Thesis (M.S.P.H.)--University of Michigan, 1939.

Thesis (M.S.P.H.)--University of Michigan, 1942.

Thesis (M.S.P.H.)--University of Michigan, 1942.

Thesis (M.S.P.H.)--University of Michigan, 1942.

Thesis (M.S.P.H.)--University of Michigan, 1939.

Thesis (M.S.P.H.)--University of Michigan, 1942.

Thesis (M.S.)--University of Wisconsin--Madison, School of Nursing, 1974. / eContent provider-neutral record in process. Description based on print version record. Includes bibliographical references (leaves 57-60)

The histogenesis of two squamous-carcinoma cell lines, originally derived from the same human cervical tumor but with different morphology in vitro and in vivo, was compared histochemically and electron microscopically. Inoculation into hamster cheek pouches was used to distinguish in vitro from in vivo responses in the two cell lines. Normal cervical epithelium was grown in vitro to determine which histogenetic characteristics were due to the malignant nature of the cell lines. Possible mechanisms underlying histogenetic variation were investigated by chemical and physical modifications of the in vitro environment. In vitro cells of line C-4c were more cohesive, less deformable and less adhesive to substrata than the cells of line C-4s and they formed more compact and highly stratified colonies. Line C-4s responded to crowding in vitro by cell separation (dispersal) while line C-4c increased stratification. Histochemically the cell lines were essentially similar but they differed in ultrastructure, particularly at the level of tissue organization. The C-4s cells were columnar, with a higher nucleo-cytoolasmic ratio, a polarity of the organelles, dispersed cytoplasmic filaments, little stratification and with terminal bars between superficial cells. In contrast, C-4c cells were oval, stratified and with superficial cells flattened, with a lower nucleo-cytoplas-mic ratio, without polarity of organelles, and with more cytoplasmic filaments condensed into bundles that were associated with desrnosomes. In both lines, but particularly in line C-4c, interdigitating microvilli provided the main intercellular contact and the cell surfaces were modified in association with both substrata and the growth medium. Most in vitro differences between the cell lines were retained in vivo, and, in addition, incomplete basement membranes were formed in both lines, though more extensively in line C-4.s. It appeared that the only major difference in vitro between benign squamous cells and the carcinoma cells lay in the ability of the latter to maintain an intercellular organization in the complete absence of any supporting tissue. Possible mechanisms underlying these observed differences in histogenesis were experimentally investigated. Ferritin uptake in vitro indicated that the difference in stratification was not due to more efficient intercellular circulation in the more highly stratified C-4c line. No difference in cell-surface charges or distribution of cell-surface acid mucosubstances could be demonstrated, suggesting no difference in contact inhibition to be associated with the stratification. It was found that C-4c cells were more cohesive and that this cohesion, in both lines, depended predominantly on the presence of divalent cations. In contrast, adhesion to the substrata required extracellular proteins, probably accompanied by the masking of acid groups on the cell surface. These and other described observations suggested a defect in deformability in these cell lines with the probability that the stronger cohesion in line C-4c was due to the more extensive microvillus population. The in vitro maintenance of cell shape, of tissue organization and of cell separation (dispersal) seemed to involve an interaction between these intercellular adhesive forces and the cytoplasmic filaments. On the other hand, the specific, modifications, such as terminal-bar formation and cell flattening, which appeared in relation to growth medium, to other cells and to substrata could be shown to be responses of the cells to the viscosity of the immediate environment and thus experimentally modifiable. The results of this investigation are discussed relative to the specific suggestion that much of the histogenetic variation could be explained by considering that the more cohesive cell line (C-4c) had retained properties of normal epithelium stratum spinosum cells, while the spreading cell line (C-4s) exhibited characteristics of basal cells, and relative to the interaction of differentiation capacity and malignancy in tumor development. / Science, Faculty of / Zoology, Department of / Graduate

A research report submitted to the Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, in partial fulfilment of the requirement for the degree of Master of Science in Epidemiology and Biostatistics. November 2017. / Introduction: Pancreatic ductal adenocarcinoma (PDAC) is a devastating diagnosis for anyone as it is associated with a global mortality rate of about 4%, and has few therapeutic interventions that prolong survival as compared to other cancers. Frequent epidemiological reports on PDAC are available in the developed countries, but in South Africa, there is a paucity of epidemiological data on this aggressive cancer. Understanding risk factors will help to assess and develop relevant interventions for asymptomatic high-risk patient populations. Aim: To investigate and explore how various risk factors were associated with PDAC at two public academic hospitals in Johannesburg between 2013 and 2015. Method: This was a secondary unmatched case-control study to assess risk factors for developing PDAC at two public academic hospitals, namely the Chris Hani Baragwanath Academic Hospital (CHBAH) and the Charlotte Maxeke Johannesburg Academic Hospital (CMJAH). All cases of PDAC were histologically and/or cytologically confirmed. All participants were >18 years of age, including 139 cases and 139 controls. Data exported from REDCap database included patient demographics and social and medical histories. Proportions used the chi-square test and bivariate logistic regression estimated ORs between individual variables and PDAC. Multivariate logistic regression analysis investigated all possible confounders present in the data. The likelihood ratio test with a p-value of <0.20 was accepted to assimilate data fitting into the model. Results: Eighty two percent of the study population was black. The 50-59 age group accounted for 37% of the cases. Multiple logistic regressions showed the following odds ratios 95% CI and p-values for ages (i) 20-29 [0.11(0.11-1.00) p=0.05] and (ii) 50-59 [2.63(1.03-6.70) p=0.04]. As for diet, the following odds were observed (i) high white meat [0.18(0.04-0.86) p=0.03], (ii) low fish intake [2.17(1.06-4.45) p=0.03], (iii) low consumption of fried food [0.48(0.23-1.00) p=0.05] and (iv) high consumption of vegetables [0.17(0.05-0.61) p=0.007]. In terms of occupation, general workers had the following likelihood [1.79(0.93-3.45) p=0.08] of developing PDAC. Conclusion: Being 50-69 years of age and employed for longer periods than the general norm, was positively associated with PDAC. Additionally, increased consumption of vegetables and white meat was protective against PDAC, whilst a low intake of fish increased PDAC risk. Keywords: Pancreatic cancer, risk factors, epidemiology and case-control. / LG2018

Background: Oesophageal cancer is the fourth most common cancer in black South African males with an age-standardised incidence rate of 7.31 per 100 000 in 2010, and sixth in black South African females, 3.59 per 100 000. The adoption of lifestyle behaviours such as tobacco and alcohol use was on the rise in that decade, amongst black South Africans living in Johannesburg. Aim: The overall aim was to investigate the association of smoking and alcohol use on oesophageal cancer amongst 18 to 74-year-old black South Africans in Johannesburg from 1999 to 2009. Methods: Secondary analysis were done based on a matched case-control study design. We used the conditional univariate and matched multiple logistic regression for statistical analysis as the main analysis. Gender was ascertained as an effect modifier therefore the analysis were done separately for males and females. Sensitivity analysis of the findings were tested using the unconditional univariate and unmatched multiple logistic regression. Results: Heavy smokers had an increased likelihood of oesophageal cancer as shown in the conditional multivariate logistic regression (AOR = 9.0; 95% CI: 5.2 – 15.5) in males and (AOR = 5.2; 95% CI: 1.3 – 20.1) in females. Alcohol consumption was a much stronger risk factor for oesophageal cancer among female heavy drinkers (AOR = 2.1; 95% CI: 1.5 – 2.9) relative to the light drinkers, controlling for other variables. There was a significant interaction of tobacco use and alcohol drinking as they acted synergistically to increase the likelihood of oesophageal cancer among current heavy smokers and heavy drinkers, (AOR = 15.0; 95% 7.2 – 31.3) in males and (AOR = 2.7; 95% 0.6 – 11.2) in females, compared to non-drinkers. iv Conclusion: We established that over the ten-year study period, smoking and alcohol use were both associated with oesophageal cancer independently and as combined exposures. An increase in sin tax on cigarettes and alcohol as well as increased education on the risk factors associated with the development of oesophageal cancer could be used as interventions to decrease the burden of this disease. Keywords: Smoking, Alcohol, Cancer, South Africa / LG2018