Antimicrobial resistance (AMR) is a significant and growing threat to public health. When
microorganisms (such as bacteria) develop and change in ways that make them resistant to previously
effective treatments (for example, antibiotics), there can be significant impacts on human health. AMR is
particularly problematic in hospitals, where it can impact highly vulnerable patients, such as premature
infants in the neonatal intensive care unit. When treating infections in hospitals, healthcare providers look
at recent patterns from human test results over the past 1 or 2 years to understand which resistances might
be present; this data is provided in reports called antibiograms. However, these antibiograms have
limitations, and novel solutions are needed.

Environmental surveillance has emerged as a powerful approach for infectious disease surveillance.
Taking swab samples from the floor can provide estimates of infectious disease burden within a wing of
a building or ward of a hospital. We propose to conduct a pilot study to collect floor swabs from the
neonatal intensive care unit at Mount Sinai Hospital for the surveillance of local AMR prevalences. DNA
will be extracted from the floor swabs, and will undergo metagenomic and targeted metagenomic
sequencing. Through the sequencing, we will detect antimicrobial resistance genes and organisms of
interest. This proposed non-invasive technique will help complement existing antibiograms as a more
timely, less invasive, and more spatially refined approach. Environmental AMR surveillance has the
potential to provide valuable information for infection control, gauging the risk of specific resistant
infections and informing responsible, effective prescribing.

There is an “arms-race” between bacteria and humans. Bacteria develop ways of “resisting” these antimicrobials (aka. antimicrobial resistance [AMR]), and humans develop ways of overcoming these bacterial mechanisms (e.g. alternative antimicrobial, bacterial enzyme inhibitors etc). WHO describes AMR as a “silent pandemic” and one of the top ten most crucial global health problems. AMR is responsible for ~ 1.3 million deaths/year, growing to 30 million by 2050. AMR can result in resistance to most or all antimicrobials, rendering no alternatives. Carbapenems are considered the ‘last resort antibiotics’ for the treatment of severe Gram-negative infections, however, resistance through enzymes referred to as carbapenemases are becoming common. Canada is not exempt from this; CNISP/TIBDN report within Canada, carbapenemases have more than doubled in a few years, and are almost doubling every two years within the Toronto-Peel Region. Within the region, Peel has double the incidence versus Toronto, and Trillium Health Partners (THP) have the second highest incidence within Toronto-Peel.

Bacteria that harbour carbapenemases are detrimental to patient outcomes and drastically affect the economy of healthcare systems. Genetic material containing carbapenemases and additional AMR genes can easily spread between different bacterial species within the same patient, between patients, and from the environment-to-humans. Therefore, early detection and prevention of spread is critical. In a hospital setting, a “resistome” may be one intervention; monitoring and investigating known and novel AMR determinants may act as an early warning system in the hospital environment or patients, allowing early intervention strategies. Most current methods require expertise that is rarely available in a hospital environment. We propose to validate a method of AMR detection using whole genome sequencing technology by our collaborators, THP, the National Microbiology Laboratory (NML), and Public Health Ontario (PHO) to implement in a large community hospital and potentially more remote settings.

Latently infected immune cells in people with HIV are the main obstacle to cure despite improved treatment because this virus can integrate its genes into our DNA. We will attempt to silence such integrated HIV by harnessing a molecular pathway that our cells evolved against genomic intruders like HIV. This pathway consists of PIWI proteins and their guide molecules called piRNA. This duo can recognise, tag, and permanently silence genes of molecular invaders, such as endogenous retroviruses, and degrade their products. Information about potential targets for PIWI-mediated destruction is stored in our DNA in the form of piRNA clusters — essentially a ‘list of potential invaders’. We hypothesize that the PIWI pathway could be harnessed for use against HIV in several different ways by targeting the viral genes in latently infected immune cells with synthetic anti-HIV piRNA, and by programming piRNA clusters to produce ‘natural’ anti-HIV piRNA. As a proof of concept, we will use synthetic piRNA to target a T cell line called J-Lat that harbours HIV genes along with a gene of green fluorescent protein (GFP). By targeting GFP, we can measure changes in the amount of light it emits when the HIV genes are active. We have already achieved some initial success with this approach. The next step would be to introduce a GFP sequence into an active piRNA cluster and let the cell produce its own anti-GFP piRNA. If successful, we will then test the most promising anti-HIV piRNA candidates in a similar manner.

Metapopulation models are an emerging method for incorporating important community-level differences, such as sociodemographics, social contact patterns, disease burden, and seasonality, into mathematical models of infectious diseases. Despite gaining attention for their ability to offer valuable insights during the COVID-19 pandemic, particularly in the United States and Europe, these methods have received limited attention to date in Canada. However, a crucial first step to developing these models is identifying an appropriate framework for incorporating between-community mobility patterns. There are many options available, and the choice of framework has significant implicants for a model’s epidemiological accuracy.
Our objective is to compare the ability of two common statistical models (i.e., gravity, radiation) to accurately capture travel patterns between Ontario’s 34 public health units (PHUs). We will compare model-based estimates of between-PHU travel frequency to empiric de-identified 2021 Census data capturing workplace mobility using common statistical measures of agreement (e.g., percent agreement, Cohen’s kappa), reliability (e.g., interclass correlation coefficients), and accuracy (e.g., root mean squared error).

This research will provide a foundation for the development of infectious disease metapopulations and local Canadian capacity in this emerging methodological area. Most immediately, this will support our team’s planned work to develop a new metapopulation model for respiratory syncytial virus (RSV) to support the evidence-informed selection of emerging immunization strategies to reduce RSV disease in infancy. However, this work also offers a foundation for developing metapopulation models for other infectious diseases (e.g., measles, emerging viruses of pandemic potential) and may be replicated in other provinces/territories.