Medicine (Austin & Northern Health) - Theses

Permanent URI for this collection

Search Results

Now showing 1 - 10 of 156
  • Item
    Thumbnail Image
    The Immunological Tumour Microenvironment in Small Cell Lung Cancer and Pulmonary Neuroendocrine Tumours
    Rivalland, Gareth ( 2022-11)
    Background There is a limited understanding of the immunological tumour microenvironment (TME) in pulmonary neuroendocrine tumours (NETs), including well-differentiated (carcinoid) and small cell lung cancer (SCLC). A fuller understanding of the mechanisms by which these tumours escape immune surveillance is required to build on the success using checkpoint inhibitors in pulmonary NETs. This study presents two retrospective cohorts: one of pulmonary carcinoid tumours, and the second of SCLC examining of the IHC analysis of common checkpoints, their correlation with gene expression, and the impact of both on prognosis. In addition, a smaller group of SCLC patients underwent gene expression profiling (GEP), with clustering studies based on T-cell inflammation signatures and their impact on prognosis. Methods Tissue microarrays from 153 histologically-confirmed pulmonary carcinoids and 101 SCLC specimens were scored for IHC staining of immune checkpoints on tumour alone or TME (combined positive score, CPS). Immune checkpoint expression was analysed for impact on overall survival. A subgroup of 30 patients underwent GEP using the Nanostring nCounter Pan-cancer immune panel. Two IFN-G GEP signatures were examined for supervised clustering and impact on survival. Results In carcinoid tumours PD-L1 >5% was seen in 19/153 (12.4%), but only 3/153 (2.0%) expressed PD-L1 >50%. Expression of PD-L1 >5% was more frequent in atypical carcinoid tumours (6/34; 17.7%) than typical carcinoids (13/119; 10.9%; p 0.39 for significance). No significant tumour expression of PD-L2 and TIM3; LAG3 was expressed at low levels (2/153; 1.3%). Checkpoint expression in TME was low, with PD-L1 CPS >10% in 14/153 (9.1%). No significant prognostic impact was seen with checkpoint expression in well-differentiated tumours. PD-L1 was the most frequently expressed checkpoint in SCLC. PD-L1 >5% was seen in 17/101 (16.8%). PD-L2 in 2/101 (1.9%); TIM3 had 1/101 (1.0%); LAG3 in 1/101 (1.0%). Expression was more frequent in the tumour microenvironment with PD-L1 CPS >10% in 18/101 (17.8%); TIM3 CPS >10% in 50/101 (49.5%); LAG3 CPS >10% in 27/101 (26.3%) and PD-L2 CPS >10% in 2/101 (2.0%). 12/101 (11.9%) had co-expression of PD-L1, TIM3 and LAG3. PD-L1 >5% on tumour was associated with improved median OS, 27.7 months vs 12.9 months (p 0.04), as was LAG3 CPS expression, with median OS 18.0 months vs 11.9 months (p 0.02). The greatest improvement in survival was seen in cases with co-expression of checkpoints, median OS 33.0 months vs 12.5 months (p 0.04). Multivariable analysis of survival showed that these differences became non-significant when accounting for stage. Significant heterogeneity was observed in gene expression. Gene expression by mRNA level did not correlate with protein expression by IHC. Supervised clustering with the 18-gene and 28-gene IFN-G GEP signatures showed significant clustering but no significant association with survival. Conclusion Well-differentiated pulmonary NETs express PD-L1 at low rates in typical carcinoids and moderately low rates in atypical carcinoids. PD-L2, LAG3 and TIM3 were not expressed at significant levels on tumour or in the TME. The prognostic impact of checkpoint expression is unclear due to the excellent survival in carcinoid tumours with conventional therapy. In the SCLC cohort, tumour expression of PD-L1 and LAG3 in the TME were associated with improved survival on univariate analysis. Co-expression of immune checkpoints occurred in 16.7% SCLC and was associated with improved survival. These differences were accounted for by stage on multivariable analysis. The cohort showed clustering with T-cell inflamed/IFN-G GEP but there was no significant impact of T-cell inflammation on survival.
  • Item
    Thumbnail Image
    Detection of Somatic Mutations in Sporadic Focal Epilepsies
    Ye, Zimeng ( 2022)
    Epilepsy is one of the most common neurological conditions worldwide. Approximately 20–30% of epilepsy cases are attributable to acquired causes such as stroke or head injury. Many of the remaining cases are believed to have genetic contributions. Traditional genetic studies utilize leukocyte-derived DNA to search for germline variants, which are inherited or arise de novo in parental gametes. Over the past decade there has been growing evidence that pathogenic somatic variants arising post-zygotically may also make a major contribution to focal epilepsies and brain malformations. The current route to study the brain somatic variants is via sequencing brain tissues acquired from resective surgery or autopsy. The major limitation is access to brain tissues which are only available from a minority of patients. Novel approaches are required to detect somatic variants when brain tissues are not available. In this project, I explored and established three strategies to detect somatic variants in the absence of conventional surgical samples or autopsy. The first strategy was deep sequencing of peripheral blood samples to detect low-level mosaicism that is below the limit of detection of conventional genetic testing. In Tuberous Sclerosis Complex (TSC) 10-15% patients have no mutations identified in TSC1 or TSC2 on conventional genetic testing. I investigated 31 sporadic patients who were negative on conventional testing using 1,000-4,000x deep sequencing of blood-derived DNA. I solved 71% of cases, most had mosaic mutations at low variant allele fractions. In addition, I investigated 8 families with likely parental mosaicism and found low-level mosaicism in 5 families. The second strategy was the analysis of cell-free DNA (cfDNA) in cerebrospinal fluid (CSF) to detect somatic mutations confined to the brain. CSF liquid biopsy had been used to detect somatic mutations in malignant brain tumours, but its application in non-malignant neurological disorders was not explored. I established the first proof-of-concept that CSF liquid biopsy can be used to detect brain somatic mutations causing focal epilepsies. I first measured the CSF cfDNA level in patients with epilepsy to show the feasibility. Then I detected somatic mutations in CSF cfDNA in three patients with focal epilepsy and brain malformations to demonstrate efficacy. Finally, cell-of-origin analysis was performed to confirm that CSF cfDNA is largely brain-derived. The third strategy was screening of DNA from trace brain tissues obtained from stereo- electroencephalography (SEEG) electrodes. I employed this approach for the first time to a patient with non-lesional multi-focal epilepsy undergoing surgical evaluation. I amplified trace DNA from anatomically pooled electrodes and demonstrated a mosaic gradient for a novel KCNT1 stopgain variant predicted to lead to nonsense-mediated decay. I found the mosaic gradient correlated strongly with the SEEG findings in the patient as variant allele fraction was highest in the right posterior quadrant, her most epileptogenic seizure focus. Together, these new genetic diagnostic methods will shed light on the “hidden genetics” of epilepsy and provide insights for future precision medicine approaches in the clinic.
  • Item
    Thumbnail Image
    Improving prognostic and therapeutic precision in high-risk diffuse large B-cell lymphoma
    Wight, Joel Cameron ( 2022)
    Diffuse large B cell lymphoma (DLBCL) is the most common lymphoid malignancy and follows an aggressive clinical course. Standard chemotherapy including rituximab, cyclophosphamide, vincristine, doxorubicin and prednisolone (R-CHOP) can be expected to cure ~60% of patients, but those who experience relapse after initial disease control, or have lymphoma which is refractory to front-line combination chemotherapy have limited salvage options and often succumb to their disease. Identification of the clinical and biological factors underlying these poor responses to R-CHOP remains imprecise. One of the most devastating events is dissemination into the central nervous system (CNS), which is usually fatal despite dedicated intensive CNS-penetrating chemotherapy regimens. The objectives of this thesis were to examine the clinical, biological and immunological drivers behind DLBCL refractoriness and relapse in general, with a particular focus on a) the pathobiology and outcomes of CNS dissemination, b) the toxicity of CNS-directed therapy. Clinically, patients with secondary CNS lymphoma (SCNSL) have an aggressive clinical course and the optimal treatment is not defined. This thesis demonstrates that the primary cause of death in these patients is the CNS disease, and subsequently the importance of adding cytarabine to high-dose methotrexate to adequately treat the CNS component. The development of CNS disease, either synchronously at diagnosis or in relapse, occurs for largely unknown reasons. Clinical risk scores for determining CNS risk are imprecise and pathobiological data are lacking to develop rational drug combinations. This thesis compared genetic and functional drivers of secondary CNS DLBCL (SCNSL) samples with primary CNS lymphoma (PCNSL) and systemic-only DLBCL. SCNSL carries multiple aberrations in antigen presentation mechanisms (MHC, B2M, CD58, CIITA) in similar numbers to PCNSL and far more than systemic-only DLBCL suggesting that failure of antigen presentation is a possible mechanism of CNS dissemination with potential therapeutic implications. High-dose methotrexate itself carries substantial toxicity, the underlying risk for which has been undefined. This thesis evaluated the modifiable risk factors for toxicity with high-dose methotrexate and documented the most critical were dose and interacting/nephrotoxic drugs, and suggests how these observations can impact practice. Determining the likelihood of relapse or refractoriness to R-CHOP in systemic DLBCL has been an area of expansive research since the original international prognostic index was published in 1993. With rare exception, improved prognostication has not led to changes in therapy. Real-time assessment of disease response to R-CHOP using FDG-PET has been shown to improve prognostication, but this thesis demonstrates that it can also be used to identify early those who will have refractory disease. Specifically, those with a Deauville score of 5 following 2 cycles of R-CHOP, or a change in the maximum standard uptake value (SUVmax) of <66% following 4 cycles of R-CHOP are highly likely to have refractory disease and should be evaluated for alternative salvage or clinical trial. In order to relapse, DLBCL must escape immune surveillance, and a lack of diversity of the T-cell receptor repertoire at diagnosis has been associated with relapse. Work in this thesis assessed the T-cell diversity in paired diagnostic and relapsed tumour samples. It was observed that these tumour microenvironments become increasingly dysfunctional at relapse though further narrowing of the T-cell receptor repertoire and change in dominant clones, highlighting that it is not only clonal evolution that plays a role in relapse but also defunct immune responses. The high-risk DLBCL groups evaluated in this thesis demonstrate substantial clinical and biological heterogeneity. Identifying and managing the highest risk subgroups remains a challenge that requires comprehensive and multi-faceted approaches to overcome.
  • Item
    Thumbnail Image
    Understanding how gender affirming hormone therapy affects bone during puberty and in adulthood in male-tofemale mouse models
    Nie, Tian ( 2022)
    Sex steroids are essential for peak bone mass accrual during puberty as well as bone maintenance in adulthood, but the long-term effects of gender affirming hormone therapy (GAHT) on bone and fracture in transitioning adolescent transgender girls and adult women are poorly understood. We hypothesized that in both young pubertally suppressed male mice and adult intact male mice, that exogenous estradiol treatment will not be sufficient to restore cortical and trabecular bone volume or bone strength to levels observed in male sham controls. To address my hypothesis, I developed two mouse models. The first mimics transitioning in adolescent transgender girls by first pubertal suppression via orchidectomy (ORX) at a pre-pubertal age, with estradiol treatment at a later age. The second models transitioning in adult women, where intact adult male mice were given estradiol. Our aim was to determine the effcts of GAHT administered during puberty and adulthood on bone microstructure and strength in mouse models of male-to female transition. As expected ORX in the pre-pubertal male-to-female mice markedly decreased trabecular and cortical bone 3 weeks post-surgery. The dose of estradiol treatment administered to pubertal ORX male-to-female mice (0.9 mg) led to serum estradiol concentrations analogous to the peak estradiol concentrations that occur during the proestrus stage of the estrus cycle in female mice (Handelsman et al., 2020), whilst a 1.3 mg estradiol dose in adult male-to-female mice resulted in a 4.7-fold higher mean circulating concentration of estradiol than those observed during the proestrus stage. In both mouse models, estradiol treatment in young ORX males or adult intact males was associated with a marked increase in trabecular and cortical bone compared to age matched controls, consistent with the actions of estradiol to inhibit bone resorption and stimulate the endocortical deposition of bone. The increase in both the trabecular and cortical bone fractions following estradiol treatment in the ORX males and intact males led to an increase in the maximum force the bones could withstand prior to fracture compared to male controls, indicative of increased bone strength. These data in both the adolescent and adult mouse models of male-to-female transition are contrary to our hypotheses, where estradiol in the high physiological or supraphysiological range were able to increase both trabecular and cortical bone and thus circumventing the requirement of local synthesis of estradiol within bone via the aromastisation of testosterone. Taken together these data show in young mice, the deficit in peak bone mass accrual that occurs with pubertal suppression in male mice can be rescued, thereby preserving bone strength, if sufficient serum estradiol concentrations are achieved in late puberty. In adult mice, supraphysiological estradiol treatment increased bone mass and subsequently bone strength. Clinically, it may be possible to preserve bone health and prevent fractures in trans girls and trans women treated with gender affirming hormone therapy, if sufficiently high levels of estradiol can be achieved without adverse side effects.
  • Item
    Thumbnail Image
    Could estimation of kidney function in people with diabetes be improved?
    Zafari, Tina ( 2022)
    There are about 450 million people with diabetes in the world, 30-40% of whom develop diabetic kidney disease (DKD) and end up with end stage kidney disease at some stage in their life. Progression of DKD onto end-stage kidney disease could be reduced by timely initiation of multifocal, target-driven interventions to control its risk factors. Therefore, accurate diagnosis of diabetic kidney disease is pivotal for its management. The main diagnostic tool to identify those with DKD or at increased risk of developing DKD in clinical practice is estimating glomerular filtration rate (eGFR) by multivariate equations based on serum creatinine, age, and sex. The aim of this dissertation was first to systematically review performance of the most commonly used eGFR equation, the chronic kidney disease epidemiology collaboration (CKD-EPI), in assessing kidney function. I then aimed to evaluate diagnostic performance of modification of diet in renal disease (MDRD) and CKD-EPI as well as the newer eGFR equations (i.e., revised Lund-Malmo, full age spectrum (FAS), Cambridge GFR version2 (CamGFRv2), and European Kidney Function Consortium (EKFC)) in a cohort of people with diabetes attending Austin Hospital, Melbourne, Australia. Next, I aimed to assess alternative ways of improving diagnostic performance of these eGFR equations. My suggested alternatives were i) optimising eGFR cut points used in important clinical decision makings; ii) using artificial intelligence neural network models instead of equations to estimate GFR. I assessed diagnostic performance of eGFR equations by both continuous (i.e., bias, precision, accuracy, Bland-Altman plot, and reduced major axis regression) and categorical (i.e., sensitivity, specificity, area under the receivers operating curve, and Kappa inter-rater coefficient) outcome measures. Results of my thesis indicated that there was no consensus on the statistical methods to report diagnostic performance of eGFR equations on a continuous scale. Regardless, it seemed that the current widely used eGFR equation, is not precise and accurate in people with diabetes. Newer equations particularly revised Lund-Malmo seemed to improve diagnostic performance of eGFR, however, their performance was still suboptimal. Optimised GFR cut points as well as our artificial neural network derived model (RenoTrueTM) could potentially improve diagnosis of DKD. Nonetheless, our suggested optimised eGFR cut points and RenoTrueTM model need to be further validated in other study populations.
  • Item
    Thumbnail Image
    The Mechanisms Underlying the Beneficial Effect of AMPK Activators in Renal Injury and Fibrosis
    Harley, Geoffrey Alexander ( 2022)
    Altered energy metabolism within tubular cells plays a key role in determining the outcome following acute kidney injury (AKI) and is thought to mediate the severity of the injury as well as the degree of resultant fibrosis. Despite the high prevalence of acute kidney injury and chronic kidney disease with permanent fibrosis of kidney tissue, there is a paucity of therapies available for the prevention of AKI. AMP-activated protein kinase (AMPK) has a key role in regulating both fatty acid oxidation and glycolysis, the primary methods of energy generation within the kidney, through its actions on the acetyl-CoA-carboxylase (ACC) and phosphofructokinase (PFK) pathways. When activated by cellular energy depletion, AMPK phosphorylates its substrates leading to inhibition of energy-consuming and promotion of energy-producing pathways. AMPK activators, such as metformin, have previously shown to be protective in models of kidney fibrosis, however clinical use of metformin in human AKI is limited by lactic acidosis. This thesis seeks to further examine the mechanisms of AMPK activators through which they have beneficial effects against kidney fibrosis. The importance of metformin’s effects on glycolysis was examined using an inactivating knock-in mutation of a key regulatory point in glycolysis. 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase (PFKFB) has a critical role in regulating the activity of phosphofructokinase-1 (PFK-1), the key rate-limiting step in glycolysis. Mice with inactivating mutations of the phosphorylation binding sites of PFKFB2 (PFKFB2 KI mice), the predominant isoform of PFKFB within the kidney, were examined in a unilateral ureteric obstruction (UUO) model of fibrosis with and without added metformin. Mice undergoing UUO had decreased amounts of fibrosis, as measured by Western Blot and PCR markers of alpha-smooth muscle actin and fibronectin when treated with metformin. This effect was preserved in the PFKFB2 KI likely due to ongoing effects on fatty acid oxidation, suggesting that PFKFB2 phosphorylation was not essential to the anti-fibrotic effects of metformin. A double knock-in mutation was then examined with the combination of ACC1 and PFKFB2 inactivating mutations of phosphorylation sites. Mice were then subjected to UUO with or without metformin. ACC exists as two isoforms within the kidney; ACC1 in the cytosol which is thought to be far more predominant in the kidney and important, and ACC2 within the mitochondria. Since ACC1 and PFKFB2 are thought to be the key enzymes regulating fatty acid oxidation and glycolysis respectively through which AMPK activators have their effect, it was hypothesized that metformin would have minimal protective effects against fibrosis. Contrary to this hypothesis, metformin continued to demonstrate protective effects in the dual ACC1/PFKFB2 KI mice in this UUO model. To explore this further, isolated ACC2 KI mice were generated and subjected to a UUO model. ACC2 KI UUO mice had evidence of worse fibrosis as measured by alpha-smooth muscle actin compared to their WT counterparts. Metformin did not demonstrate a protective effect against fibrosis in ACC2 KI UUO mice, suggesting ACC2 within the mitochondria may play a more important role in mediating the effects of metformin than previously thought, a novel finding. These experiments were then further extended to examine an injury-based model of AKI before fibrosis is fully established. This was to explore if the degree of resultant fibrosis in these models is due to worse injury at the time of AKI or purely due to worse resulting fibrosis following this injury. Cisplatin nephrotoxicity, a well-established model of acute kidney injury with relevance in clinical practice, was selected for this. Mice with inactivating mutations of both ACC1 and ACC2 were subjected to cisplatin intraperitoneal injection and displayed markers of worse injury compared to their WT counterparts, indicating the kidney’s ability to regulate fatty acid oxidation is critical to reducing the amount of injury following cisplatin-nephrotoxin exposure. Finally, 0304, a novel AMPK activator demonstrated to have pleiotropic beneficial effects in other tissues, was examined in regards to its role in the kidney. Tubular epithelial cells were exposed to O304 and demonstrated increased ACC-Ser79 phosphorylation consistent with O304 being an effective AMPK activator in kidney tissues as well. An in vivo model of renal fibrosis with UUO in WT and ACC1/2 KI kidneys did not demonstrate a significant effect of O304, likely due to methodological differences in O304 administration. Overall, the work in this thesis further elucidates the mechanisms of AMPK activators in the kidney, linking its effects on fatty acid oxidation and glycolysis to injury and fibrosis models of kidney disease. This may prove clinically relevant in the future development of treatments for AKI.
  • Item
    Thumbnail Image
    Biological Actions of Estradiol in Men
    Russell, Nicholas David Fletcher ( 2022)
    ABSTRACT Context: Accumulating evidence suggests that many of the biological actions of testosterone (T) in men are dependent on endogenous aromatisation of T to estradiol (E2). However, few studies have been designed to measure effects of E2 in men directly, in the absence of T. The most prevalent contemporary cause of severe hypogonadism in older men is the use of gonadotrophin-releasing hormone analogs (GnRHa) as androgen deprivation therapy (ADT) for prostate cancer. ADT represents a unique situation in which the usual clinical and ethical imperative to treat severe male hypogonadism with T is absent. Men receiving ADT experience accelerated osteoporosis leading to fractures, accumulation of fat mass, loss of muscle mass, sexual dysfunction, hot flushes, and, reportedly, adverse neuropsychological effects. This thesis reports on a project designed to test the hypothesis that low dose E2 treatment of men receiving GnRHa for prostate cancer would mitigate some of the adverse effects of the induced hypogonadism, by restoring E2 action (E2 ‘add-back’). There were two overarching aims: first, to directly examine the isolated effects of E2 in men in the absence of T in order to provide new information on the biological role of E2 in men; and second, to test the efficacy of transdermal E2, a readily available and cheap therapeutic, for the mitigation of ADT-induced side effects. Methods: Two randomised controlled trials (RCTs) were conducted in men undergoing ADT for prostate cancer. These trials tested the hypotheses that transdermal E2 would: 1. Reduce the ADT-associated unbalanced and accelerated bone remodelling, leading to better maintenance of volumetric bone mineral density (vBMD). 2. Reduce the ADT-associated increase in fat mass. 3. Reduce ADT-associated changes in cognition. 4. Reduce hot flushes and thereby improve quality of life. Trial 1 aimed to assess the effects of two low doses of transdermal E2 on serum E2 concentrations and to assess the effects of E2 add-back on bone remodelling markers and hot flushes. This trial was a 28-day RCT, randomising 37 participants to transdermal E2 0.9mg/d, 1.8mg/d, or matching placebos. Fasting morning pre-dose serum sex steroids were measured by liquid chromatography mass spectrometry. Hot flushes were measured by prospective 7-day diary. Carboxyl-terminal type 1 collagen telopeptide (CTX) and pro collagen type 1 amino-terminal propeptide (P1NP) were measured by electrochemiluminescence. Trial 2 aimed to assess the efficacy of transdermal E2 0.9mg/d on total vBMD at the tibia and total fat mass (co-primary endpoints), as well as on cognition, hot flushes, and quality of life. This trial was a 6-month RCT, randomising 78 participants to transdermal E2 0.9mg/d, or matching placebo. Total vBMD was measured by high resolution peripheral quantitative CT at baseline and study end. Body composition was measured by dual x-ray absorptiometry at baseline, month 3 and month 6. At each visit, sex steroids, bone remodelling markers, and hot flushes were measured as per trial 1. Quality of life was measured by the Functional Assessment of Cancer Therapy – Prostate and the Aging Males’ Symptom Scale instruments. Cognition was assessed by a validated, tablet computer-based cognitive battery. Results: In trial 1, transdermal E2 0.9-1.8 mg daily increased median serum E2 concentrations into the reference range reported for healthy men, but with substantial variability. Day 28 serum E2 concentrations ranged from 106 to 870 pmol/L in the 0.9mg dose group (median 208 pmol/L; interquartile range: 157-332) and 96-1814 pmol/L in the 1.8mg dose group (median 200 pmol/L; interquartile range 144-660). Over 28 days, transdermal E2 reduced serum CTX, increased serum P1NP, and reduced hot flushes. In trial 2, serum E2 increased in the E2 group over 6 months compared to the placebo group, mean adjusted difference (MAD) 207 pmol/L (95% CI 123-292), p<0.001. E2 add-back did not significantly change total vBMD at the distal tibia, MAD 2.0 mg hydroxyapatite per cm3 (95% CI -0.8 – 4.8), p=0.17. E2 increased cortical vBMD at the distal radius, and estimated failure load at the tibia and radius, and reduced CTX and P1NP. E2 add-back did not significantly change total fat mass, although the confidence interval suggested a likely increase in the E2 group, MAD 1007g (95% CI 124-1891), p=0.09. E2 increased android fat, MAD 164 g (95% CI 41-286), p=0.04. E2 did not change performance over time in any pre-defined cognitive domain. E2 reduced daily hot flush frequency, mean adjusted difference (MAD) -1.6 hot flushes per day (95% CI -2.7 to -0.5), p=0.04. but had no significant effect on quality of life. Conclusion: These two trials offer new insights into the biological actions of E2 in men by providing direct observations of E2 effects, in the absence of T. Over 6 months, E2 add-back reduced the bone remodelling rate, and had beneficial effects on bone to increase aBMD at lumbar spine and ultra-distal radius, cortical vBMD at radius, and estimated failure load, a measure of bone strength, at both radius and tibia. Contrary to the hypothesis, these data suggest that E2, in the absence of T, increases fat mass, suggesting that the effect of E2 on fat might be more complex than reported in studies inferring E2 effects to reduce fat from paradigms in which T is present. E2 had no observed effect on cognition, suggesting that any short-term effects of E2 on cognition in men, are likely to be subtle. Transdermal E2 add-back durably reduced hot flushes, thus providing high quality evidence for this intervention to be used clinically.
  • Item
    Thumbnail Image
    Investigating the role of UBL5 and the mitochondrial unfolded protein response in pancreatic beta cell function
    Haralambous, Christian Savvas ( 2022)
    Type 2 diabetes is characterised by hyperglycaemia caused by insulin resistance and an impairment in beta cell function. The hyperglycaemia associated with type 2 diabetes causes oxidative stress and mitochondrial dysfunction in pancreatic beta cells, which can lead to apoptosis. Oxidative stress is the process where reactive oxygen species bind to biological molecules such as proteins, lipids and DNA causing them to not to function correctly. Cells have a variety of stress responses to prevent any further damage to the cell from oxidative stress, in the mitochondria one of the most vital is called the mitochondrial unfolded protein response (UPRmt). The UPRmt serves to upregulate mitochondrial specific chaperone proteins and proteases to help remove, replace and restore function to proteins damaged by oxidative stress. Ubiquitin-like protein 5 (UBL5) is a protein that has been shown in the literature to be vital to the activation of the UPRmt in the invertebrate C. elegans. To investigate the UPRmt in pancreatic beta cells a tamoxifen inducible beta cell specific UBL5 KO mouse was generated and the effect of heterozygous and homozygous UBL5 knockout in pancreatic beta cells. Complete deletion of UBL5 in pancreatic beta cells resulted in decreased glucose tolerance and insulin secretion caused by a reduction in beta cell mass and mitochondrial respiration from increased oxidative stress and apoptosis. This was also accompanied by a global decrease in UPRmt transcription. Conversely, partial deletion of UBL5 resulted in an opposing phenotype. The heterozygous UBL5 pancreatic beta cell knockout mice showed an increase in insulin secretion, beta cell mass, mitochondrial respiration and UPRmt gene transcription. Following the characterisation of the pancreatic beta cell UBL5 knockout mice, we aimed to investigate the place of UBL5 in the UPRmt pathway using RNA interference (RNAi) and pharmacological methods. However, RNAi knockdown of UBL5 and SPG7 failed to illicit changes in the UPRmt in MIN6 cells. Metformin was able to dose dependently upregulate transcription of UPRmt genes in MIN6 cells and was chosen for further investigation in animal models. Prior work for our laboratory demonstrated that metformin could increase insulin secretion and mitochondrial respiration in the New Zealand obese (NZO) mouse, a model of spontaneously develops obesity and diabetes. The NZO mice were shown to have decreased mitochondrial respiration, insulin secretion and UPRmt gene expression in their pancreatic islets compared to C57BL/6 control mice. Metformin treatment significantly increased their insulin secretion, mitochondrial respiration and transcription of UPRmt genes. Many of the proteins associated with the UPRmt are known to protect cells from damage caused by oxidative stress. The heterozygous UBL5 knockout mice had a global upregulation of UPRmt genes. Therefore, we next wished to investigate if this could offer any protective effects against a 10-week high fat diet. There was no difference in glucose tolerance, insulin secretion or beta cell mass between control and heterozygous UBL5 knockout mice on the high fat diet. However, heterozygous UBL5 knockout mice on the chow diet for 10 weeks displayed higher insulin secretion and glucose tolerance than the mice in earlier experiments that had only had UBL5 deletion for 2 weeks. In conclusion, this thesis demonstrated that UBL5 is vital for the correct function of pancreatic beta cells in mice. It also showed the importance of UBL5 in the regulation of the UPRmt in mammalian cells. Furthermore, it proved the significance of the UPRmt in mammalian cellular function and how changes in the regulation of the UPRmt can have either a protective or detrimental effect on cellular function.
  • Item
    No Preview Available
    Neuromodulation for Lennox-Gastaut syndrome
    Dalic, Linda Jane ( 2022)
    Lennox-Gastaut syndrome (LGS) is a treatment-resistant, childhood-onset generalised epilepsy, defined by multiple seizure types, specific electroencephalography (EEG) abnormalities and cognitive impairment. Affecting ~5% of people with epilepsy, long-term prognosis is poor and new treatment strategies are urgently needed. Deep brain stimulation (DBS) is an emerging treatment for drug-resistant epilepsies where resective neurosurgery is not indicated. A small number of DBS studies in LGS provide preliminary evidence that stimulation to the thalamic centromedian nucleus (CM) is likely to be beneficial in reducing seizures but randomised-controlled trials are required to further evaluate benefits and side effects. This thesis explored the role of neuromodulation in LGS, by conducting a prospective, double-blind randomised controlled trial of duty-cycle CM-DBS in young adults with LGS (the ‘ESTEL’ study: Electrical Stimulation of the Thalamus in Epilepsy of the Lennox-Gastaut phenotype). Following three-month pre- and post-implantation periods, half of the ESTEL participants received three months of stimulation (blinded phase), then all received three months of stimulation (unblinded phase). The primary outcome of ESTEL was the proportion of participants with a ≥50% reduction in diary recorded seizures in stimulated versus control participants, measured at the end of the blinded phase. A secondary outcome was the proportion of participants with ≥50% reduction in electrographic seizures on 24-hour ambulatory EEG at the end of the blinded phase. Study 1 reports the efficacy and safety of CM-DBS, from the ESTEL study. Between November 2017 and December 2019, 20 patients with LGS (19-37 years; 13 women) underwent bilateral CM-DBS at Austin Health, Australia, with 19 randomised to treatment (treatment, n=10 and control, n=9). The primary outcome was not met with five (50%) patients in the treatment group achieving a ≥50% reduction in diary recorded seizures, compared with two patients (22%) in the control group (adjusted OR=3.1, 95% CI = 0.44 – 21.45, p=0.25). However, for electrographic seizures, 89% of the stimulation group had ≥50% reduction at the end of the blinded phase, compared with none of the controls (OR=23.25, 95% CI = 1.0 – 538.4, p=0.05). Across all participants, median seizure reduction (baseline vs study exit) was 46.7% (IQR=28-67%) for diary recorded seizures and 53.8% (IQR = 27-37%) for electrographic seizures), providing supporting evidence of treatment effect. Serious adverse events occurred in seven (35%) participants and were unrelated to stimulation. One participant developed infection, requiring DBS device removal prior to randomisation. Transient post-operative drowsiness was common after surgery (12/20 participants), with three participants having more profound drowsiness, corresponding to imaging showing marked cerebral oedema (i.e., mega-oedema) along the DBS electrodes course. Stimulation related adverse events were low; three participants reported ipsilateral hand/face/lip paraesthesia in the blinded phase which resolved over hours to days. Study 2 detailed cognitive and behavioural outcomes of participants, following CM-DBS of at least three months duration as part of the ESTEL trial. Six standardised cognitive and behavioural measurements were administered, finding improved epilepsy severity (GASE) and disability (GAD) scores after three months of CM-DBS stimulation. No decline in cognition (NIHTB-CB), adaptive skills (ABAS-3), quality of life (QOLIE-31) or depression (PHQ-9) scores were observed, providing supporting evidence that CM-DBS is safe in patients with LGS. Study 3 assessed whether the epileptic activity of LGS is detectable in CM in the first 12 ESTEL participants undergoing CM-DBS surgery. EEG recordings and phase transfer entropy (PTE), a recently introduced phase-based measure of effective connectivity, assessed the roles of the cortex and CM during the epileptic activity of LGS. Recordings showed that epileptic activity of LGS was detectable in CM and that generalised paroxysmal fast activity (GPFA), a key EEG feature of patients with LGS, occurred a median of 98 msec earlier at the scalp (reflecting superficial cortical activity) than the thalamus. PTE analysis supported an initiating role for the cortex, with information transfer from scalp to thalamus prior to GPFA. Study 4 explored whether the burden of GPFA can be used as biomarker of seizure susceptibility in patients with LGS. The burden of GPFA, quantified over a two hour window of sleep EEG was associated with changes in the daily rate of diary recorded seizures (p<0.001, η²p = 0.30-0.48). Baseline median seizures per day on seizure diaries were 2.6 (IQR:1.4 - 5), compared with 284 (IQR:120.5 - 360) electrographic seizures per day, confirming that diaries capture only a fraction of seizure burden. Within an individual, the ratio between GPFA and diary recorded seizures appeared stable, hence changes in GPFA may allow estimation of diary recorded treatment response in participants undergoing DBS. This thesis provides evidence for CM-DBS as a new treatment for LGS, with the ESTEL study demonstrating safety, efficacy and improved epilepsy disability and severity scores. It also confirms that the CM is a part of the epileptic network of LGS and that the cortex leads in its epileptic discharges. Finally, this thesis demonstrates a relationship between the burden of GPFA and clinical treatment response amongst ESTEL participants, suggesting that GPFA may be a suitable biomarker in LGS to monitor for treatment effect.
  • Item
    No Preview Available
    The cardiometabolic consequences of cancer therapies across the breast cancer spectrum
    Cheung, Yee-Ming Melody ( 2022)
    With earlier detection and advances in treatment, there has been a notable improvement in the survival of women with breast cancer over the past decade. However, this has ultimately led to an ageing and relatively comorbid breast cancer survivorship population. As cardiovascular disease (CVD) remains the leading competing cause for mortality in older breast cancer survivors, it is concerning that many of the approved anti-neoplastic agents used in the treatment for breast cancer may be associated with increases in CVD and cardiometabolic risk factors. This thesis aims to expand our knowledge and understanding of the inadvertent impact various anti-neoplastic agents may have on the cardiometabolic health of individuals with breast cancer. It explores the mechanisms in which these agents potentially alter cardiometabolic risk markers and also investigates and describes methods in which to best monitor and manage such risks. The work presented in this thesis demonstrates that in individuals initiating aromatase inhibitors (AI), treatment for the first 12 months does not appear to increase visceral adipose tissue or other cardiometabolic risk markers in postmenopausal with early oestrogen receptor positive breast cancer. Diabetes and poor glycaemic control however, are associated with increased overall mortality in long-term metastatic breast cancer survivors. Hyperglycaemia also appears to be a common cause for premature cancer therapy discontinuation and dose reductions. Finally, this thesis highlights how our progressive understanding of pertinent onco-endocrinological themes have translated into informing our day-to-day clinical practice. As our understanding of the interaction between cancer therapies and markers of cardiovascular and cardiometabolic risk continue to grow, so too will our strategies of monitoring and caring for this vulnerable and complex patient population.