Analysis of Antimicrobial Resistance Patterns, Genetic Basis of Resistance and Phylogenetic Relatedness in clinical Pseudomonas aeruginosa isolates from in-patients at Kenyatta National Hospital

Abstract

This study sought to determine antimicrobial susceptibility profiles of Pseudomonas aeruginosa strains recovered from in-patient population at the Kenyatta National Hospital, the largest referral hospital in Kenya. A total of 188 P. aeruginosa strains were obtained from different in-patient wards from August 2015 to January 2016. Minimum inhibitory concentrations (MICs) were conducted on the Vitek 2-Compact (Biomereux company-France). High resistance in P. aeruginosa isolates was recorded towards tetracycline (92%) with an MIC of ≥ 128µg/ml followed by cefotaxime (88.8%) and ceftriaxone (86.2%) with MICs of ≥ 64µg/ml. Lowest resistance was recorded towards piperacillin (25%) and amikacin (46.3%). P. aeruginosa isolates recovered from the Intensive Care Unit (ICU) recorded the highest resistance proportion of 83% to all the antimicrobial tested while least resistance proportions was observed among strains from the Newborn unit (NBU) ward (38%). On the other hand, majority (92%) of isolates obtained from urine specimen were resistant to any given antimicrobial drug tested while lowest resistance was recorded among isolates obtained from blood (29%).Resistance to CAZ, CIP, CN, and AMK was 82.4%, 80.9%, 88.2% and 78% respectively. A high proportion (86%) of the MBL positive strains were recovered from patients in the CCU, followed by the medical wards (13%), while new born unit was the least affected (1%). From a total of 127 P. aeruginosa that were resistant to meropenem, (68%) were positive for blaNDM, a carbapenemase while 64 isolates harbored blaVEB, an ESBL gene. A total of 45 isolates tested positive for both blaNDM, blaVEB and for class 1 integron. A single isolate from tracheal aspirate sample from a 39 years old female admitted in the ICU harbored NDM, VEB, integron class 1 and 3. Plasmid screening revealed 3 types of incompatibility groups, incW, incFIB and incFIB. One P. aeruginosa isolate had both incW and incFIB, while another had an incN. Phylogenic cluster analysis using the Gelcompar2® revealed four major clusters based on age, specimen type and wards. The four clusters had a significant genetic similarity of >80% amongst P. aeruginosa strains obtained from different wards which is indicative of cross-infection. The high resistance recorded in this study is therefore worrying and may impair our ability to combat severe P. aeruginosa infections. Resistance to β-lactams, aminoglycosides and fluoroquinolones further narrows down the available treatment options considering carbapenems are not readily available in Kenya like other developing countries. Even worse, in absence of proper usage of these agents it could partially explain why resistance to carbapenems is on the rise. If left unchecked, this resistance may lead to drastic increase in pan-resistance strains that may cause in turn high mortality in hospital and community settings. The strong evidence of clonal spread in various wards show that this problem is not confined in a specific unit and therefore all relevant bodies should engage to help contain this problem.