Is Imaging of Bacteria with PET a realistic goal?

Introduction: Since FDG was introduced as a PET radiotracer, significant efforts have been made to develop other more specific radiotracers and increase spatial resolution from centimeter range to millimeters. Taking it a step further, direct PET imaging of bacterial presence, severity, and progression at the site of infection has gained popularity over the last decades, and this would indeed be clinically desirable. However, there have been mixed experiences among researchers. Through a literature review, we aim to investigate the opportunities and constraints of whether imaging bacteria with PET is a realistic goal or a misconception.

Methods: We searched Google Scholar and PubMed databases for literature on identifying bacteria using PET scans using keywords including “PET, Imaging, Bacteria, and Infection.” We read and analyzed the most cited articles thoroughly in keeping with our objectives and included literature outside this list to ensure that the limitations of studies aiming to identify bacteria with PET were sufficiently covered. We compared the opportunities and limitations of the subject and made results and conclusions based on the data obtained from those articles.

Results: Most infections cause inflammation, which is successfully visualized by ¹⁸F-FDG PET, and FDG and PET/CT are already clinically important for the diagnosis of infection, monitoring response to treatment, and overall improving patient care. FDG does not image the microorganism but rather the downstream immune activation. Thus, it is a nonspecific tracer of activated immune cells at the site of inflammation, no matter if the underlying infection is bacterial, fungal, parasitic, or viral. It should be considered advantageous as clinical symptoms are usually nonspecific, and the underlying etiology is usually unknown at the initial workup. Nonetheless, as more specific tracers like radiolabeled white blood cells, [⁶⁸Ga] citrate, and Ubiquicidin peptides are used clinically, others are being explored at the pre-clinical stage; as most biological molecules can be labeled, researchers are looking for more specific tracers by targeting biochemical characteristics of living bacteria or by labeling antimicrobial agents. However, this may have limited clinical usefulness as specific types of tracers are required for any particular class or group of bacteria, and it is practically impossible to image all patients with a large panel of tracers. Moreover, antibiotic resistance has challenged the use of antimicrobial agents as radiotracers. Further, owing to the physical limitations of PET, a gross imaging modality, the claims of identifying small quanta of bacteria at a cellular and subcellular level are questionable. Hence, it is not surprising that results similar to that of ¹⁸F-FDG PET for detecting inflammation have not been achieved by attempts to visualize bacteria with the help of various tracers at the sites of infection directly. Indeed, results of most of these tracers studied in animal models have not been successfully translated to humans, and a recent systematic review found significant heterogeneity in study designs with little or no consensus on a basic methodology which hampers general comparability and applicability.

Conclusions: Specific tracers may have merit in some settings but should not always be too specific as this, together with physical limitations of PET imaging resolution, may limit clinical use. The large number of tracers being investigated for decades to visualize specific types or groups of bacteria directly may provide scientifically interesting knowledge, but until now, with limited success in translating results to humans, in part due to lack of standardization. Thus, the clinical use remains limited presently, and research resources could arguably be put to better use in further exploring the much more versatile FDG.