Investigation of orally administrable bacteria for cancer therapy

Diminished progress in improving prognoses and treatment outcomes of several cancers, using conventional drugs, is being countered by investigations into a range of new, advanced formulations, of which bacteria-based anticancer therapies are one. Bacteria-based treatments hold an especially intriguing place within this group of disruptive medical technologies for the well-documented ability of bacteria to selectively colonise tumours following IV administration and replicate therein. Based on this property, bacteria have been studied both clinically and preclinically to evaluate their potential to promote tumour regression and confine noxious effects to the malignant tissue. Some of these studies have attempted to engineer the bacteria genetically to reduce their virulence and/or accentuate their capability to exert tumour-selective damage, but efforts to date have failed to penetrate clinical barriers and produce a safe, effective, and marketable bacteria-based product for cancer therapy. This thesis examines ways in which bacteria that are truly optimised for cancer treatment might be acquired and/or developed, using data derived from patient (Chapters 2, 3, and 4) and preclinical studies (Chapters 3, 4, and 5). The thesis can be divided into three main hypotheses: i) if naturally occurring bacteria can be identified in human tumour tissues, then these may be better tumour-targeting agents than the laboratory and probiotic strains that are routinely used in tumour-targeting studies (Chapters 2 and 3); ii) if bacteria that are administered orally to cancer patients can successfully colonise their tumours, then these may be safer and equally effective alternatives to intravenously administered microbes (Chapter 4); and iii) if bacteria are continually recycled through tumours via IV administration, then this may result in strains with improved tumour selectivity over their parental counterparts (Chapter 5). A prevailing concept throughout this work is that exposure to tumours, either via purposeful human intervention or natural colonisation, and the resultant evolutionary effects could adapt bacteria to the intratumoural environment more effectively than what genetic engineering would allow. This thesis contributes considerably to several bodies of knowledge, including the human tumour microbiota and the feasibility of administrable bacteria as vehicles for tumour-selective transmission of therapeutics. Chapters 2 and 3 use sequencing and culture techniques to develop an accurate and comprehensive picture of the human breast tumour microbiota. A culture-based assay was able to define the microbial diversity and community structure of breast tumours, when placed in the context of refined bioinformatic analyses. Despite a lack of evidence for viable, endogenous, tumour-selective bacteria, breast tumours contained a characteristic microbiological signature, as determined by deep sequencing, that warrants further exploration in varying tumour types to establish the diagnostic potential of this finding. Chapter 4 reports clinical data in relation to the translocation of probiotic bacteria from the GIT to distal tumours, using both FFPE specimens and fresh tissues. Results of this study, and corresponding animal experiments, shed light on the viability, or lack thereof, of this route of administration for prospective bacterial tumour-targeting studies in the clinic. Lastly, Chapter 5 is an account of the effects of tumour-to-tumour cycling on probiotic bacterial strains. The data in this chapter indicate that recycling bacteria through murine tumours can alter intratumoural growth dynamics and metabolic capacity, which was reflected in enhanced tumour persistence and biofilm formation. This work emphasises tumour recycling as a method by which bacteria can be adapted to the complex, heterogenous environment of cancerous tissue in a relatively simplistic way, which may have implications for the success of this treatment modality as it gains more clinical traction