Prolonged Outbreak of Carbapenem and Colistin-Resistant Klebsiella pneumoniae at a Large Tertiary Hospital in Brazil

Abstract

Multidrug-resistant gram-negative bacteria, such as carbapenem and colistin-resistant Klebsiella pneumoniae (ColR-CRKP), represent a major problem for health systems worldwide and have high lethality. This study investigated the genetic relationship, antimicrobial susceptibility profile, and resistance mechanisms to ColR-CRKP isolates from patients infected/colonized in a tertiary hospital in Salvador, Bahia/Brazil. From September 2016 to January 2018, 46 patients (56 ColR-CRKP positive cultures) were enrolled in the investigation but clinical and demographic data were obtained from 31 patients. Most of them were men (67.7%) and elderly (median age of 62 years old), and the median Charlson score was 3. The main comorbidities were systemic arterial hypertension (38.7%), diabetes (32.2%), and cerebrovascular disease (25.8%). The average hospitalization stay until ColR-CRKP identification in days were 35.12. A total of 90.6% used mechanical ventilation and 93.7% used a central venous catheter. Of the 31 patients who had the data evaluated, 12 had ColR-CRKP infection, and seven died (58.4%). Previous use of polymyxins was identified in 32.2% of the cases, and carbapenems were identified in 70.9%. The minimum inhibitory concentration (MIC) for colistin was > 16 μg/mL, with more than half of the isolates (55%) having a MIC of 256 μg/mL. The bla _KPC gene was detected in 94.7% of the isolates, bla _NDM in 16.0%, and bla _GES in 1.7%. The bla _OXA-48, bla _VIM, and bla _IMP genes were not detected. The mcr-1 test was negative in all 56 isolates. Alteration of the mgrB gene was detected in 87.5% (n = 49/56) of the isolates, and of these, 49.0% (24/49) had alteration in size probably due to IS903B, 22.4% (11/49) did not have the mgrB gene detected, 20.4% (10/49) presented the IS903B, 6.1% (3/49) had a premature stop codon (Q30*), and 2.1% (1/49) presented a thymine deletion at position 104 - 104delT (F35fs). The PFGE profile showed a monoclonal profile in 84.7% of the isolates in different hospital sectors, with ST11 (CC-258) being the most frequent sequence type. This study presents a prolonged outbreak of ColR-CRKP in which 83.9% of the isolates belonged to the same cluster, and 67.6% of the patients evaluated had not used polymyxin, suggesting the possibility of cross-transmission of ColR-CRKP isolates.

Keywords: Enterobacterales infections; carbapenem-resistant Klebsiella pneumoniae; colistin-resistant Klebsiella pneumoniae; multidrug-resistant gram-negative bacteria; outbreak.

FIGURE 1

Flowchart of ColR-CRKP positive culture cases identified during the study period.

FIGURE 2

Alignment of the amino acid sequences of mgrB from ColR-CRKP strains that were not disrupted by insertion sequences. Dots indicate identical residues; asterisk indicates stop codon; letter indicates amino acid substitution. The first sequence in the alignment belongs to the wild-type strain Klebsiella pneumoniae MGH 78578 (CP000647.1).

FIGURE 3

Dendrogram showing the relatedness of ColR-CRKP based on PFGE patterns after digestion with XbaI. Isolates showing 80% similarity are considered related. ARG, Antimicrobial Resistance Genes; ID, identification; KPC, Klebsiella pneumoniae carbapenemase; NDM, New Delhi metallo-beta-lactamase; PFGE, Pulsed-field electrophoresis technique; ST, sequence typing; WT, Wild type; fs, Frameshift.

FIGURE 4

Timeline and transmission opportunities among patients during the outbreak including the index patient according to the PFGE profile, culture date and period of hospitalization. The 31 patients who had their medical records evaluated and the index case were placed on the Y axis. The numbers in black on the Y-axis indicate patients who were admitted to the general ICU and in red patients from other sectors of the hospital (24: UAVC; 25: emergency; 27: medical clinic ward; 28: emergency ICU; 30: emergency ICU; 45: emergency ICU and 46: ICU: neurological). The length of stay was placed monthly on the X-axis. Each bar indicates a patient, and the length of the bar signals the length of stay. The transverse lines on each bar indicate the moment of identification of ColR-CRKP. The color of the bars indicates the PFGE profile.

FIGURE 5

Temporal distribution of 46 patients with ColR-CRKP and the index case according to the PFGE profile and intervention points performed in the hospital. Full black line: removal of bed curtains, intensification of cleaning of the environment, greater availability of alcohol gel for hand hygiene, removal of adornments from health professionals, introduction of rectal swab at the admission of patients to the sector and daily control of prescription of antimicrobials. Black dotted line: functional cohort of staff for patients colonized/infected by ColR- CRKP, intensification of hand hygiene, universal contact precaution in all patients admitted to the ICU and weekly rectal swab in all patients (in addition to the swab rectal and nasal surveillance at admission of all patients). Dashed black line with dot: physical restructuring of the unit, removal of damaged mattresses, terminal cleaning twice (by different hygiene teams) of the beds of patients who were discharged or died before the hospitalization of another patient.