Oesophageal Carcinoma: a neglected volcano in Sri Lanka

Abstract

Background: Risk factors of Oesophageal Carcinoma (OC) specific to Sri Lanka are important for implementing primary prevention. Owing to late symptoms and absence of routine endoscopic screening, delayed presentation leads to severe outcomes. Thus, developing a simple low-cost risk prediction tool to identify high-risk individuals based on population-specific risk will enable early diagnosis and treatment as secondary prevention. Though quality of life (QOL) is important in tertiary prevention of OC, its status following the initial treatment has not been assessed lately in Sri Lanka. Objectives: To identify the risk factors of OC specific for adults in the Western province, to develop and validate a risk prediction model to identify high-risk individuals, and to assess the QOL of patients at diagnosis and one month after completing the initial treatment. Methodology: Component one: A case-control study was conducted among 49 incident cases of OC recruited from the National Cancer Institute, Maharagama using a non-probability sampling method, and unmatched hospital controls (n=196) excluded of oesophageal carcinoma recruited from the endoscopy unit at the National Hospital of Sri Lanka using an incidence density sampling method. Data were collected using an interviewer administered questionnaire at the patients’ residence. Component two: A risk prediction tool to identify high-risk individuals was developed based on weighted scores derived from their risk factor profile and expert opinion. An independent case-control study was conducted among 30 cases and 138 unmatched controls recruited in the same manner as in component one for assessing its criterion validity against histological diagnosis of OC. Data were collected through an interviewer administered questionnaire. Component three: A cohort of 51 incident cases was followed up to assess their QOL at diagnosis and one month after completion of their initial treatment. Data were collected at the patients’ residence using previously validated EORTC QLQ –C30 and EORTC –OES 18 questionnaires. Results: Bivariate analysis followed by logistic regression analysis identified the following risk factors of OC adjusted for confounding effect: age >65 years (OR: 4.0; 95% CI: 1.1-14.2), family history of cancer (5.0; 1.3-19), sub-optimal consumption of fiber (3.6; 1.1-12.3), antioxidants (7.0; 2.2-22.5), deep fried food (6.7; 2.0-22.6), ‘low’ total lifetime sports and exercise activities (5.8, 1.5-23), ‘high risk’ alcohol consumption (11.7; 2.8-49.4), ever betel quid chewing (6.1; 1.9-20), ever exposure to agrochemicals (6.6; 1.4-30.3), consumption of pipe-borne water (5.6; 1.7-18.9) and ever exposure to radiation (4.6; 1.4-15.5). Risk predictors in the newly developed risk prediction tool were all the significant risk factors in the logistic regression plus ever tobacco smoking. The tool demonstrated valid predictions (96.7% sensitivity; 84.1% specificity; AUC=0.97; 95% CI: 0.94-0.99) to identify high-risk individuals for OC at 34.5 cut-off. On a scale of 0-100, the overall health status/QOL (mean score=49.8; SD=22), and role (42.2; SD=34), physical (53.1; SD=29), emotional (53.4; SD=26) and social functioning (57.2; SD=23) were relatively low at diagnosis, which further deteriorated (difference >5 points) following the initial treatment (p<0.05). Dysphagia (mean=54; SD=27) was the main symptom at diagnosis, which improved significantly (p<0.05) in contrast to dry mouth (mean=39.2; SD=34) that worsened (p<0.05) following initial treatment. Family support and financial difficulties were adversely affected (p<0.05) during the initial treatment. Conclusions and Recommendations: Risk factors specific for OC were mainly lifestyle related including a few explained by environmental factors. To minimize delayed diagnosis, the risk prediction model should be implemented to identify and prioritize high-risk individuals for endoscopy screening. The deterioration of several dimensions of QOL of patients following the initial treatment highlights the need for preserving and improving it in OC management.

Keywords

• oesophageal carcinoma, risk factors, risk prediction, quality of life, prevention, sri lanka