Leon Liebenberg

The mechanism by which moderate alcohol consumption influences coronary heart disease

BackgroundModerate alcohol consumption is associated with a lower risk for coronary heart disease (CHD). A suitably integrated view of the CHD pathogenesis pathway will help to elucidate how moderate alcohol consumption could reduce CHD risk.MethodsA comprehensive literature review was conducted focusing on the pathogenesis of CHD. Biomarker data were further systematically analysed from 294 cohort studies, comprising 1 161 560 subjects. From the above a suitably integrated CHD pathogenetic system for the purpose of this study was developed.ResultsThe resulting integrated system now provides insight into the integrated higher-order interactions underlying CHD and moderate alcohol consumption. A novel ‘connection graph’ further simplifies these interactions by illustrating the relationship between moderate alcohol consumption and the relative risks (RR) attributed to various measureable CHD serological biomarkers. Thus, the possible reasons for the reduced RR for CHD with moderate alcohol consumption become clear at a glance.ConclusionsAn integrated high-level model of CHD, its pathogenesis, biomarkers, and moderate alcohol consumption provides a summary of the evidence that a causal relationship between CHD risk and moderate alcohol consumption may exist. It also shows the importance of each CHD pathway that moderate alcohol consumption influences.

A practical quantification of blood glucose production due to high-level chronic stress.

Blood glucose (BG) is the primary metabolic fuel for, among others, cancer cell progression, cardiovascular disease and inflammation. Stress is an important contributor to the amount of BG produced especially by the liver. In this paper, we attempt to quantify the BG production due to chronic (in the order of weeks) high-level psychological stress in a manner that a lay person will understand. Three independent approaches were used. The first approach was based on a literature survey of stress hormone data from healthy individuals and its subsequent mathematical manipulation. The next approach was a deductive process where BG levels could be deduced from published stress data of large cardiovascular clinical trials. The third approach used empirical BG data and a BG simulation model. The three different methods produced an average BG increase of 2.2-fold above basal for high levels of stress over a period of more than a day. The standard deviation normalized to the average value was 4.5%.

How do high glycemic load diets influence coronary heart disease

BackgroundDiet has a significant relationship with the risk of coronary heart disease (CHD). Traditionally the effect of diet on CHD was measured with the biomarker for low-density lipoprotein (LDL) cholesterol. However, LDL is not the only or even the most important biomarker for CHD risk. A suitably integrated view of the mechanism by which diet influences the detailed CHD pathogenetic pathways is therefore needed in order to better understand CHD risk factors and help with better holistic CHD prevention and treatment decisions.MethodsA systematic review of the existing literature was conducted. From this an integrated CHD pathogenetic pathway system was constructed. CHD biomarkers, which are found on these pathways, are the only measurable data to link diet with these CHD pathways. They were thus used to simplify the link between diet and the CHD mechanism. Data were systematically analysed from 294 cohort studies of CHD biomarkers constituting 1 187 350 patients.Results and discussionThe resulting integrated analysis provides insight into the higher-order interactions underlying CHD and high-glycemic load (HGL) diets. A novel “connection graph” illustrates the measurable relationship between HGL diets and the relative risks attributed to the important CHD serological biomarkers.The “connection graph” vividly shows that HGL diets not only influence the lipid and metabolic biomarkers, but also the inflammation, coagulation and vascular function biomarkers in an important way.ConclusionA focus primarily on the low density lipoprotein cholesterol biomarker for CHD risk has led to the traditional guidelines of CHD dietary recommendations. This has however inadvertently led to HGL diets. The influence of HGL diets on the other CHD biomarkers is not always fully appreciated. Thus, new diets or other interventions which address the full integrated CHD impact, as shown in this paper, are required.

Complexity of metabolic cancer control: Can we exploit the superior metabolic position of glucose?

We respond to a paper by Ko and colleagues (“Glutamine fuels a vicious cycle of autophagy...”). They hypothesize that tumor stroma are the central fuel generators for cancer growth, the so-called “reverse Warburg effect.” Irrespective of whether this mechanism or the extensively quoted “Warburg effect” accounts for energy production and biosynthesis, we propose (based on recent findings) that all cancer cell metabolites are either glucose-dependent or glucose-derived. We ask whether metabolic cancer control may be simpler than expected. Ko and colleagues show that neighboring cancer-associated fibroblasts are stimulated to convert to glycolytic metabolism through cancer cell-initiated oxidative stress. The fibroblasts thus become glucose dependent. This is the reason why the FDG-PET image of fibroblasts may be even larger than that of the primary tumors. Complexity of metabolic cancer control Can we exploit the superior metabolic position of glucose?

Development of a motion platform for an educational flight simulator

Flight simulators are regularly used in the undergraduate and postgraduate training of mechanical and aeronautical engineers. Due to advances in computing technology, several flight simulation-related tasks can now be accomplished in real-time using low-cost PC platforms and inexpensive commercial software. The difficulty in realising an educational flight simulator system with motion platform therefore lies with the design and construction of an effective motion platform. Costs become exorbitant when simulation platforms of more than two degrees of freedom (i.e. pitch and roll) are attempted. This paper describes the development of a drive system for a motion platform with two degrees of freedom (pitch and roll) for use in undergraduate engineering training. Use was made of off-the-shelf PC equipment and flight simulation software and hardware, together with commercial actuators and drive systems. The motion platform was manufactured from square tubing and consisted of three frames: the stationary main frame and, rotating inside this, the roll frame and pitch frame. These rotated relative to each other and were actuated by two similar-sized DC motors and gearbox/chain transmissions. The system effectively simulated the pitch and roll motions of commercial airliners, using a low-cost, easily maintainable motion platform. The educational value of the simulator was twofold: first, it was to be displayed in the science exploratorium (SciEnza) of the University of Pretoria; and second, it provided a platform on which mechanical (as well as electrical, electronic and computer) engineering students could conduct practical work in courses such as dynamics and control, and on which final-year and postgraduate students could conduct research.

Is knowledge of brain metabolism the key to treating highly glycolytic cancers and metastases

In their recent article, Weller et al1 delineate the minor improvement in survival rates among patients with glioblastoma treated with various chemotherapeutic strategies. Malignant gliomas resort in the group of highly glycolytic cancers and metastases (HGCM) responsible for 90% of cancer-associated deaths.2 Is it possible to improve the outcome in these patients by considering the brains strong metabolic regulation? The brain is the healthy bodys most glycolytic organ, with a typical 18fluorodeoxyglucose (FDG) PET-derived standard uptake value of 8.22.3 Of interest, this is close to the mean standard uptake value of 9.25 (variance, 5–22) for the HGCM that we investigated previously.3 This means that the level of cerebral blood glucose (BG) metabolism is similar to that of the mean HGCM. Therefore, the brain with its strong energy regulation should provide the ideal microenvironment for highly glycolytic metastasis. This is indeed what is found in practice (US data), with 40% of all metastases also occupying the brain.4 With the high BG microenvironment, brain metastases are also lethal within months.1,4 Because brain BG needs are similar to those of HGCM, any aggressive antiglycolytic treatment must account for the brains BG needs. This is indeed what is starting to emerge in practice. Before entering phase II clinical trials, the glycolytic inhibitor 2DG showed brain toxicity.5 Should we not devise methods to harm HGCM in a controlled manner while still satisfying the brains minimum metabolic needs?3 In vivo data 6,7 suggest that such an approach (via extracorporeal glucose deprivation)3 could potentially work. Furthermore, the latest BG control technology may make such an approach practical.8 When devising clinical trials for antiglycolytic treatments, it should be borne in mind that rodent models will not manifest the key invasive properties of malignant gliomas.9 This is attributable to the vastly different regulation strengths that the rodent and the human brains have on their respective BG environments.10 There is an urgent need to more fully explore antiglycolytic treatments for patients with glioblastoma. We hypothesize that the human brain metabolism may be the limiting factor for any aggressive antiglycolytic HGCM treatment. Exploiting such a fact could result in more successful therapies targeting this tumor-protective niche. We believe that this merits further investigation.

Isolated Limb Perfusion: Is it Possible to Increase Cancer Treatment Efficacy by Simultaneous Glucose Deprivation?

In their recent paper, Deroose et al. [1] elucidate the excellentoverall response rate of 71% for soft-tissue sarcoma patients treatedwith TNF-a and melphalan-based hyperthermic isolated limb perfu-sion (HILP). Is it possible to improve on these rates by imposingnovel microenvironmental challenges to cells and tissues duringHILP?Isolation from systemic circulation with HILP permits the admin-istration of high doses of cytotoxic chemotherapy. However, the useof extremely high doses of TNF-a and melphalan presents a life-threatening situation in the case of systemic leakage [2]. It maytherefore also be asked whether it is possible to better sensitize iso-lated tumours to chemotherapeutics, thus allowing for lower, safer,concentrations of cytotoxic agents during HILP.There is a growing interest in the Warburg effect focusing onglucose metabolism in highly glycolytic cancers and metastases [3].Furthermore, it was recently shown that glucose appears at the top ofthe cancer cell (and cancer-associated fibroblast) metabolic hierarchy[4]. Therefore, if cancer cells and associated fibroblasts can bedeprived of glucose, the cancer cells will not only be less able tometabolise glutamine but will also be deprived of de novo producedlactate, fatty acids and ketone bodies [5].Recent research indeed shows that glucose deprivation inducescancer cell apoptosis whilst not adversely affecting healthy cells [6–8]. Glucose deprivation treatment additionally increases tumour cellsensitivity to chemotherapeutics [6–8] as well as radiotherapy [9].This might suggest that lower (and safer) levels of chemo- andradio-therapeutics should be achievable when treating a cancer pa-tient with HILP.Such an approach could also be practical as resting skeletal mus-cle rely mainly (approximately 60%) on plasma fatty acids for ener-gy source supply, with the balance being provided by blood glucose[5]. Furthermore, normal cells are more metabolically adaptable thantheir cancerous counterparts with their glycolytic phenotypes [5].The surrounding normal cells of the isolated limb should thereforebe able to utilise other oxidisable substrates, such as fatty acids,when deprived of glucose [5].It is thus possible to safely remove glucose (and glutamine) froma patient’s isolated limb by means of haemodialysis [10]. This isachieved by arranging lower concentrations of glucose and glutaminein the dialysate, and by selection of an appropriate dialyser mem-brane (for diffusion of glucose and glutamine).We believe that this approach merits further investigation.

Quick Estimates for Analysis and Prediction of the Flight Mechanics of Unmanned Aerial Vehicles

Unmanned aerial vehicles (UAVs) are commonly employed in undergraduate engineering curricula. Limited literature is, however, available for the lay design engineer or engineering student regarding the modelling, simulation and analysis of the flight dynamics of small UAV systems, especially pertaining to flight dynamics modelling. There is great demand for unskilled UAV designers to predict the stability of new designs, quickly, cheaply, and with relative ease, preferably during the conceptual design stage. This paper summarizes some salient techniques for performing quick characterization of the longitudinal dynamics of a small, electrically propelled UAV, by using freely available software such as Datcom+, AVL, XFLR5 and MotoCalc. The simulation outputs compare favourably with experimental results from a wind tunnel. The software was also used to provide accurate estimates of coefficients required for performing an analysis of the UAVs longitudinal dynamics. The proffered analytical techniques should greatly benefit lay design engineers and engineering students venturing into the realm of UAV research.

Improved rodent models of human brain metastases

We respond to a recent paper by Daphu et al. [1]. The authors inter alia report that most metastases studies have focussed on a characterization of the tumor cell (‘‘seed’’) and neglect potential contribution from the microenvironment of the target organ (‘‘soil’’). We agree and thus focus on the energy microenvironment in this letter. Brain metastasis has a very high glucose metabolism with FDG standard uptake values (SUVs) of around 7.8 [2]. We therefore suspect the following: similarity in the blood glucose (BG) microenvironment (and its control) between the rodent and human brain should be critical when comparing highly glycolytic brain metastasis in the two species. The normal human brain (cerebellum) also has a very high glucose metabolism with FDG-SUV values of typically around 8.2 [3]. Interestingly this is close to the average SUV of 9.25 for highly glycolytic cancers and metastases (HGCM) [4]. It appears that brain BG metabolism is similar to that of the ‘‘average’’ HGCM. We thus suspect that the human brain should provide the ideal ‘‘soil’’ for highly glycolytic ‘‘seeds’’ [5]. This is indeed supported by US clinical data where 40 % of all human metastases also occupy the brain [6]. Additionally, if the brain’s BG microenvironment (‘‘soil’’) and its control are more favourable for highly glycolytic metastases (‘‘seeds’’) than other organs, we would expect that primary brain tumours will spread within the human brain, but rarely metastasize to other organs. This is also confirmed clinically [7, 8]. Also, the normal rodent brain has nearly five times lower glucose metabolism (and BG control) than the human brain with FDG-SUV values of around 1.7 [9, 10]. We therefore suspect that naturally occurring brain metastasis in rodents will be rare. This is indeed observed in practice [1, 11]. The less ideal BG energy microenvironment and control in rodents might also explain why experimental gliomas regress spontaneously in the rodent brain [12], but not in the human brain [11]. The only option we see at present to improve the representativity of in vivo rodent models is to equalise the rodent and human BG microenvironments and their control [13]. One practical way of obtaining BG similarity and control is to employ glucose haemodialysis [4, 14]. This can be achieved by either downregulating the human’s BG or by upregulating the rodent’s BG levels. It should however be easier to perform such treatment with humans considering maturity of the human haemodialysis process. (It has not escaped our notice that such extracorporeal control of BG levels could potentially also be used to treat other HCGM [15].) Daphu et al. [1] also explain the utility of xenotransplantation rodent models. Here again we have to investigate the microenvironments and their consequences. The average tissue microenvironment in healthy humans comprises of approximately 6 mmol/L BG [16]. However, xenograft transplantation models typically employ cell cultures grown in 25 mmol/L (or higher) glucose medium [17, 18]. Highly glycolytic cancer cells derived from 25 mmol/L solution are expected to be more sensitive to experimental chemotherapeutics than those from 6 mmol/L solution E. H. Mathews L. Liebenberg (&) Center for Research and Continued Engineering Development, North-West University, Pretoria Campus, Suite No. 91, Private Bag X30, Lynnwood Ridge 0040, South Africa e-mail: lliebenberg@researchtoolbox.com

Preliminary Application of a New Bolus Insulin Model for Type 1 Diabetes

BACKGROUND Previous work derived a theoretically rigorous bolus insulin model. It was shown that the new model predicts insulin response of subjects without diabetes substantially better than the carbohydrate counting method (CHOcm). As most individuals with type 1 diabetes use the CHOcm, this article investigates if the new model can be applied to them. METHODS Equations are derived to characterize a person with type 1 diabetes. These are implemented on a cell phone that calculates bolus insulin dosages. In a small feasibility study the cell phone was used by 11 patients. Basal insulin remained unchanged. The subjects were experienced in the CHOcm and were using it at the start of the study. Continuous glucose monitoring data were recorded to capture blood glucose (BG) control elements such as average BG and tightness of control, as well as hypoglycemic and hyperglycemic events. A new rating method was proposed to estimate BG control. It (or a derivative thereof ) may in the future become a valuable measure of total glycemic control. We used the status quo ante versus status quo approach to find indicative results. BG control for the same group when using CHOcm (status quo ante) was compared with BG control when using the new application (status quo). RESULTS Patients found the new application on the cell phone practical. Indicative results also showed BG control improvements, although the subjects were more experienced in the CHOcm. Depending on the weights assigned to the underlying control elements an improvement of between 26% and 64% was found. CONCLUSION An indicative study (status quo ante vs. status quo) on 11 patients with type 1 diabetes showed that the new method can be practically and successfully applied on a cell phone. Glycemic control even improved. A new BG rating method was proposed. We believe there is enough preliminary indication to warrant a more detailed clinical trial in future by an institution with adequate funds and access to more patients.