Summary

Microbial-based cancer treatments are an emerging field, with multiple bacterial species evaluated in animal models and some advancing to clinical trials. Noninvasive bacteria-specific imaging approaches can potentially support the development and clinical translation of bacteria-based cancer treatments by assessing the tumour and off-target bacterial colonization. Noninvasive, bacteria-specific, positron emission tomography (PET) imaging has the potential to support the development and translation of tumour-targeting bacteria by providing information on the localization and proliferation of bacteria at the tumour site while also monitoring bacterial colonization in other organs (off-target effect).

¹⁸F-Fluorodeoxysorbitol (¹⁸F-FDS) has been characterized as a bacteria-specific PET tracer in animal studies to specifically localize and monitor infections due to the Enterobacterales order of gram-negative bacteria, which include gram-negative bacteria such as Escherichia coli, Klebsiella spp., Salmonella spp., Serratia spp., Yersinia spp., etc.

It was evaluated whether ¹⁸F-FDS PET could be used to accurately monitor colonization of breast cancer tumours by a genetically modified, attenuated strain of Yersinia enterocolitica T3P-Y004 lacking effector proteins (the “Yops”). Y. enterocolitica T3P-Y004 was found to selectively colonize solid tumours in multiple models, and a T3P-Y004 strain engineered to deliver therapeutic proteins is currently in clinical development (NCT05120596).

Results from nanoScan® PET/CT

15-minute PET acquisition and CT were performed using the nanoScan PET/CT (Mediso). PET imaging was performed 120 minutes after intravenous injection of 7.8 MBq of ¹⁸F-FDS. ¹⁸F-FDS PET/CT was performed at postadministration days −1 (baseline preadministration), 2 and 4 for the intratumour group, or at days −1, 5, and 10 for the intravenous group. ¹⁸F-FDG PET/CT (PET acquisition performed 45 minutes after intravenous injection of 7.8 MBq) was used as a control in the intravenous group and performed at days −2, 4, and 9 in the same animals. Animals not injected with bacteria were used as controls. 

Figure 1. Experimental outline. A, Following the injection of Yersinia enterocolitica T3P-Y004 into the triceps of mice, the animals were injected with ¹⁸F-FDS and imaged with PET/CT. Tissues were harvested for gamma counting and to quantify bacterial burden. To determine the ability of ¹⁸F-FDS PET/CT to localize the bacteria, animals were injected with Y. enterocolitica T3P-Y004 directly into the tumour (B) or intravenously (C). Subsequent ¹⁸F-FDS or ¹⁸F-FDG PET/CT scans were performed in the same cohort of animals. 

To quantify the PET signal, spherical volumes of interest (VOIs) were drawn at the sites of pathology and unaffected sites of the same tissue (for the myositis model) to calculate the target-to-nontarget tissue ratio with the CT as reference using AMIDE 1.0.4. For the tumour models, the whole tumour volume was measured using the CT as a reference. Graphical representations were obtained with VivoQuant 2020 (Invicro). Quantification of PET signal in all organs is represented as mean standardized uptake value (SUV_mean).

The in vitro accumulation of ¹⁸F-FDS in Y. enterocolitica T3P-Y004 was 49.25% (SD 3.46%) and 63.83% (SD 3.93%) after 120 minutes of incubation at 28°C and 37°C, respectively (Supplementary Figure 1) and comparable to the reference strains. In a mouse model of myositis, the ¹⁸F-FDS PET signal at the site of Y. enterocolitica T3P-Y004 infection (right triceps) was significantly higher compared to the contralateral muscle (left triceps) with sterile inflammation (target-to-nontarget tissue ratio 14.52; interquartile range [IQR], 13.48–15.60; P = .028; Supplementary Figure 1).

Supplementary Figure 1. In vitro uptake of ¹⁸F-FDS and myositis model. Y. enterocolitica T3P-Y004 and a reference strain of E. coli were incubated in vitro with ¹⁸F-FDS over 120 min at 28°C (A) or 37°C (B). Data represented as % uptake with a minimum of six replicates per condition per time point. (C) The in vitro accumulation of ¹⁸F-FDS was also evaluated in a reference strain of Y. enterocolitica (ATCC 23715) at 37°C. (D) Mice were injected with Y. enterocolitica T3P-Y004 in the right triceps (yellow arrow) and heat-killed bacteria in the opposite triceps (red arrow). After 20 hours of incubation, the animals were imaged with ¹⁸F-FDS PET/CT. The maximum intensity projection, coronal, transverse, and sagittal views of a representative mouse are shown. (E) The PET signal was quantified and represented as SUV_mean. Statistical comparison made with two-tailed Mann-Whitney test, n = 4 animals.

These studies demonstrated that Y. enterocolitica showed robust ¹⁸F-FDS uptake and whole- body ¹⁸F-FDS PET was able to differentiate the tumors with Y. enterocolitica colonization from those not colonized, in murine models utilizing direct intratumour or intravenous infection.

There was heterogeneity of ¹⁸F-FDS uptake in the tumour, which is in line with bacterial heterogeneity noted in vivo by other groups. To account for this heterogeneous distribution, the VOIs to calculate tumour SUV_mean were drawn to include the whole tumour volume based on CT.

Figure 2. Intratumour model and ¹⁸F-FDS PET/CT. A, MIP, coronal, sagittal, and transverse sections of a representative mouse injected with bacteria directly into the tumour and imaged with ¹⁸F-FDS PET/CT on day 4. The ¹⁸F-FDS signal can be visualized in the tumour site with minimal background in other organs. B, Quantification of the ¹⁸F-FDS PET signal represented as SUV_mean (error bars represent interquartile range). On days 2 and 4, the ¹⁸F-FDS PET signal was 1.6 times and 2.9 times higher in the animals injected with bacteria, compared to controls (n = 7 animals injected with bacteria and 3 controls, Mann-Whitney test, 2-tailed, *P = .016). Abbreviations: CT, computed tomography; ¹⁸F-FDS, ¹⁸F-fluorodeoxysorbitol; MIP, maximum intensity projection; SUV_mean, mean standardized uptake value.

Figure 3. Intravenous model evaluated with ¹⁸F-FDS PET/CT. A, MIP, coronal, sagittal, and transverse sections of a representative mouse injected intravenously with bacteria and imaged with ¹⁸F-FDS PET/CT on day 5. The ¹⁸F-FDS signal can be visualized in the tumour site with minimal background in other organs. B, MIP, coronal, sagittal, and transverse sections of a control mouse imaged with ¹⁸F-FDS PET/CT on day 5. C, Quantification of the ¹⁸F-FDS PET signal represented as SUV_mean (error bars represent interquartile range). For ¹⁸F-FDS on days 5 and 10, the ¹⁸F-FDS PET signal was 3.2 times and 2.4 times higher in the animals injected with bacteria, compared to controls (n = 7 animals injected with bacteria and 3 controls, Mann-Whitney test, 2-tailed, *P = .016).

• enterocolitica demonstrated excellent ¹⁸F-FDS uptake in in vitro assays.
• Whole-body ¹⁸F-FDS PET demonstrated a significantly higher PET signal in tumours with enterocolitica colonization compared to those not colonized
• ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) PET signal was not different in enterocolitica colonized versus uncolonized tumours.
• ¹⁸F-FDS PET could be utilized as a complementary approach supporting the development and clinical translation of enterocolitica-based tumour-targeting bacterial therapeutics.
• ¹⁸F-Fluorodeoxysorbitol positron emission tomography (PET), a bacteria-specific imaging approach, was used to visualize an attenuated strain of Yersinia enterocolitica, currently in clinical trials as a microbial-based cancer treatment, in murine models of breast cancer.

Full article on academic.oup.com