Prostate Cancer Determinants, Disease Severity and Treatment Outcomes at the Sweden Ghana Medical Center in the Greater Accra Region of Ghana

Abstract

Background: Prostate cancer is a global health concern, with varying epidemiological patterns across populations. This study investigates the demographic/physical attributes, disease characteristics, and treatment outcomes of prostate cancer patients at the Sweden Ghana Medical Centre (SGMC) to provide insights into the Ghanaian context. Methods: An analysis of retrospective data from 852 prostate cancer patients who visited SGMC from 2011 to 2023 was conducted. Demographic information, which included age , ethnicity, marital status, occupation, socio-economic status, level of activity, place of residence, family history, alcohol consumption, and tobacco use, was assessed. Disease characteristics included clinical stage, risk stratification, and PSA levels . Treatment outcomes studied included, PSA response and incidence of toxicity. Descriptive statistics, correlation analysis, logistic regression analysis plus modelling was done and odds ratios calculated; all at an alpha of 0.05%. Results: Mean age was 67.5 years; median was 68.0 years. Median PSA at diagnosis was 29.0 ng/ml. There were notable variations in weight (mean 76.8 kg, SD 12.7), height (mean 1.71 meters, SD 0.07), and body mass index (BMI) (mean 26.3 kg/m², SD 4.1). Ethnicity, was predominantly Akan (55.52%), Ga (13.38%), Ewe (14.91%), men from Northern Ghana (6.34%), Nigerians (8.57%), Other Africans/Jamaicans (0.95%), Caucasians/Asians (0.34%). Patients, aged > 65 years, (OR = 2.34, p = 0.022) and high BMI (OR = 2.34, p = 0.022), were associated with high PSA; and BMI alone, with high risk localised prostate cancer, but a reduced propensity to metastatic prostate cancer. Ga and Ewe Ethnicities were associated with low risk localised prostate cancer on DRE (OR = 0.52, p = 0.049). High socio-economic status (59.32%) and sedentary occupations (56.5%) predominated. Urban residence was prevalent among patients (74.79%), with a noteworthy number of foreigners (10.18%). Family history (24.65%), alcohol consumption (31.3%), and tobacco use (8.6%) exhibited varying prevalence rates. Comorbidities were relatively uncommon, but with hypertension being the most frequent (10.55%). Late diagnosed prostate cancer was commonplace; 46.81%. Metastatic prostate cancer rate was 28.72% of cases. Risk stratification indicated a considerable proportion of overall high-risk disease in patients with localised disease (44.75%). External beam radiotherapy was the primary treatment modality (76.88%), often combined with hormonal therapy (55.06%). The mean PSA response per unit of treatment dose was 31.60 ng/ml per Gray; and response rate was 53.16%. Treatment toxicity was infrequent (6.09%). Overall survival and quality of life data were however, not available. PSA, BMI, DRE, ISUP, predicted metastasis-probability in prostate cancer in a mathematical model with Sensitivity of 85.65%; Specificity of 95.45%; Positive Predictive Value of 90.84%; Negative Predictive Value of 92.65% and an accuracy of 75.31%. AUC of ROC curve = 90.55%; YIELDp= 68.71% and YIELDs= 95.45%. A prostate cancer case-detection model which estimates an individual’s prostate cancer risk(between 0 and 1) using PSA, BMI, linear weignt-to height ratio, age, marital status, ethnicity, family history of prostate cancer, socio-economic status, tobacco use and sedentary, or non-sedentary occupation in one model, was also obtained. It outperformed a PSA alone model, considerably. Conclusions: Though limited by secondary data, Age, BMI, Ethnicity, marital status, occupation, and socio-economic status were identified as key determinants. Aggressive preventive measures on all fronts are are needed to diagnose more early stage prostate cancer, nationally, to improve prostate cancer outcomes.