Combination of pre-adapted bacteriophage therapy and antibiotics for treatment of fracture-related infection due to pandrug-resistant Klebsiella pneumoniae

Abstract

A 30-year-old bombing victim with a fracture-related pandrug-resistant Klebsiella pneumoniae infection after long-term (>700 days) antibiotic therapy is treated with a pre-adapted bacteriophage along with meropenem and colistin, followed by ceftazidime/avibactam. This salvage therapy results in objective clinical, microbiological and radiological improvement of the patient's wounds and overall condition. In support, the bacteriophage and antibiotic combination is highly effective against the patient's K. pneumoniae strain in vitro, in 7-day mature biofilms and in suspensions.

Fig. 1. Timeline of the most relevant surgical procedures (green), microbiological results (blue), antibiotic therapies (dark grey), and phage therapy (magenta).

Phage therapy related events are indicated in fuchsia. AMB, liposomal amphotericin-B (600 mg, q24h); AMK, amikacin (1800 mg, q24h); CAZ-AVI, ceftazidime/avibactam (2 g/0.5 g, q8h); CIP, ciprofloxacin (400 mg, q8h); CIR, clarithromycin (500 mg, q12h); CST, colistin (up to 20 MIU, q24h); EMB, ethambutol (1200 mg, q24h); FLX, flucloxacillin (2 g, q4h); FOM, fosfomycin (4 g, q4h); GEN, gentamicin (400 mg, q24h); MEM, meropenem (up to 2 g, q4h); POS, posaconazole (200 mg, q8h); TMP-SMX, trimethoprim/sulfamethoxazole (1660 mg, q12h); MXF, moxifloxacin (400 mg, q24h); PDR, pandrug-resistant; RMP, rifampicin (450 mg, q24h); TGC, tigecycline (100 mg, q12h); TPZ, piperacillin/tazobactam (4.5 g, q6h); VAN, vancomycin (30 mg/kg/day); XDR, extensively drug-resistant.

Fig. 2. Characteristics of phage M1.

Transmission electron micrograph showing an icosahedral head (120 × 120 nm) and a contractile tail (150 × 22 nm) (a). Lytic activity of phage M1 against 141 Klebsiella spp. clinical isolates obtained from Georgia (n = 88), France (22), Switzerland (21), Singapore (6), and China (4) (b). Biological characteristics of phage M1 as determined in propagation strain 1a. Adsorption curve (c), single-step growth curve (d), pH stability (e), and temperature stability (f). BS, burst size; LP, latent period; PFU, plaque-forming unit. Source data are provided as a Source Data file.

Fig. 3. Relevant genomic and proreomic characteristics of phage M1.

Genome representation of phage M1 and comparison, using a BLASTn analysis, with Slopekvirus type species KP15 (a). Each white or colored arrow represents a predicted open reading frame. In orange, genes encoding packaging and lysis-associated proteins are displayed, in yellow structural proteins and in blue DNA- and metabolism-associated proteins (adapted from EasyFig 2.2.2). The single missense mutation found in the preadapted phage M1 isolate used for therapy is indicated (gp269). AlphaFold2 model of the hinge connector of the distal tail fiber (b). The mutation of a Threonine to an Arginine is indicated (position 94). Alpha helices are indicated in red and beta sheets in yellow. Green stretches exhibit no specific structures. The model quality scores are presented to the right of the model. lDDT local distance difference test.

Fig. 4. Bandage visualization of the short-reads (Illumina) and hybrid (Illumina + Nanopore) assembly graphs of Klebsiella pneumoniae isolates Kp040741, Kp040762, and Kp36336.

The different replicons can be distinguished in the hybrid assembly graphs and include the chromosome and up to five plasmids (isolate Kp040762).

Fig. 5. Visualization of the functional content of the pandrug-resistant Klebsiella pneumoniae isolate Kp36336 among the different replicons.

The antibiotic resistance genes (ARGs) and single nucleotide polymorphisms (SNPs), as well as the prophage content and genomic islands, found in the three distinct K. pneumoniae isolates were added (Supplementary Data 2–4).

Fig. 6. Pre-adapted phage M1 activity.

The activity of pre-adapted phage M1 against Klebsiella pneumoniae isolated on day 170 (isolates Kp04741and Kp04762) and day 702 (isolate Kp36336) postinjury were determined. Phage activity was measured at different multiplicities of infection (MOI): MOI 10.0 (a), MOI 1.0 (b), and MOI 0.1 (c). Controls consisted of the K. pneumoniae isolates without phages. The curves represent bacterial proliferation (cellular respiration). Results are presented as mean values of biological replicates (n = 3) with error bars representing the standard deviations of the means. Source data are provided as a Source Data file.

Fig. 7. Comparison of the patient’s condition before and after combined phage-antibiotic treatment.

Relevant clinical, biological, microbiological, and radiological parameters before (within 3 days before, or – for microbiological parameters – during surgery) and after treatment (a). CRP, C reactive protein; CT, computed tomography; ESR, erythrocyte sedimentation rate; PDR pandrug-resistant, WBC white blood cells. Antero-external view of the patient’s left hip and thigh 18 months before and 3 months after treatment (b). Successive computed-tomography scanners of the left femur before (admission, June 2017 and February 2018) and after treatment (June 2018 and June 2019) (c). Neutralization of pre-adapted phage M1 by antibodies produced by the patient’s adaptive immune system upon phage treatment. Bars represent the mean values of biological replicates (n = 6), each represented by a circle, with error bars representing standard deviations of the means (d). Source data are provided as a Source Data file.

Fig. 8. Activity of preadapted phage M1, ceftazidime/avibactam, and combinations thereof against PDR K. pneumoniae isolate Kp36336 residing in mature biofilms.

Seven-day mature biofilms of K. pneumoniae isolate Kp36336 were exposed daily for 3 subsequent days to phages (7.7 × 108 PFU), moderate concentrations of ceftazidime/avibactam (just below and just above the MIC value of 2.0/0.5 μg/ml), and combinations thereof. After exposure for 1–3 days, bacteria in the biofilms were recovered by sonication (10 min at 40 kHz) in saline (or saline supplemented with 10 µM ammonium iron (II) sulphate hexahydrate to neutralize residual phage activity) and subsequently plated on Mueller Hinton agar for microbiological determination of bacterial counts. Results are presented as the median (horizontal bars) of biological replicates (n = 5), each dot representing the sum of bacterial counts from pools of 6 wells (technical replicates). The significance of the difference in bacterial counts between samples from biofilms exposed to antibiotic/phage combination and samples from biofilms exposed to the antibiotic alone was determined using a two-sided Mann Whitney U tests, *p < 0.05 (p = 0.0238), **p < 0.01 (p = 0.0079). Source data are provided as a Source Data file.

Fig. 9. Phage-antibiotic synergy.

The activity of preadapted phage M1 at different multiplicities of infection (MOIs), of antibiotics (ceftazidime/avibactam, meropenem, and colistin) at different concentrations, and of combinations thereof, against pandrug-resistant (PDR) Klebsiella pneumoniae isolate Kp36336 (isolated 702 days postinjury) were determined. Ceftazidime/avibactam at different concentrations (0.125, 0.25, and 0.5 mg/l) and phage M1 at MOI 1.0 (a), 0.1 (b), or 0.01 (c), and combinations thereof. Meropenem (4.0 mg/l) at MOI 1.0, 0.1, and 0.01 (d). Colistin (1.0 mg/l) at MOI 1.0, 0.1, and 0.01 (e). Controls consisted of the K. pneumoniae isolate Kp36336 without phage M1 or antibiotics. Bacterial proliferation (cellular respiration) is presented. Efficacious phages, antibiotics, and combinations thereof, suppress bacterial proliferation. Results are presented as mean values of biological replicates (n = 3) with error bars representing the standard deviations of the means. Source data are provided as a Source Data file.