.
Due to an increase in isolation of NDM-producing CRE at the health care institution, a retrospective descriptive analysis of these isolates was performed and includes epidemiological and microbiological characteristics, as well as management and outcome measures.

Results
Twenty-six NDM-producing CRE isolates from 22 children were included in the study.

Demographics
Most children included were female (59%

Risk factors
All patients had at least one known risk factor for CRE infection, 6 (Figure 1a).

Management and outcome
The clinical characteristics, management and outcome of all patients are summarised in were treated with antibiotics that did not demonstrate any in vitro susceptibility (Table 1).

Discussion
The increase in CRE infections, and in particular NDMproducing S. marcescens, has been identified as a public health concern with the potential for outbreaks and being untreatable. 9,10 In keeping with this trend, the authors noted increasing numbers of NDM-producing CRE and specifically NDM-producing S. marcescens from clinical samples at the healthcare institution. A total of 81% of children who isolated NDM CRE had ≥ 3 risk factors present, with the majority being the presence of comorbidities, prior hospitalisation and antibiotic exposure in the preceding three months, and ICU admission, highlighting the vulnerable population that NDM CRE affects. New Delhi metallo-β-lactamase was first identified in K. pneumoniae, and then alarmingly in S. marcescens, which has subsequently become the predominant NDM-producing organism at the institution. Resistance to antibiotics was higher in S. marcescens than in K. pneumoniae with the majority being pan drug-resistant (9/11, 82%) and having high meropenem MICs (≥ 32 μg/mL) (8/10, 80%), when compared to the lower proportion of pan drug-resistant K. pneumoniae isolates (2/15, 13%) and those with high meropenem MICs (5/9, 56%). However, it is also important to be aware that BL susceptibility testing of MBLproducing organisms is influenced by the presence of zinc in the testing medium, which if depleted may render lower MICs or false susceptibility to BL antibiotics, making interpretation additionally challenging. 3 Further complicating this are NDM CRE case reports and simulated human models describing successful clinical outcomes and reduction in bacterial loads, respectively, using carbapenems, despite high MICs. 11,12 Treatment of CRE infections in children, and those caused by NDM-producers, is confounded by a paucity of datadriven evidence and availability of effective antibiotics. Treatment is usually individualised based on in vitro susceptibility profiles and pharmacokinetics of available but less optimal antibiotics, such as combination or monotherapy with carbapenems, aminoglycosides, fluoroquinolones, polymyxins, and tigecycline. These antibiotics are used with various degrees of success depending on the severity and site of infection, and whether source control can be attained. Outcomes in children vary, with descriptions of lower mortality in those with NDM infection when compared to non-NDM infection being reported, 12  Antimicrobial Stewardship Programmes through institutional and individualised guidance play a major role in the management of these complex infections and perhaps, more importantly, avoiding these infections by limiting selection pressure through the judicious use of broadspectrum antibiotics. The use of CRE surveillance data as well as technological advances in the identification of (multi-drug resistant) micro-organisms can also limit unnecessary prolonged exposure to broad spectrum antibiotics. 13 Finally, it is critical to always reinforce compliance with infection prevention and control such as hand hygiene, barrier nursing, and enhanced environmental cleaning measures, to limit the spread of these organisms within institutions, as developing new antibiotics only partially address the problem of multidrug-resistant organisms and is neither an immediate nor sustainable solution to the crisis.
Limitations of this study analysis include not estimating the longitudinal prevalence rate of CRE at the Red Cross War Memorial Children's Hospital. The impact of delays in appropriate directed therapy, due to limited access to expedited pheno-and genotypic susceptibility testing, could not be analysed. The differences in severity of illness, which is an important outcome determinant, were also not compared. In addition, as whole-genome sequencing of CRE was not performed, resistome and virolome determinants, phylogenetic and plasmid analysis, and thus transmission dynamics at the institution is unknown. This may be important to elucidate the underlying reasons for the replacement of K. pneumoniae by NDM S. marcescens and its current dominance. Going forward, such data are pivotal, not only to inform data-driven patient management, but also diagnostic and antibiotic stewardship and infection prevention strategies.

Conclusion
Reports of NDM CRE are increasing, with NDM-producing S. marcescens specifically being identified as a serious public health threat. This study highlights the paucity of antibiotics available to treat these infections, especially those caused by NDM S. marcescens, which appear to be more resistant to the backbone treatment options for CRE infection.