Successful phage-antibiotic therapy of P. aeruginosa implant-associated infection in a Siamese cat

Abstract

Antibiotic-resistant pathogens are a growing global issue, leading to untreatable infectious diseases in both humans and animals. Personalized bacteriophage (phage) therapy, the use of specific anti-bacterial viruses, is currently a leading approach to combat antibiotic-resistant infections. The implementation of phage therapy has primarily been focused on humans, almost neglecting the impact of such infections on the health and welfare of companion animals. Pets also have the potential to spread resistant infections to their owners or the veterinary staff through zoonotic transmission. Here, we showcase personalized phage-antibiotic treatment of a cat with a multidrug-resistant Pseudomonas aeruginosa implant-associated infection post-arthrodesis surgery. The treatment encompassed a tailored combination of an anti-P. aeruginosa phage and ceftazidime, precisely matched to the pathogen. The phage was topically applied to the surgical wound while the antibiotic was administered intramuscularly. After two treatment courses spanning 7 and 3 weeks, the surgical wound, which had previously remained open for five months, fully closed. To the best of our knowledge, this is the first case of personalized phage therapy application in felines, which provides further evidence of the effectiveness of this approach. The successful outcome paves the way for personalized phage-antibiotic treatments against persistent infections therapy in veterinary practice.

Keywords:

• Bacteriophage
• phage therapy
• pseudomonas aeruginosa
• antibiotic resistance
• implant-associated infection

Introduction

Antibiotic-resistant infections pose a formidable challenge in modern medicine, impacting both humans and companion animals. In 2019, infectious diseases emerged as the second leading cause of mortality in humans, surpassed only by cardiovascular diseases (Murray et al. 2022). Among companion animals, a notable prevalence of antibiotic-resistant strains worldwide is also observed, leading to increasingly challenging treatment scenarios (Sobkowich et al. 2023; Marco-Fuertes et al. 2022; Caneschi et al. 2023; Habib et al. 2023; Yaovi et al. 2022; Ju et al. 2023; Leite et al. 2023; Burke and Santoro 2023).

One of the most prevalent bacterial infectors is Pseudomonas aeruginosa, an opportunistic pathogen known for causing a wide variety of infections in humans and animals, including respiratory infections, soft tissue infections and urinary tract infections (Qin et al. 2022; Hillier et al. 2006; Sharma et al. 2019; Feßler et al. 2022). Due to its high healthcare burden and high prevalence of antimicrobial resistance, it has been acknowledged by the World Health Organization as a critical priority pathogen in human health (Tacconelli et al. 2018). The European Food Safety Authority has also identified the significance of this pathogen, as it has the potential to be transmitted from pets to their owners and medical staff (Nielsen et al. 2022). These recognitions warrant immediate attention for the development of innovative therapeutic strategies to address the challenges P. aeruginosa and many other dangerous antibiotic-resistant bacterial pathogens pose (Jin et al. 2023).

Bacteriophage (phage) therapy is the utilization of bacteria-specific viruses against bacterial pathogens. Although it was developed over a century ago, this therapeutic approach has experienced a resurgence in interest worldwide in the last decade as a potential solution to combat persistent infections (Uyttebroek et al. 2022). Phages provide a means to overcome resistance that may emerge during treatment. This can be achieved through various methods, including isolating new phages from the environment, employing in vitro adaptation or “training” techniques to enhance their infectivity against the bacterial host, and phage engineering (Dedrick et al. 2019; Rohde et al. 2018). Phages also demonstrate a remarkable ability to penetrate and disrupt bacterial biofilms (Ferriol-González and Domingo-Calap 2020). Furthermore, phages, display a high level of specificity towards their bacterial hosts, ensuring precise targeting and elimination of the pathogen while preserving the microbiome (Altamirano and Barr 2019).

However, due to their specificity, precise matching of the phage to the specific bacterial pathogen is essential before treatment can be administered effectively (Yerushalmy et al. 2023). For this reason, many successful phage-therapy treatments in recent years for various human infections utilized phages in a personalized approach, including by our team (Schooley et al. 2017; Law et al. 2019; Nir-Paz et al. 2019; Jennes et al. 2017; Israeli Phage Therapy Center (IPTC) Study Team 2023). In these treatments, phages are administered in conjunction with antibiotics, taking advantage of the synergistic effect that can appear between the two agents, leading to better killing effect of the pathogen (Altamirano and Barr 2019; Gelman et al. 2018).

In veterinary medicine, a significant portion of in-vivo phage therapy studies was conducted in controlled settings involving farm or laboratory animals (Loponte et al. 2021; Marshall and Marsella 2023). However, when it comes to companion animals, phage therapy cases are scarcely described, with most of the published works demonstrating the potential of phages against pathogens isolated from pets primarily in in vitro settings (Lerdsittikul et al. 2022; Santos et al. 2022; Freitag et al. 2008). To the best of our knowledge, the only significant documented treatment in companion animals was a partially controlled clinical trial reported by Hawkins et al. (2010). In this trial, a cocktail of six bacteriophages was administered to dogs suffering from antibiotic-resistant otitis media. In this trial, improvement of the infection was commonly noticeable after 48 h, and no adverse reactions were observed (Hawkins et al. 2010).

Here, we present a case of phage-antibiotic therapy personally tailored to treat a cat suffering from an implant-associated P. Aeruginosa infection that developed post-arthrodesis surgery. The positive outcome of this case highlights the potential of phages as a valuable tool in combating antibiotic-resistant infections in companion animals.

Material and methods

Bacterial strains

All five Pseudomonas aeruginosa strains were isolated from the surgical wound using cotton swab samples and cultured in LB broth medium or LB-agar plates at 37 °C for 18–24 h. The identification and antibiogram (Table 1) of the first isolated strain were conducted by AML Lab Services in Herzliya, Israel (https://www.aml.co.il/en/). The identity of the four strains isolated throughout the treatment was validated by colony and microscopic morphology. Bacterial lawns were prepared by transferring 100 µL of an overnight bacterial liquid culture to 4 mL of pre-heated agarose (0.3%) and pouring the mixture onto LB-agar plates.

Table 1. Antibiogram of the cat’s pathogen conducted by AML Lab services in Herzliya, Israel. The bacterium was sensitive to ceftazidime and imipenem.

Antibiotic	Sensitivity	MIC (µg/mL)
Ceftazidime	Sensitive	2
Imipenem	Sensitive	1
Amikacin	Intermediate	8
Gentamycin	Intermediate	4
Enrofloxacin	Resistant	≥4
Marbofloxacin	Resistant	≥2

Clinical phage microbiology

To match a phage-antibiotic combination against the isolated P. aeruginosa, we followed the Israeli Phage Center Clinical Phage Microbiology protocol (Yerushalmy et al. 2023). Briefly, the anti-P. aeruginosa phage collection (34 phages) from the Israeli Phage Bank (Yerushalmy et al. 2020) was screened against the bacterial pathogen isolated from the wound by spotting 5 µL of each phage on the pathogen bacterial lawn. After overnight incubation, unique phages that showed clear plaques were tested separately for plaque-forming units (PFU) count by spotting one 10 µL drop from each of the six serial dilutions of each phage onto the bacterial lawn of the pathogen and a reference strain. After overnight incubation, the efficiency of plating was calculated by dividing the titer on the target strain by the phage titer on its designated reference strain. For the growth kinetics assay, we selected phage ΦPASB7 and ceftazidime. ΦPASB7 was selected as it demonstrated high efficiency of plating value (∼1.0), and we had an available high-titer stock of it. Ceftazidime was selected due to the antibiogram. The growth kinetics assay was conducted using a 96-well plate reader (BioTek Eon Microplate Spectrophotometer, Winooski, VT). At the kinetics assay endpoint, a viable cell count was conducted for all wells that showed complete growth inhibition by colony-forming units (CFU) assay. Results graphs and statistics were prepared using GraphPad Prism 8.4.3.

The phage ΦPASB7

ΦPASB7 was previously isolated from a water sample collected in Jerusalem in 2022 as part of our lab phage “hunting” efforts. The sample was centrifuged (5000 × g) for 10 min and filtered (0.22 µm). The supernatant was enriched with a lab strain P. aeruginosa 160 (PA160) and LB medium. Following an overnight incubation, the sample was filtered again, and a plaque assay was conducted on a lawn of PA160 bacteria. The next day, a single plaque was streaked on a fresh PA160 lawn. After five generations of plaque isolation streaks, a single plaque was transferred into an LB medium and incubated with PA160 for amplification.

The genome of the phage was sequenced as previously describedKhalifa et al. 2016. Briefly, DNA was extracted using a phage DNA isolation kit (Norgen Biotek, Thorold, Canada) (Summer 2009), and libraries were prepared using the Illumina Nextera XT DNA kit (Illumina, San Diego, CA). Normalization, pooling, and tagging were performed with a flow cell (1 × 150 bp single-end reads). Sequencing was performed using the Illumina NextSeq 500 platform at the Hebrew University of Jerusalem sequencing unit at Hadassah Campus. Trimming, quality control, read assembly, and analyses were performed using Geneious Prime 2023.2.1 and its plugins (https://www.geneious.com). Genome assembly was performed using the SPAdes plugin of Geneious Prime. Annotation was performed using PHAROKA (https://github.com/gbouras13/pharokka). The phage genome was scanned for resistance genes and virulence factors using Abricate (Seemann T, Abricate, https://github.com/tseemann/abricate) based on its databases: NCBI, CARD, Resfinder, ARG-ANNOT, EcOH, MEGARES, PlasmidFinder, Ecoli_VF, and VFDB. ΦPASB7 sequence was uploaded to the NCBI GenBank, accession number: OR509539.

TEM visualization of ΦPASB7 was performed in the microscopy department of the intradepartmental unit of Hebrew University as described (Khalifa et al. 2016). Briefly,10⁹ PFU/mL ΦPASB7 suspension was centrifuged (13,500 × g) for 2 h, after which the supernatant was discarded, and the pellet was resuspended in 200 μL of 5 mM MgSO₄ and left overnight in 4 °C. From this suspension, 10 µL were spotted on a carbon-coated copper grid with an addition of 2% uranyl acetate and allowed to incubate for 1 min. Afterward, the excess was removed, and the sample was visualized using a TEM 1400 plus from Joel, Tokyo, Japan. Images were captured with a charge-coupled device camera (Gatan Orius 600).

Phage preparation for treatment

A 100 mL ΦPASB7 phage suspension (10⁹ PFU/mL) was washed and concentrated by centrifugation (4000 × g, 5 min) using phosphate-buffered saline (PBS) and 30 kDa Amicon® centrifuge tubes, resulting in a 5 mL phage suspension (10¹⁰ PFU/mL). The concentrated phage suspension was then carefully transferred to the top of an ultracentrifuge tube containing an iodixanol (OptiPrep^TM, SIGMA-ALDRICH) density gradient (15%, 25%, 40%, 60%). Following ultracentrifugation for 2 h at 350,000 × g, each gradient fraction was collected separately, and iodixanol was washed from each fraction with PBS in 30 kDa Amicon® centrifuge tubes (4000 × g, 5 min, 3 times). Plaque assays were performed for each fraction to quantify the phage titer. Additionally, the endotoxin levels of all fractions were evaluated using the LAL Chromogenic Endotoxin Quantitation Kit (Charles River Laboratories), following the manufacturer’s instructions. Briefly, each fraction underwent a two-fold serial dilution with endotoxin-free water. Subsequently, 100 μL from each dilution, as well as from the original fractions, were dispensed into a 96-well plate. To the same plate, a serial dilution of known LPS concentrations (50, 5, 0.5, 0.05, 0.005 Endotoxin Units (EU)/mL) was added as a standard. The optical density was measured at a 405 nm wavelength every 2 min using a plate reader for 40 min.

The fraction that was used for the phage suspension preparation was the one that exhibited the highest phage titer (10⁸ PFU/mL) and the lowest endotoxin units (50 EU/mL). For the final preparation, the pure phage was diluted 1:10 with PBS and distributed into syringes, each containing 1 mL of purified phage (10⁷ PFU/mL, 5 EU/mL). These syringes were stored at 4 °C until use, with a maximum storage period of up to 3 weeks.

Ethical approval

The current case study is a compassionate treatment that was administered by the Vet Holim Medical Center (https://www.vet-holim.com/blank-page) under the guidelines and the regulations of the World Small Animal Veterinary Association (WSAVA, https://wsava.org) using the American Veterinary Medical Association (AVMA, https://www.avma.org) protocols and the American Animal Hospital Association (AAHA, https://www.aaha.org) guidelines. The treatment followed the signing of informed consent by the cats’ owners. The case study is reported following the ARRIVE guidelines 2.0 (https://arriveguidelines.org/).

Phage-antibiotic treatment

The treatment was administered at the Vet Holim Medical Center and the cat owners’ home under veterinary supervision and in accordance with Israeli law, following the signing of informed consent. The ΦPASB7 suspension (1 mL, 10⁷ PFU/mL, 5 EU/mL) was directly applied to the wound by the cat’s owners, according to the veterinarian’s instructions and demonstration. The phage was given twice daily, in conjunction with changing the wound’s bandage. Additionally, the cat’s owners administered ceftazidime (30 mg/kg) via intramuscular injection four times daily. The wound was sampled and measured using cotton swabs throughout the treatment period.

Detailed case description

Case background

A five-year-old Siamese cat was brought to the JVMC “Vet Holim” animal medical center after a high-rise trauma. The cat was diagnosed with multiple comminuted fractures in both hind legs and impaired surrounding soft tissues and subsequently underwent arthrodesis surgery with internal fixation, debridement, and irrigation (Figure 1). Two weeks post-surgery, necrosis led to the amputation of the left hind leg, while a Pseudomonas aeruginosa implant-associated infection developed in the right hind leg. The infection persisted despite various antibiotic treatments over a four-month period following the operation: amoxicillin, marbofloxacin, doxycycline, cefovecin, and azithromycin. Throughout the antibiotic treatment efforts, the surgical wound remained open, secreting, with the metal implant visible. Prior to considering a complex implant-replacement surgery, adding phage therapy was explored as a potential treatment option.

Figure 1. The cat’s right hind leg post-arthrodesis surgery.

Phage matching: Clinical phage microbiology

Initial screening of the Israeli Phage Bank’s anti-Pseudomonas phage collection against the bacterium revealed 17 phages had an inhibitory effect on the pathogen (Figure 2(A)). Out of the 17 phages, we selected the phage ΦPASB7 for further analysis as it showed good lytic activity in the plaque assay. Further tests revealed that it had a high titer of 10⁹ PFU/mL (Figure 2(B)). Comparing this titer to the titer of phage ΦPASB7 on its reference strain P. aeruginosa PA160, which was used to isolate and characterize it, resulted in efficiency of plating (EOP) value of ∼1. In the growth kinetics assay, ΦPASB7 was able to inhibit the pathogen growth for 15 h after which a small regrowth appeared (Figure 2(C)).

Figure 2. Matching phages and antibiotics to treatment. (A) Plaque assay of the anti-P. aeruginosa phages panel against the cat’s pathogen. ΦPASB7 is marked with an arrow. (B) Plaque assay of the selected ΦPASB7 phage on the cat’s pathogen. (C) Growth kinetics assay (GKA) of the cat’s pathogen treated with ΦPASB, ceftazidime, and the combination of the two agents. The phage-antibiotic combination inhibited the growth of the bacterium for the duration of the test. (D) Colony-forming units count was conducted at the endpoint of the GKA, demonstrating that the phage-antibiotic combination had a synergistic effect that eradicated the bacterium.

ΦPASB7, a short-tail anti-P. aeruginosa phage (Figure 3(A)) was isolated in 2022 from a water sample collected from a fountain in Jerusalem as part of our ongoing phage-hunting efforts. Genomic analysis categorizes this phage within the Schitoviridae family. Its’ genome, consisting of 72,635 base pairs, is a double-stranded DNA structure and comprises 101 predicted genes (Figure 3(B)). Particularly noteworthy is the presence of three RNA polymerase genes, among them a large virion RNA polymerase, a distinctive characteristic within the Schitoviridae family (Wittmann et al. 2020). Additionally, the phage contains six recognized structural genes, as well as the cell membrane disruption genes holin, small proteins that permeabilize the bacterial membrane (Cahill and Young 2019), and spanin, proteins that disrupt the cell membranes by fusing the inner and outer membranes (Cahill and Young 2019), both leading to cell lysis. We could not detect endolysins, which are the third group of common lysing proteins in phages. Worth highlighting is that 71% of the predicted genes remain hypothetical, representing unknown products and functions. We next examined the genome for evidence of lysogenic-associated genes and found no genes related to integrase family proteins nor any recognized phage repressors. Furthermore, BLAST analysis failed to indicate any part of the genome as a component of bacterial genomes.

Figure 3. The phage ΦPASB7. (A) Electron microscopy imaging of ΦPASB7, a short-tailed phage. (B) The genome organization map of phage ΦPASB7 created by PHAROKKA. Each ORF is annotated with its respective transcription direction and highlighted using distinct colors corresponding to their functional roles.

These cumulative insights confidently lead us to select ΦPASB7 for this treatment to be administered alongside the antibiotic ceftazidime. Ceftazidime, a third-generation cephalosporin, is very effective against P. aeruginosa infections and is well-tolerated in cats (Albarellos et al. 2008). The antibiotic was selected due to the low minimal-inhibitory concentration value it had for the pathogen in the antibiogram test (Table 1). Testing ΦPASB7, in combination with ceftazidime, revealed a synergistic effect between the two agents as the phage-antibiotic combination was able to eradicate the pathogen during the 24-h test (Figure 2(D)).

Treatment, monitoring, and outcome

The first dose of phage ΦPASB7 (1 mL, 10⁷ PFU/mL) was topically applied to the wound by the treating veterinarian at Vet Holim Veterinary Medical Center. After a 15-min interval, ceftazidime was administered intramuscularly at a dosage of 30 mg/kg to monitor for any specific side effects of the phage solution. The cat was kept under surveillance for an additional 30 min, during which no adverse effects were observed. All other phage and antibiotic doses were administered at home by the cat’s owners in accordance with the veterinarian’s guidelines. The phage was administered twice daily during bandage changes to minimize the risk of contamination. Simultaneously, ceftazidime was administered four times a day following the standard care protocol. The treatment promptly halted the secretions within the first week of treatment and resulted in a significant reduction (65%) in wound size after a seven-week course, leading to almost complete wound closure, at which point the treatment was stopped (Figure 4).

Figure 4. Treatment. (A) The progress of wound healing during the period of phage therapy. Within the first few days of treatment, the secretions halted, and the wound began to reduce in size. The wound closed completely on day 115. (B) The wound’s size throughout the phage therapy period. (C) The cat’s left hind leg five months after the end of treatment. The wound remains closed, and there are no signs of infection.

However, a month later, the wound began to expand again, and it was accompanied by the reappearance of secretions and a positive P. aeruginosa culture. As a result, ceftazidime was reintroduced and administered twice daily. In addition, a topical application of 0.1% gentamicin cream was applied to the wound twice per day with bandage change to prevent new reinfection. After three weeks without satisfying results, ΦPASB7 was also added as before. Remarkably, within 10 days of the reintroduction of the phage, a complete wound closure was achieved (Figure 4). The treatment was sustained for an additional 2 weeks before being concluded. To date, 8 months after treatment, no signs of infection are evident (Figure 4(C)).

Throughout the period of phage therapy, no side effects were observed. All five P. aeruginosa strains isolated from the wound during the treatment period were tested against ΦPASB7 alone or in combination with ceftazidime to monitor for any emerging resistance to the specific phage-antibiotic treatment. While the P. aeruginosa isolated at the end of the first course of treatment showed a decreased sensitivity to ceftazidime in our in vitro growth kinetic assay compared to the previously isolated strains, it remained highly susceptible to the phage-antibiotic combination (Figure 5(E and F)(. Therefore, the phage-antibiotic combination wasn’t altered during the two courses of treatment.

Figure 5. Sensitivity of the pathogen during the treatment. (A–E) Growth kinetic assays of the P. aeruginosa strains, isolated pre- and during the phage therapy treatment, with the phage ΦPASB7 and ceftazidime. All bacteria showed the same sensitivity to ΦPASB7, which inhibited each strain’s growth for 15–17 h before small bacterial regrowth appeared. The sensitivity to ceftazidime was reduced at the end of the first treatment course (E). The ΦPASB7–ceftazidime combination was able to inhibit the growth of all stains. The multiplicity of infection for all five tests ranged between 1.1 and 1.7. F. Colony-forming units assay performed at the end of the checkerboard assay revealed that the phage-antibiotic combination was able to eradicate all strains regardless of their sensitivity to each of the two agents.

Discussion

In this work, we presented a successful case of phage-antibiotic therapy applied to a P. aeruginosa implant-associated infection in a Siamese cat. The treatment entailed a meticulously designed phage-antibiotic combination specifically tailored to target the pathogen. To the best of our knowledge, this is the first personalized phage-based treatment administered in companion animals.

The prevalence of surgical site infections following orthopedic surgery in companion animals is estimated to be as high as 8.5%, with the emergence of antibiotic-resistant bacteria becoming more prominent in recent years (Husi et al. 2023; Schmökel 2023). These infections pose a substantial burden in terms of morbidity, mortality, and associated costs with surgical procedures (Schmökel 2023; Nicoll et al. 2014). Consequently, there is a critical need for novel adjuvant therapeutics to effectively address this growing concern.

Phages could offer a promising solution to this problem. In recent years, 85% of phage-based treatments for orthopedic infections in humans have involved topically applied phages, resulting in a high success rate (IPTC Study Team 2023; Genevière et al. 2021). Positive outcomes in treating chronically infected wounds and diabetic foot ulcers in humans have also been demonstrated through topical application of phages (Uyttebroek et al. 2022; Young et al. 2023). Therefore, in this case, we chose to take advantage of the direct access to the implant via the open surgical wound and instructed the cat owners to administer the phages directly to it. By applying the phage topically, we aimed to avoid the complexities and costs associated with intravenous treatment and ensure that the phages would reach the infection site at the highest possible concentration.

While the adjunctive use of the phage opens the possibility that the clinical improvement could be attributed to continuous appropriate antibiotic administration, certain factors diminish this likelihood. Specifically, the decreased sensitivity of the pathogen to ceftazidime, as observed in our CPM (5E-F), after the first 7-week course and the failure of wound closure during the second course of treatment when ceftazidime was administered by itself. Moreover, our tests indicated that neither ceftazidime nor ΦPASB7 alone could effectively eradicate the bacteria. However, their combined application achieved successful bacterial eradication in vitro across all cultured isolates throughout the treatment (Figure 5(F)). Moreover, the treatment was efficient even though the sensitivity to ceftazidime alone was reduced. Thus, this study highlights the advantageous synergistic effect attainable through phage-antibiotic combination, further supporting their concurrent use.

The successful conduct of the treatment at home by the cat’s owners provides compelling evidence that personalized phage therapy combined with antibiotics can be effectively extended to pets, who also face the challenges of antimicrobial resistance, similar to humans. A recent survey questioning whether veterinarians and pet owners are open to the possibility of phage therapy showed positive results, indicating that the need is indeed pertinent (Rhys-Davies and Ogden 2020). Phage therapy in companion animals holds the potential to enhance the health and well-being of not only beloved pets but also their owners by preventing zoonotic transmission.

Additionally, extending phage therapy to companion animals can play a crucial role in bridging the gap created by the limited number of phage clinical trials, as most data regarding the safety and efficacy of phages are currently derived from case reports of compassionate use (Uyttebroek et al. 2022; Altamirano and Barr 2019). It will allow phage centers to draw conclusions regarding potential benefits and limitations of phage therapy against a wide range of bacterial infections, some of which may be less common in humans and refine phage-matching, phage-adapting, and treatment protocols accordingly. The resulting data can be applied to both veterinary and human cases, leading to improved treatment outcomes. Phage therapy in pets can educate both healthcare professionals and the public about the potential benefits of this treatment approach. Increased awareness to phage therapy among veterinarians and pet owners can lead to greater acceptance and adoption of phages in various medical settings, paving the way for phages to be accepted as a standard medicine for infectious diseases.

Author contributions statement

The contributions of each author were as follows: Conception and design: RB, GH, OY, AB, RH. Preparation of phages for treatments: RB, OY, SOA, AR. Cat treatment: GH, AB. Treatment management: ON, HM. Analysis and interpretation of the data: RB, GH, OY, AB, RH. The drafting of the paper: RB, SCG, RH. All authors agree to be accountable for all aspects of the work.

Acknowledgments

We would like to extend our gratitude to the cats’ owners, Larry, Ana, and Milette Barcly, for willingly participating in this unique treatment and for their full compliance throughout the process. Additionally, we express our appreciation to Eddie Bernstein for preparing and providing the electron microscope visualization images.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The authors confirm that the data supporting the findings of this study are available within the article.

Additional information

Funding

This work was supported by the EveryCat Health Foundation, Wyckoff, New Jersey, under GRANT EC23—0000000060.