The increasing trends in weight gain and obesity and the subsequent rises in the incidence of associated comorbid conditions are a growing concern in public health. Obesity is an established risk factor for the development of both breast and ovarian cancers. The role of obesity in the survival of women with these cancers is less well understood. Early studies of obesity and breast cancer survival have been summarised in two meta-analyses, but the latest of these only included studies that recruited women diagnosed as recently as 1991. Studies examining the role of obesity in ovarian cancer survival have provided conflicting results. Currently, the causal mechanisms underlying the potential association between obesity and cancer survival are poorly defined and need further investigation.
The overall aim of this Thesis was to use epidemiological methods to further understand the relationship between obesity and survival in women with breast and ovarian cancers. Methods The analytical studies that form this Thesis used multiple data sources and statistical methods. Systematic reviews and meta-analyses of existing evidence were conducted to provide updated pooled estimates of the degree of association between obesity and survival in women with breast and ovarian cancers, as well as to explore important sources of variability. An extensive medical record review of the chemotherapy dosing practices in women (n=354) receiving adjuvant chemotherapy for breast cancer at a major tertiary hospital in Australia between 2002 and 2010 was conducted to explore the potential role of chemotherapy under-dosing (measured by planned dose-capping and relative dose intensity (RDI)) as a causal mechanism for the survival disadvantage observed in obese women. To further explore the association between obesity and ovarian cancer survival, a large cohort (n=1404) of Australian women, diagnosed with ovarian cancer between 2002 and 2006, were followed up until October 2011 and the association between obesity and survival was analysed. Statistical methods used in this Thesis include: Random effects meta-analysis and meta-regression to quantitatively summarise the existing literature; multivariate logistic regression analysis to assess whether obesity is an independent predictor of chemotherapy under-dosing; and Cox proportional hazards regression modelling to measure the association between obesity and survival in Australian women diagnosed with ovarian cancer.
Forty three studies were identified that have investigated the association between obesity and breast cancer survival. The meta-analysis showed poorer survival among obese compared with non-obese women with breast cancer (hazard ratio (HR) 1.3). This survival differential was similar regardless of how obesity was measured or how survival was defined (cancer-specific versus overall survival), however larger differences were observed if women were in observational cohorts (HR 1.36) compared to treatment cohorts (HR 1.2). In the medical record review of women who underwent chemotherapy for breast cancer at a major teaching hospital in Australia, obesity appears to be a major predictor of planned dose-capping (odds ratio (OR) 6.6). While, on average, obese women received lower RDI of chemotherapy compared to non-obese women (mean RDI: 88% in obese and 91% in non-obese), obesity was not a significant predictor of receiving a RDI of less than 85% (p=0.68). Comparatively few studies have examined the association between obesity and ovarian cancer survival. The meta-analysis of 14 studies showed slightly poorer survival among obese than in non-obese women (HR 1.2), however there was significant heterogeneity in the definition and timing of obesity measurement and most studies were based on relatively small sample sizes. The results of our large cohort of Australian women with ovarian cancer show results consistent with the meta-analysis (HR 1.2). Obesity conferred a survival disadvantage for both short- and long-term survival (<5 and ≥5 years respectively), with a stronger association in long-term survivors.
The findings of this Thesis confirm the adverse role of obesity in the survival of women with breast and ovarian cancers and provide updated estimates of the strength of the association, based on the available literature. Chemotherapy under-dosing in obese women being treated for breast cancer still appears to be a problem in Australian practice, and may indeed explain some of the survival differential between obese and non-obese women with breast cancer. Australian oncology guidelines should be updated, in line with American guidelines, to discourage the practice of chemotherapy under-dosing based on a woman’s body size, especially in the curative setting. The role of obesity in ovarian cancer survival is less defined. More large scale, population-based studies need to be conducted with a specific focus on examining potential differences by histologic subtype, as well as the role of obesity-related comorbidities, such as diabetes, and survival from ovarian cancer.