Clinical characteristics and categorization of nosocomial isolates
Of the 14 drug-resistant study isolates, two were isolated from tap waters of the CTVS-ICU, while the remaining 12 were clinical isolates. Clinical characteristics of patients at the time of the isolation of the nosocomial strains include urological abnormalities (8 nos), Congenital Heart Disease (CHD) (3 nos) and Rheumatic Heart Disease (RHD) (1 no). (Table 1). Clinical and tap water isolates recovered in the study were identified as P. aeruginosa by Vitek-2 and re-confirmed by 16s rDNA sequencing. The study isolates were tested against Penicillins (carbenicillin, piperacillin+tazobactum and ticarcillin+clavulanic acid), cephalosporins (ceftazidime, cefipime and cefoperazone and sulbactum), monobactum (aztreonam), aminoglycosides (amikacin, gentamicin, netilmicin and tobramycin), carbapenems (imipenem, meropenem and doripenem), fluoroquinolones (ciprofloxacin, levofloxacin and ofloxacin), and polymyxins (colistin and polymyxin – B). Further, they were categorized as MDR, XDR and PDR based on the antibiotic sensitivity patterns (Table 2).
Seven of the 14 isolates were found to be MDR with resistance to more than three categories of the tested antibiotics. Six isolates which exhibited resistance to more than six categories of antibiotics were grouped as XDR. XDR strains were found to be sensitive to colistin. Notably, one isolate (208) was found to be PDR with resistance against all the eight categories of antibiotics tested and an MIC value of >=16 μg/ml against colistin. Of note, the other 13 isolates had MIC values of 0.5 to 2 μg/ml against colistin (Table 2).
Phylogenetic analyses
Phylogenetic analyses using PFGE and RAPD finger print analyses unambiguously revealed the presence of four distinct clusters. Clinical isolates were grouped into three different clusters, while the two tap water isolates (T3, T4) grouped into an independent cluster. Strains 199, 200, 208, 217, 225, 239, 251 and 255 isolated from the patients of the urology ward were found clustered into a single group. The Second cluster comprised the isolates 236 and 237 isolated from the CTVS-ICU patients. Strain 227 (VAP- sputum) although isolated from a patient in CTVS-ICU, was found to be distinct and formed the third cluster. Based on the PFGE (Figure 1) and RAPD (Supplementary figure 1) analyses, it may be inferred that the clonality of the study isolates is ward and location specific.
Biofilm production and MBIC
In vitro, all the 14 study isolates were found to produce strong biofilms (mean absorbance: >2.224 @OD 600nm) (Figures 2a, 2b and supplementary table 3). In this experiment, P. aeruginosa ATCC 27853 was used as a negative control as it does not produce biofilms. As expected, the study isolates were found to exhibit higher MBIC values compared to their corresponding MICs (supplementary table 4). In comparison to their MIC values, the MBIC values of the tap water isolates was highest against levofloxacin, while the PDR isolate 208 exhibited the highest MBIC (1024 μg/ml) against imipenem. Five isolates (208, 225, 239, 251 and 255) showed maximum MBIC value (1024 μg/ml) against piperacillin-tazobactam. Similarly, three isolates (241, 251 and 255) exhibited higher MBIC values (1024 μg/ml) compared to their corresponding MIC values against ceftazidime.
Carbapenem and fluoroquinolone resistance
Our analysis revealed the presence of NDM-1 gene among five of the study isolates (199, 200, 208, 217, 225, 239, 251 and 255). Sequencing of the QRDR (Quinolone Resistance Determinant Regions) revealed that eleven isolates (199, 200, 208, 217, 225, 227, 236, 237, 239, 251 and 255) had mutations in both gyrA (83Thr–Ile) and parC (87Ser–Leu) genes. Mutations in the QRDR region of the gyrB were observed in twelve isolates. A total of five different mutations were identified in the gyrB gene among the study isolates. 468Glu–Asp mutation was found in seven isolates, 504Asp–His mutation in five isolates, 503Val–Leu and 533 Gln–His mutations in two isolates and 490Leu–Val mutation was seen in one isolate (Table 3).
Int1 gene and 5̍ - 3 ̍ conserved segment of class 1 integrons were observed among twelve of the analyzed isolates (Table 3). Notably, isolate 208 which was resistant to all the tested antibiotics did not possess class 1 integrons raising the possibility that other mobile genetic elements may have played a role in its emergence into a PDR phenotype.
Mutations in the efflux pump and efflux pump regulatory genes
Mutations in efflux pump genes were identified among some of the study isolates. In the mexB gene, isolate 208 was found to have 576Val–Gly and 578Thr–Asn mutations; isolate 227 had 607Ser–Thr, 611Thr–Ala and 612Val–Glu mutations; while isolate 237 had 615Phe–Leu mutation and isolate 239 possessed 589Gln–Pro mutation. mexF gene analysis revealed that isolate 217 had 792Glu–Ala and 816Lys–Met mutations while T3 and T4 isolates had 843Ala–Thr mutation (Table 3). Only the tap water isolate T4 was found to possess 93Leu–Val and 131Gln–His mutations in the nfxB gene. 126Val–Glu mutation was observed in the mexR gene among six of the isolates (227, 236, 237, 241, T3 and T4). Further, isolates 236 and 237 were found to possess an additional mutation (143Pro–Leu) in the mexR gene.
Interventions
Repeated chlorine treatment of water for one week was performed to eliminate the presence of the pathogens in the hospital water sources. Further, disinfection of all potential reservoirs (overhead water tanks, sumps and taps) was attempted and strict infection control measures were implemented. Additionally, plumbing (pipes and faucets) in the CTVS-ICU where P. aeruginosa spp. were isolated from tap water has been totally replaced. A new overhead water tank with hyper-chlorinated (2ppm) water was installed for the purpose of scrubbing in all the ICUs and operation theaters. Enhanced surveillance, awareness programs for the staff and continuous monitoring has been initiated by the infection control team. Follow up environmental surveillance in and around the hospital premises (repeat sampling over a period of three months) did not reveal the presence of P. aeruginosa and the incidence of P. aeruginosa infections subsided. However, regular screening of water samples in the hospital premises has been made mandatory to identify any potential source of infection.