The overall aim of the research outlined in this PhD thesis is to assess the feasibility and acceptability of early infant male circumcision (EIMC) as an HIV prevention intervention in Zimbabwe in order to inform roll out. Mathematical modelling estimates that circumcising 1.9 million Zimbabwean men aged 15-49 by 2015 could avert 42% of new HIV infections that would have otherwise occurred by 2025. Since 2009, Zimbabwe has provided voluntary medical male circumcision (VMMC) to over 300,000 adult and adolescent men. In order to ensure that the protective effect of male circumcision is sustained in the longer-term, Zimbabwe intends to roll out EIMC alongside adult MC, starting 2015. Although EIMC’s effects on HIV will take longer to realise, infant circumcision is easier, safer and cheaper than adult MC. Further, EIMC may more effectively prevent HIV acquisition as the procedure is carried out before the individual becomes sexually active, negating the risk associated with acquisition or transmission of HIV during the healing period. Since large-scale EIMC for HIV prevention, or indeed for other reasons, has never been practised in Zimbabwe or more widely in Southern Africa, there are concerns around its feasibility and acceptability. Clearly, acceptability of infant MC will have a bearing on uptake, roll out and subsequent effectiveness in preventing HIV. In Zimbabwe, there are also concerns about the feasibility of rolling out EIMC for HIV prevention within the context of existing health services, many of which are already overburdened and understaffed. The PhD research is in two phases. The first phase describes a systematic review and thematic synthesis I conducted to explore parental reasons for non-adoption of infant MC for HIV prevention in sub-Saharan Africa. Additionally, this phase qualitatively explored hypothetical acceptability of EIMC among parents and wider family as well as hypothetical feasibility and acceptability of EIMC among health-care workers. Findings from the first phase informed the design of a study to pilot EIMC roll out. The second phase was nested within a trial that assessed the feasibility, safety, acceptability and cost of rolling out EIMC using devices in Zimbabwe. It explored actual acceptability of EIMC among parents and wider family as well as actual feasibility and acceptability of EIMC among health-care workers. Findings from both phases informed recommendations for a demand generation intervention for EIMC which is currently being developed and will subsequently be tested for impact. Given that EIMC has been identified as a key HIV prevention intervention for sustaining the prevention gains anticipated through VMMC across sub-Saharan Africa, the findings of this research are likely to have broad implications for HIV prevention across the region.

Antiretroviral therapy (ART) has transformed HIV infection from a death sentence into a chronic condition. In sub-Saharan Africa, the area most affected by this disease, availability of ART has increased dramatically over the last few years. Nevertheless, many people are still not receiving ART either because they are not aware of being HIV-positive or because they struggle to access ART or to engage in HIV care. It is fundamental to take decisions which maximise the health benefits with the limited resources available. When I was writing this thesis, there were countless discussions regarding whether the recommendation on when to start ART had to be modified to a CD4 count threshold higher than 350 cells/μL, given the compelling evidence that ART reduces substantially the risk of transmission in heterosexual serodifferent couples. In this thesis I evaluated the effectiveness and cost-effectiveness of alternative ways of increasing the number of adults receiving ART in South Africa: increasing the CD4 count threshold at which a person is eligible to be initiated on ART, or maintaining the eligibility criteria to CD4 count below 350 cells/μL but expanding the number of people who are diagnosed and engaged in care. In particular, I focused on the impact these two alternatives would have on the development and transmission of resistance. To inform the model on the extent to which NNRTI resistance mutations are present in people who have interrupted NNRTI, I conducted an analysis using data from the UK resistance database. In addition, since I found that the most cost-effective strategy was to expand the number of people engaged in HIV care without modifying the CD4 threshold at which a person is eligible to receive ART, I evaluated at which steps in the current leaky cascade of HIV care it was most cost-effective to intervene. Finally, as new evidence regarding the accuracy and acceptability of HIV self-testing came up, I decided to evaluate the cost-effectiveness of introducing HIV self-testing in a setting such as Zimbabwe.

Computed tomographic colonography (CTC) is a technique employed to examine the whole colon for cancers and premalignant adenomas (polyps). Oral preparation is taken to fully cleanse the colon, and gas insufflation maximises the attenuation contrast between the enoluminal colon surface and the lumen. The procedure is performed routinely with the patient both prone and supine to redistribute gas and residue. This helps to differentiate fixed colonic pathology from mobile faecal residue and also helps discover pathology occluded by retained fluid or luminal collapse. Matching corresponding endoluminal surface locations with the patient in the prone and supine positions is therefore an essential aspect of interpretation by radiologists; however, interpretation can be difficult and time consuming due to the considerable colonic deformations that occur during repositioning. Hence, a method for automated registration has the potential to improve efficiency and diagnostic accuracy. I propose a novel method to establish correspondence between prone and supine CT colonography acquisitions automatically. The problem is first simplified by detecting haustral folds which are elongated ridgelike endoluminal structures and can be identified by curvature based measurements. These are subsequently matched using appearance based features, and their relative geometric relationships. It is shown that these matches can be used to find correspondence along the full length of the colon, but may also be used in conjunction with other registration methods to achieve a more robust and accurate result, explicitly addressing the problem of colonic collapse. The potential clinical value of this method has been assessed in an external clinical validation, and the application to follow-up CTC surveillance has been investigated. MRI has recently been applied as a tool to quantitatively evaluate the therapeutic response to therapy in patients with Crohn's disease, and is the preferred choice for repeated imaging. A primary biomarker for this evaluation is the measurement of variations of bowel wall thickness on changing from the active phase of the disease to remission; however, a poor level of interobserver agreement of measured thickness is reported and therefore a system for accurate, robust and reproducible measurements is desirable. I propose a novel method which will automatically track sections of colon, by estimating the positions of elliptical cross sections. Subsequently, estimation of the positions of the inner and outer bowel walls are made based on image gradient information and therefore a thickness measurement value can be extracted.

In this thesis, the challenges of establishing the very first human in vivo 23Na magnetic resonance imaging in the United Kingdom are presented. A comprehensive framework for quantitative human in vivo studies is established and translated for clinical research imaging. Quantitative measures to obtain the sodium bioscale using external calibrants are discussed and results from a scan re-scan reproducibility study on five healthy volunteers are presented. The protocol was subsequently used in a clinical research study and published by a clinical collaborator in a high impact journal. Improvements in acquisition are achieved by implementation and 23Na adapt- ation of the state of the art 3D-Cones pulse sequence. For evaluation, it is compared against more established 3D-radial k-space sampling schemes featuring cylindrical stack-of-stars (SOS) and 3D-spokes kooshball trajectories on five healthy volunteers in a clinical setting and numerical phantoms. Signal- to-noise ratio (SNR) as a measurement of sequence performance was compared between the sequences and the results are presented. The results were published in a special issue on X-nuclei imaging in the journal of Magnetic Resonance Materials in Physics, Biology and Medicine. The work was subsequently shortlisted and presented for the Young Investigator Awards at the annual meeting of the European Society for Magnetic Resonance in Medicine and Biology. Reconstruction improvements by means of sophisticated k-space weighting schemes are presented on numerical and in vivo data and its effects on image appearance, SNR and total tissue sodium concentration estimates are discussed. The work is currently in peer review for journal publication. A protocol for clinically feasible in vivo 23Na relaxometry measurements of the transverse relaxation time constant T2 is established and results for a range of anatomical white and grey matter locations is presented using both a bi-exponential two-component fit and an unrestricted continuous distribution model. The implications of the results on the underlying tissue sodium environment are discussed. This work has subsequently been presented at international conferences of the International Society for Magnetic Resonance in Medicine and European Society for Magnetic Resonance in Medicine and Biology and has been submitted for peer review as a journal publication. As a conclusion I discuss how the methods presented here can be used to obtain unprecedented spatial and temporal resolution in in vivo 23Na imaging at 3T. Preliminary results are presented.

Spinal and bulbar muscular atrophy (SBMA), also known as Kennedy’s Disease, is a late-onset, X-linked, progressive neuromuscular disease, which predominantly affects males. There is currently no effective treatment for this disease. SBMA is caused by a CAG repeat expansion in the gene that encodes the androgen receptor (AR) protein. Disease manifestation is androgen dependent and results at least in part from a toxic gain of AR function. AR100 transgenic mice which carry 100 CAG repeats in the human AR gene develop lower motor neuron degeneration with accompanying progressive neuromuscular phenotype, which closely recapitulates the human disease. In this Thesis, a longitudinal physiological and histological characterisation of disease progression in AR100 mice was conducted. The experiments described show that the disease first manifests in skeletal muscle, prior to any motor neuron degeneration, which only occurs in late stage disease. These findings challenge the prevailing view that SBMA is primarily a motor neuron disorder, and suggest that muscle-targeted therapeutics may be effective in SBMA. The effects of a potential therapeutic agent on disease progression in AR100 mice were examined. SBMA is a hereditary disease in which individuals carrying the mutation can be genetically identified prior to symptom onset, allowing for commencement of early treatment. Therefore AR100 mice were treated with arimoclomol, a pharmacological co-inducer of the heat shock response (HSR), from the point at which motor deficits first manifest. Treatment of mice with arimoclomol had no beneficial effect on disease. These results are surprising, as treatment of AR100 mice with arimoclomol from a post-symptomatic time point has previously been shown to modify disease progression. The final Results Chapter describes the development of MRI to monitor disease progression in AR100 mice and evaluates its potential as an outcome measure in preclinical trials, focusing on imaging of lower hindlimb muscles. This technique is directly translatable to the clinical setting and may provide a more direct transition between preclinical trials and patients.

Mixed methods investigate the association between HIV, ART and fertility following scale-up of HIV treatment and care in South Africa. Two longitudinal analyses of surveillance data from the Africa Centre for Health and Population Studies compare factors associated with live birth by HIV and ART exposure. Semi-structured interviews with women enrolled on ART and healthcare providers explore perceptions of childbearing and contraceptive use. A quantitative study reports on one open cohort analysis and a subsequent closed cohort. Crude Birth Rates declined since 2005 and an open cohort analysis (2007-2013) found consistently lower birth-rates amongst women on-ART, compared to HIV-positive ART-naïve women and HIV negative women. One exception was found in the 25-29year age group: incidence was 38% higher to women on ART than ART-naïve women. Crude incidence of live birth was 6.6 births/100 women-years and decreased with increasing age, higher parity, poorer self-reported health, urban area of residency, knowledge of own HIV status, being single or engaged/married, not living with a partner, awareness of the benefits of ART, use of contraception and use of injectable methods. Annual likelihood (aHR, 0.39; 95% 0.347 – 0.441) was 61% lower to HIV positive versus negative women in multivariable Poisson analysis and exposure to ART was associated with 38% reduced likelihood (aHR, 0.62; 95% 0.487 – 0.799). In a subsequent closed cohort, HIV ‘unknown’ women demonstrated similar incidence and associated factors of live birth compared to HIV negative women. HIV positive women were less likely than HIV negative and HIV ‘unknown’ women to use contraception. Women described inconsistent injectable use in semi-structured interviews due to side effects and perceptions that injectables make women ‘watery’ or are unnecessary on ART. Family planning counselling was under-prioritised within the health care service and women were unaware of safer conception topics. Recent pregnancies were considered unintended and most women desired to avoid childbearing considering current family size, economic and health risks. Partner expectations could override strong concerns for health, however, and HIV positive women were at similar risk of live birth to HIV negative women when in a regular relationship or living with a partner.

Auricular reconstruction using sculpted autologous costal cartilage following congenital microtia, trauma and cancer is effective, but complex, time consuming and may incur donor site morbidity. Current conventional synthetic alternatives are associated with infection and extrusion. A novel nanocomposite polymer based on polyhedral oligomeric silsesquioxane (POSS) covalently bonded to poly(carbonate-urea-urethane) (PCU), has developed and patented in our laboratory and displayed promising clinical applications in first-in-human. POSS-PCU has been utilised to develop synthetic 3D auricle construct for use as auricular cartilage replacement, which aims to reduce extrusion rates by mimicking the mechanical properties of human ears and by encouraging desirable cellular interactions. Methods & Material: A novel custom-made ear-shaped 3D negative glass-mould was fabricated from positive 3D-printing ear model of a CT scan obtained from the external part of the human ear. This glass-mould was used to manufacture 3D auricular scaffolds using POSS-PCU materials. Two novel fabrication techniques solvent-evaporation/porogen-leaching (POSS-PCUs) and phase-separation/porogen-leaching (POSS-PCUc) were developed to produce nanocomposite scaffolds. The fabrication, physicochemical properties (including nanoscale topography) and in-vitro cellular interactions of these novel fabricated nanocomposites were extensively studied and compared to current commercially available synthetic material (Medpor®). Further cell-material interactions were completed on the optimized POSS-PCUs and Medpor® to examine the angiogenic and inflammatory responses to these materials using human fibroblasts and macrophages. Finally, POSS-PCUs scaffolds with two different pore sizes ranging from 50-100µm & 150-250µm were subcutaneously evaluated compared with Medpor® in-vivo in a rodent model for up-to 3 months. Results: The glass-mould design using 3D-printing technology allowed the manufacture of customised auricular POSS-PCU scaffolds that replicated the original human auricle. Two different polymeric fabrication methods resulted in the production of two types of nanocomposite scaffolds with same average pore size (150µm) but different porosity (63.47±1.35% in POSS-PCUs & 85.21±1.19% in POSS-PCUc). POSS-PCUs had similar elastic modulus (5.73±0.17 MPa) to human ear cartilage (5.02±0.17 MPa) compared with POSS-PCUc (0.58±0.12 MPa) and Medpor® (140.9±0.04 MPa) elastic modulus which were significantly different (P<0.001). Optimised POSS-PCUs nanocompoiste showed greater protein adsorption, and subsequently increased fibroblast adhesion, growth, and collagen production in vitro, compared to Medpor®. A higher volume fraction of fibrotic tissue ingrowth with negligible capsular formation was achieved in vivo in POSS-PCUs with pore-sizes of 150-250µm than Medpor®. No evidence of chronic inflammation was observed in the implanted scaffolds. Conclusion: These findings as well as other clinical studies in man suggest that POSS-PCUs nanocomposite can be a promising biomaterial for auricular reconstruction.

Monitoring patient health in low and middle-income country HIV care programmes is challenging, as, without evidence, measurement tools derived from high-income country studies have been adapted and paper-based monitoring systems quickly developed. An accurate understanding of the population in care may be compromised. This thesis examines aspects of HIV care: access to Cotrimoxazole preventive therapy (CPT), prevalence of common mental disorders (CMD), and tools used to measure outcomes on antiretroviral therapy (ART). CPT access is frequently cited as being as low as 4% with few studies estimating long-term access. Estimated prevalence of CMD varies widely as little standardisation in measurement tools exists. And, while international ART programme monitoring recommendations exist, no study has compared the concordance, or otherwise, between information collected in different countries. The first study in this thesis, in Ugandan and Tanzanian patients, estimates time from HIV diagnosis to CPT initiation, time spent on CPT and associated factors. These estimates are compared to reported data. CPT coverage and time on CPT were poor. The absence of unique patient identifiers means monitoring data cannot distinguish patients who were diagnosed and initiate CPT in different reporting periods. Furthermore, no long-term data are officially reported. The second study estimates CMD prevalence and associated factors among HIV-positive Ugandans, and validates measurement tools for this. Prevalence was around 10% but no routinely-collected data identified at-risk patients. Measurement tool validity was poor, and their use substantially overestimates prevalence. The third study compares ART programme monitoring systems in Malawi, Uganda, Ukraine and Tanzania. There was little concordance with international recommendations, and discordance in additional data-items and paediatric age-groupings. This signalled a lack of understanding of how best to monitor the health of treated populations. Finally, a fourth study is proposed with the aim of assessing the validity and predictive value of existing programmatic monitoring systems.

Regulatory B cells (Bregs), identified by this laboratory as transitional-2 marginal zone precursor cells (T2-MZPs), restrain immune-driven diseases by inhibiting pathogenic T cell differentiation via the provision of IL-10. Despite intense investigation into their effector function, little is known regarding the cellular processes that govern Breg differentiation. Here I show that arthritic inflammation controlled by the gut-microflora promotes the differentiation of splenic Bregs. Drastic alteration of the inflammatory process imposed by anti-TNFα treatment or changes in the gut-microflora caused a deficiency in the production of IL-1β and IL-6, and concurrently in the number and functional capacity of Bregs. Furthermore, Bregs induced by IL-1β and IL-6 were critical in limiting excessive inflammatory responses, as disruption of IL-1R1 and IL-6R exclusively on B cells lead to the development of an exacerbated arthritis and reduced IL-10 production by splenic B cells. These results led to the question of how the gut-microflora influences the differentiation of splenic-Bregs? I found that splenic-T2-MZPs, expressed the gut-homing marker α4β7, and were also detectable in the mesenteric lymph nodes (MLN). Furthermore, similarly to the spleen, changes imposed on the microbiome by antibiotics treatment caused a reduction in both the number of IL-6+ cells and IL-10+Bregs in the MLN. Interestingly, further analysis revealed that MLN-T2-MZPs were not exposed to IL-1β in the MLN and were only-partially suppressive when compared to their splenic counterparts. Collectively, this data suggests a possible scenario where Breg differentiation is initiated in the MLN following exposure to gut-microbiota induced IL-6, but B cells do not become fully-committed Bregs until exposure to the combination of both IL-1β and IL-6 in the spleen.

Though haemodynamic changes underpin the pathophysiology of chronic liver disease, there are currently no robust non-invasive methods available for their assessment. I propose ‘caval subtraction’ phase contrast MRI (PCMRI) a novel method to measure total liver blood flow (TLBF) and hepatic arterial (HA) flow using PCMRI measurements of caval and portal venous (PV) flow. I validate this method at 9.4T and 3.0T to demonstrate: agreement between preclinical PCMRI and invasive transit-time ultrasound (TTUS) and fluorescent microsphere measurements of flow parameters; good consistency between clinical caval subtraction PCMRI and independent direct PCMRI measurements; encouraging correlations between PCMRI and invasive ICG clearance in patients; and good seven-day reproducibility of PCMRI derived haemodynamic parameters in normal volunteers. Using dynamic contrast enhanced (DCE) MRI on a 3.0T system, I demonstrate improved seven-day reproducibility using dual input single compartment pharmacokinetic modelling with a novel method for obtaining physiological vascular input function delays, correction of arterial input functions using PCMRI aortic flow and use of PCMRI estimations of TLBF to correct DCE MRI quantification. I also implement arterial spin labelling (ASL) at 9.4T and demonstrate a tendency for ASL to underestimate PCMRI hepatic parenchymal perfusion. Using bile-duct ligated (BDL) rats to study cirrhosis, I demonstrate that these have reduced TLBF and HA fraction at baseline, impaired HA regulation and buffer response, cirrhotic cardiomyopathy, and a failure to match hepatic circulatory demands with increased liver:body mass ratio. Acute-on-chronic liver failure (simulated using endotoxaemia) demonstrates reductions in TLBF, HA flow, absence of normal sepsis-induced hepatic hyperaemia and blunted cardiac systolic response. Studies in cirrhotic patients demonstrate increased TLBF and HA flow in higher risk portal hypertensive patients; that HA flow, HA fraction and cardiac output are important correlative parameters with hepatic venous pressure gradient and that caval subtraction PCMRI has potential in evaluating treatments for portal hypertension.