INTRODUCTION To evaluate the clinical manifestations, staging and prognostic factors for Hepatocellular Carcinoma (HCC) in Gambian patients.
METHODS All patients with suspected HCC referred to the main liver clinic In Medical Research Council, Gambia Unit (MRCG) were recruited between December 2015 to January 2019. The diagnostic criteria was based on ultrasound demonstration of liver mass ≥2 cm combined with alpha-fetoprotein (AFP) level of ≥200ng/ml and/or histopathology confirmation. Kaplan -Meier, univariate and multivariate cox regression analysis were used to assess factors related to survival in these patients.
RESULTS Two hundred and sixty patients were recruited into the study. The mean age of HCC patient was 40 years and mostly males (80.7%). The majority were rural born (74.4%), of the Wollof tribe (26.2%) and farmer/gardener (42%). The most common constitutional symptoms were weight loss (92.3%), easy fatiguability (91.9%) while the most common gastrointestinal symptoms were early satiety (89.1%) and abdominal pain (88.7%). The most common signs were hepatomegaly (83.7%) and abdominal tenderness (44%). Multi-focal lesions were the most common on ultrasound scan(67.2%) and the median fibroelastography score was 75kpa in these patients. HBsAg carriage was present in 66.4% of HCC patients with a median AFP of 3895 ng/ml. WHO performance status 3, BCLC stage C and Child-Pugh score stage B were most common among this patient group. HBsAg-positive patients with HCC were mostly males, much younger, most likely to have abdominal pain, jaundice, dark urine, abdominal tenderness, raised transaminases and decreased platelet counts as compared to patients with HCC who are HBsAg-negative. Patients diagnosed with HCC who were HBsAg-negative tended to be older, more likely to be hypertensive and had a much better median survival ( 45 days vs 31 days). Both prognostic staging systems had good stratification of survival and the median survival of these patients was 35 days. Independent factors that affect survival were (i) Patient related, (sex, M vs F:p=0.022), (ii) tumour related (abdominal pain:p=0.020, oedema:p=0.003, jaundice:p=0.040, ascites: p=0.012), (iii) disease stage at presentation (child pugh scoring system,AB vs C: p=0.025) and Biochemical indices(AFP,≥200ng/ml:P=0.001,sodium:p=0.008, glucose p=<0.001, AST:p=<0.001 and albumin:p=0.001).
CONCLUSIONS HBV is a significant factor in HCC in The Gambia. Young males who are the main workforce are disproportionately affected, are more likely to be symptomatic and have much shorter survival. In resource limited countries were screening programmes and therapeutic interventions are limited, the prognosis and survival of patients with HCC is poor, emphasizing the need for early preventive strategies.

BACKGROUND Male breast cancer is a rare disease; it accounts for less than 1% of all breast cancer cases and less than 1% of all malignancies in men. The incidence is much higher in sub-Saharan Africa. Male Breast cancer is less known and thus late presentation and delayed diagnosis in sub-Saharan Africa. Due to its rarity, there is limited literature regarding trends, incidence and relevant management of Male breast cancer.
METHODS In this study, we describe the trends and overall incidence of male breast cancer in Moi Teaching and Referral Hospital, Eldoret from 1997 to 2017 using data extracted from Eldoret Cancer Registry.
FINDINGS A total of 23,704 cancer cases were registered between Jan 1997 and Jun 2017, 1864 (7.9%) of these were breast cancer cases of which 117 (6.2%) were male. The median age of male breast cancer patients was 61 (27-91 years). About 40% were over the age of 65. Infiltrating ductal cell carcinoma was the commonest morphological subtype and late stage presentation was common.
CONCLUSION Male breast cancer remains a rare disease, although the incidence is rising. The incidences in Moi Teaching and Referral Hospital were higher than in developed countries but comparable with other Sub-Sahara African countries. Incidence increased significantly with advancing age. Opportunities to accelerate early diagnosis and prompt treatment exist through improved disease awareness and expansion of access to breast cancer screening among men. Multicentre studies are also encouraged to provide more data on male breast cancer.

OBJECTIVE To examine the association of oral health with risk of oesophageal cancer (EC) in Kenya.
METHODS A case-control study was conducted at the Moi Teaching and Referral Hospital, in Eldoret, western Kenya from August 2013 to September 2014 (pilot phase) and October 2015 to April 2018 (main study). Cases were patients aged ≥ 18 years who presented to the endoscopy unit with progressive dysphagia and suspicion of EC. Cases were histologically confirmed oesophageal squamous cell carcinoma ESCC (90%) and endoscopically visualized the tumour (10%). Age and gender frequency-matched controls were recruited from hospital visitors and hospital patients. Trained interviewers conducted face-to-face interviews, collected blood, saliva, urine samples and conducted oral examinations. Using logistic regression models for ESCC, odds ratios (ORs) and their 95% confidence intervals (CI) were estimated associated with oral hygiene, DMFT, leukoplakia, dental fluorosis and water source. Stratified analyses by interviewer and control type were conducted to investigate potential biases and by participant characteristics to examine effect modifiers. Statistical analyses were conducted in Stata version 14.0 and mapping in QGIS.
RESULTS Four hundred and thirty (430) oesophageal cancer cases 18 and 440 controls were enrolled. Tooth brushing once a week or less and using miswaki rather than a toothbrush was more common in cases than controls. Cases had higher DMFT (number of decayed + missing + filled teeth) scores (median 7 vs 3). The odds ratio (OR, 95% confidence interval), adjusted for known lifestyle risk factors, was 6.4 (3.4 to 12.1) for a DMFT score of ≥ 8 vs 0. Moderate or severe dental fluorosis, i.e. an irreversible enamel hypo-mineralization due to early-life fluoride intakes, had an OR of 14.7 (7.6 to 28.6). Oral leukoplakia was almost always (90%) present with moderate/severe fluorosis, compared to 16% with no fluorosis. Compared to having piped water, ORs were 1.6 (0.9 to 3.0) for well and 3.0 (1.6 to 5.5) for spring/river sources.
CONCLUSION Dental fluorosis, which occurs in this setting due to early-life fluoride intakes naturally occurring in ground/surface water, poor modern oral hygiene and tooth decay are strong markers of oesophageal cancer risk in Kenya. Poor oral health in combination with high-altitude susceptibility to hydro-geologically influenced exposures may underlie the steep spatial gradients in incidence rates and the striking co-location of East Africa’s oesophageal cancer corridor with the African rift valley.

OBJECTIVE Previously, we and others described a tumour-specific immune-inflammation gene expression signature in prostate tumours that was more prevalent in tumours of African American (AA) than European-American (EA) patients. The occurrence of an interferon signature, a subset of the immune-inflammation signature, in prostate tumours of AA patients was associated with decreased overall survival of these patients. The clinical impact of the inflammation status of tumours from these patients was further suggested by the finding that regular aspirin use, an anti-inflammatory drug, significantly reduced the risk of aggressive prostate cancer and disease recurrence in AA men. Combined, these findings suggested that immune inflammation contributes to prostate cancer progression. While various environmental exposures may promote inflammation, ancestry could also influence the inflammatory processes. It has been shown that allele frequencies of genetic variants in inflammation-related genes can markedly differ amongst population groups suggesting that ancestral factors may influence inflammatory processes.
METHODS To determine if men of African descent are differentially affected by a systemic inflammatory process, we measured 92 immune-inflammation proteins in serum samples of 1519 prostate cancer cases (659 Ghanaian, 411 AA, and 449 EA) and 1518 population controls (659 Ghanaian, 390 AA, 469 EA) using Proximity Extension Assay and assessed if the levels of these proteins vary in different population groups.
RESULTS Unsupervised hierarchical clustering of population controls according to serum levels of the immune-inflammation markers revealed that the Ghanaian samples grouped together and so did the AA and EA samples indicating that these markers alone are able to separate the three different population groups. Notably, the Ghanaian samples grouped closer to AA samples than EA samples suggesting that ancestry may play a role in the differences observed in the levels of these immune-inflammation analytes. Grouping these analytes by the biological process revealed immune markers associated with chemotaxis, tumour immunity suppression, vasculature & tissue remodelling were relatively elevated in both the Ghanaian and AA population controls compared to EA.
CONCLUSION Our finding provides preliminary evidence that there is a unique systemic immune-inflammation signature in men of African ancestry.

OBJECTIVE The estimated incidence of malignant melanoma in Botswana is higher than among African Americans in the United States. We sought to assess whether solar radiation or HIV could account for the observed increased incidence of melanoma in Botswana.
METHODS Cases were drawn from the Thabatse Cancer Cohort (TCC) from 2010-2017. Age-matched controls (10:1) were drawn from two sources: 1) participants in a large random population sample from rural/peri-urban communities and 2) non-melanoma and non-cutaneous cancers in TCC. We utilized exact conditional logistic regression to compare characteristics of cases and the two groups of controls.
RESULTS Thirty cases of melanoma were enrolled, including 17 (57%) women and 13 (43%) men. The median age was 64.6 (IQR 49.1 to 71.5), and 6 (22%) were HIV-infected. Twenty-eight (93%) cases reported Tswana ethnicity. The majority (63%) presented with advanced disease (stages III/IV), and 79% were acral. Eleven (37%) patients reported spending >4 hours daily in the midday sun without shading head and arms. Compared with community controls, no significant associations were identified for HIV status (OR 0.6, 95% CI 0.17 to 1.57, p=0.34) in patients with melanoma. Utilizing cancer controls, there was a significant association between Tswana ethnicity and melanoma (OR 6.8, 95% CI 1.45 to 70.39, p=0.008). No association was detected between melanoma risk and prolonged (>4 hours) midday, uncovered solar exposure (OR=0.99, 95% CI 0.41 to 2.26, p=1.00) or reported frequent utilization of hats, long sleeves, or umbrellas (OR= 0.92, 95% CI 0.32 to 2.33, p=1.00).
CONCLUSION Advanced stage and acral presentation of melanoma are common in Botswana. While conclusions are limited by the small number of cases, we did not identify HIV or heavy sun exposure has important risk factors for melanoma. Observed associations between ethnic groups deserve future study and could represent an important risk in this population.

BACKGROUND Socio-demographic characteristics can play major roles in access to oncology care and survival from Breast and gynaecological cancers in a multi-ethnic middle-income country such as South Africa (SA).
OBJECTIVES We evaluated some socio-demographic factors associated with mortality from breast, cervical and ovarian cancer in SA from 1997 to 2016.
METHODS Three unmatched case control studies of breast, cervical or ovarian cancer mortality were conducted based on the mortality data of Statistics South Africa. Mortality from breast (n=51,435), cervical (n=55,317) or ovarian (n=11,623) cancer were the cases, while controls (n= 86,054) were other female cancer mortality with no known association to breast and gynaecological cancers. Descriptive and unconditional multivariable logistic regression analysis was conducted with sociodemographic characteristics as covariates. Stata version 14 (StataCorp, USA) Software was used for analysis.
RESULTS The mean age at death was respectively 55.9 ±14.6 years, 59.7±15.5, 61.0±15.1 for cervical, breast and ovarian cancer while the mean age at death of the controls was 59.4 ±0.1 years. The likelihood of mortality from Cervical cancer or breast cancer as compared to deaths from other female cancers was about 51% (OR= 1.51, 95% CI: 1.43 – 1.60, P-value <0.001) and 18% (OR=1.18, 95% CI: 1.11 – 1.24, P-value<0.001) higher among women younger than 50years as compared to older women. But there was a 16% lesser odds of death from ovarian cancer among women younger than 50 years (OR= 0.84, 95% CI: 0.75 – 0.93, P-value=0.001). The odds of mortality from breast and ovarian cancer increases with increasing years of formal education (Ptrend <0.001) but the odds of mortality from cervical cancer decreased with increasing years of formal education (Ptrend <0.001). Other factors associated with cancer mortality includes marital status, province of residence and smoking status.
CONCLUSION Young women are more likely to die from breast and cervical cancer as compared to other female cancers in SA. Socio-economic status and place of residence play major roles in breast and gynaecological mortality in SA. Integration of breast and cervical cancer preventive efforts into sexual and reproductive health initiatives and reduction in Socio-economic inequity can reduce the burden of breast and gynaecological cancers in SA.

OBJECTIVE Endemic Burkitts Lymphoma (eBL) accounts for half of all pediatric malignancies in equatorial Africa. eBL is associated with Epstein-Barr virus (EBV) and geographically with malaria. However; eBL incidence does not mirror the high prevalence of both infections, suggesting other risk-factors may be important. Therefore, we examined risk-factors in children from Uganda, Tanzania and Kenya.
METHODS Using standardized questionnaires and protocols, cases and matched population-controls were recruited, 2010-2016. Cases were histologically and/or clinically confirmed. Adjusted odds ratios of malaria-RDT and questionnaire variables association with eBL were computed using “baseline” logistic models adjusted for age, sex, and village characteristics and further for variables with a p<0∙05 in the univariate “baseline” models.
RESULTS Of the 862 cases spotted, 697 eBL cases and 2,934 of the 2,970 invited population-controls were enrolled. Cases and controls had similar mean ages in Uganda (8∙0 versus 7∙7 years) and Tanzania (6∙8 years versus 7∙4 years) but they were younger in Kenya (6∙6 years versus 7∙4 years). eBL risk was decreased in children with higher maternal income (p-trend <0∙001) and paternal education (p-trend <0∙033), who reported malaria-attributed fever up to 6 months before enrolment (aORs=0∙48-0∙59), and tested malaria-RDT positive (aORs=0∙33-0∙43-fold). Risk of eBL was significantly elevated in children reporting indoor residual insecticide spraying (aORs=1∙71-6∙78), and ownership but non-use of bed nets (aORs=5∙74-39∙7), inpatient malaria treatment >12 months before enrolment (aORs=3∙97-2.89), non-malaria-attributed fever within 6 months (aORs=3∙88-8∙82) and HIV (ORs 4∙32-25∙6).
CONCLUSIONS We identified known eBL associations with malaria (lower socioeconomic status) or immunosuppression-related BL (HIV), and novel associations opposite of what was predicted (mass malaria suppression) or not reported before (malaria-RDT, malaria history or other fevers ≤6 months before interview). Our findings update eBL epidemiology in East Africa, strengthens the evidence for the role of socioeconomic factors, and sheds new light on eBL and malaria history.

OBJECTIVES The goal of this presentation is to provide an overview of intracranial tumor types seen in a large urban forensic population and correlate these tumour types with cause of death.
BACKGROUND Central nervous system tumours may only come to attention at autopsy. Past forensic surveys, which involve high numbers of unattended deaths, very often among vulnerable or underserved populations, have noted intracranial tumours format a frequency of 0.02-2.1%. Thus, forensic autopsies, particularly in a large, diverse urban setting, can contribute to a public health surveillance function by highlighting under-appreciated trends. Here, we sought to characterize all cases with autopsy evidence of intracranial mass lesions at the office of the Chief Medical Examiner in the City of New York, during a 12-year period.
METHODS We screened cases referred for Neuropathology consultation from 2004-2014, and 2017-present for decedents with intracranial neoplasms and mass lesions such as cysts. Demographic information, diagnoses, and cause of death were recorded.
RESULTS Among 12,295 cases referred for Neuropathology evaluation, we identified 197 cases (a prevalence of 1.6%), of which 91 were women, and 106 men. The ages ranged from 2 weeks to 92 years, with 14 cases (7%) less than 20 years of age, 53 cases (26.9%) in the 20-50 age range and 130 (65.9%) greater than 50. Most were Black/African American (n=81, 41%); 23.8% were Latino/Hispanic (n=47); 28.4% were White/Caucasian (n=56); 6.5% were Asian/Pacific Islander (n=13). There were 12 main types of primary intracranial mass lesions, including meningioma (30.4%), pituitary adenoma (9%), GBM (5.5%), lymphoma (3.5%), cysts (arachnoid, pineal, epidermoid; 9.6%), colloid cyst of the third ventricle (4.5 %), central neurocytoma (9%), subependymal giant cell astrocytoma (2.5%), and lipoma (1.6%); gliomas of grades I-III comprised 2%. Metastatic (secondary) intracranial neoplasms made up 19%. The mass lesions contributed to the cause of death in 104 (52.8%) and were previously undiagnosed in 99 (50%). Complications of therapy (including surgery) contributed to death in 17(8.6%) instances.
CONCLUSIONS Our forensic cohort shows a diverse assortment of tumour types, over half of which were undetected in life. Tumours were directly related to the cause of death in a majority of cases, including an appreciable proportion with therapeutic (operative) complications. Finally, we further emphasize the value of autopsy in diagnosis and public health surveillance for oncologic disorders and their outcomes.

BACKGROUND The incidence of breast cancer rises dramatically with age, the single most important risk factor for developing cancer. Large proportion of all new breast cancer diagnoses and the majority of breast cancer deaths occur in women 65 years and older. A large number of studies in the literature reported the benefits of post mastectomy radiotherapy in both the young and elderly breast cancer patients. Limited numbers of radiotherapy machines and practice of conventional fractionation schedule which is long in duration makes radiotherapy access more difficult especially among the eldely. Hypofractionated radiotherapy offers alternative option with shorter duration and comparable side effects and outcome to conventional fractionation.
METHODS Records of elderly patients treated with radiotherapy in the department of radiotherapy and oncology from January 2017 to January 2019 were evaluated. Post mastectomy radiotherapy was given at a dose of 45 Gy in 16 fractions of 2.81Gy over 16 days to the chest wall and the draining regional lymph nodes. Our primary end point was freedom from any grade 3 or higher toxicities and disease free survival at 2 years.
RESULTS Eighty three 83 elderly women with stages II and III breast were recruited. Majority had invasive ductal carcinoma not otherwise specific (78.3). More than half of the patients (69.9%) had the disease located in the right breast. After a follow up of 24 months, there were no grade 3 or 4 toxities. 4 and 6 patients had grade 2 skin and Nausea and vomiting toxicities respective. Seventy nine (95.2%) were disease free at 24 months while 2 (2.4%) patients each had local and distance metastasis respectively.
CONCLUSION Hypofractionated radiotherapy in the elderly breast cancer patient proved to be comparable to conventional radiotherapy in terms of toxicities and outcome in sub-Saharan Africa.

OBJECTIVE Obesity is associated with worse breast cancer (BC) outcomes in High-Income Countries (HICs). However, few studies have explored this association in Low and Middle Income Countries (LMICs). We evaluated the relationship between weight and non-metastatic BC outcomes in Haiti.
METHODS We conducted a retrospective observational study including 224 non-metastatic BC patients who presented between June 2013 and December 2016. We categorized weight at surgery by body mass index (BMI): normal weight (<25), overweight (25-29.9), and obese (>30). We described patient characteristics, diagnostic and treatment data, and outcomes. Our outcome endpoint was disease-free survival (DFS), time from surgery to recurrence, progression or death. We used Kaplan-Meier estimation to plot survival curves, censoring at last follow-up. Log-rank test was used to examine subgroup differences. Cox proportional hazard model was used to estimate the association between weight categories and DFS, while controlling for age, menopausal status, time to treatment, home location, stage, and pathologic grade.
RESULTS In this cohort, 83 (37%) were normal weight, 66 (29%) were overweight, and 75 (33%) were obese or larger. There were no statistical differences in baseline characteristics between the groups. The mean age was 49; 36% were post-menopausal; and 58% had locally advanced disease. For the cases with documented ER status, 63% were ER positive. Median follow-up time was 21.7 months. 26 patients died; 73 had disease recurrence. Median DFS for the whole cohort was 41.2 months (95% CI, 37.5 to 48.2). There was no significant DFS difference by weight status: median DFS for the weight categories were 44.4, 48.1, and 41.2 months respectively (Log-rank test P= 0.84). Compared to normal weight, overweight and obese patients had adjusted hazard ratios of 0.96 (95% CI, 0.55- 1.69), and 1.02 (95% CI, 0.59- 1.76) respectively.
CONCLUSIONS There was no difference in DFS by weight in this Haitian cohort. This finding may be explained by the disparate epidemiology of obesity in HICs compared to LMICs; higher body weights may be associated with higher socioeconomic status in LMICs. More studies are needed in LMICs to clarify the association between weight and BC outcomes.