Risk Factors Associated With Early-Onset Esophageal Cancer in Tanzania

Eastern Africa is one of several regions affected by high incidence rates of esophageal squamous cell carcinoma (ESCC). A unique epidemiologic feature of ESCC in Eastern Africa is the high incidence in young people, with one-third of cases diagnosed at age < 45 years. This study aimed to investigate risk factors for early-onset ESCC in Tanzania through a secondary analysis of a matched case-control study.

Data from Tanzania show esophageal cancer risk factors in East Africa may differ across age groups.

A Case-Control Study to Evaluate the Etiology of Esophageal Cancer in Tanzania

Abstract 75

Objective:

East Africa has been previously reported as a geographic area with a disproportionately high incidence of squamous cell esophageal cancer (EC). In the setting of an apparent geographic cluster, we sought to identify specific etiologic factors contributing to the high incidence of EC in Tanzania (TZ).

Methods:

We performed a case-control study to examine possible associations between EC and selected environmental, dietary, and lifestyle factors. Newly diagnosed EC cases were recruited at Muhimbili National Hospital and Ocean Road Cancer Institute during 2014-2015. Hospital controls were identified from patients with non-malignant conditions and matched 1:1 for gender and age ±10 years. The study team conducted in-person interviews. A matched case-control analysis used McNemar’s test with a Mantel-Haenszel estimate for odds ratios (OR).

Results:

375 cases and 375 controls were enrolled. 71% of cases and controls were male. Median age was 58 years (range 30-91) for cases and 55 years (range 31-88) for controls. Among cases with a confirmed pathologic diagnosis, squamous cell carcinoma was the most common histology (85%). Liquor consumption (OR 1.84, 95% CI: 1.05-3.22, p=0.04), burning firewood (OR 2.66, 95% CI: 1.88-3.76, p<0.001), use of kerosene (OR 3.29, 95% CI: 2.36-4.59, p<0.001), and use of surface water as a primary source of drinking water (OR 2.3, 95% CI: 1.41-3.76, p=0.001) were associated with increased risk of EC. Running water, gas, and electric utilities in the household were protective.

Conclusion:

Alcohol consumption, firewood and kerosene use as fuel, and surface water consumption merit further investigation as potential etiologic factors contributing to high EC incidence in TZ. Future research will include conducting multivariate analysis with these and other potential risk factors and a genome-wide analysis study using DNA extracted from saliva specimens obtained from cases and controls. Funding: Research support provided by the National Cancer Institute, National Institutes of Health Contract No. HH5N261200800001E.

AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST:

Elia Mmbaga No relationship to disclose Katrina Deardorff No relationship to disclose Beatrice Mushi No relationship to disclose Li Zhang No relationship to disclose William Mgisha No relationship to disclose Aparna Parikh Employment: Genentech Stock or Other Ownership: Genentech Travel, Accommodations, Expenses: Genentech Robert A. Hiatt No relationship to disclose Julius Mwaiselage No relationship to disclose Katherine Van Loon Consulting or Advisory Role: Bayer Company: OncoPlex Diagnostics

Factors Associated With Early-Onset Esophageal Cancer in Tanzania: A Case-Control Study

Abstract 89

Purpose

Previous studies have characterized geographic clusters of esophageal cancer (EC) in East Africa. Many of the epidemiologic features of EC in this context are shared globally with other clusters, including high rates, male predominance, and squamous cell histology. A unique feature in East Africa is the high proportion of young patients, with a recent case series reporting up to 24% of patients age < 45 years. The aim of the current study was to assess factors that are associated with early-onset EC in Tanzania (TZ).

Methods

We performed a secondary analysis of a previous case-control study. Patients with newly diagnosed EC were recruited at Muhimbili National Hospital and Ocean Road Cancer Institute in 2014 to 2016. Hospital controls were identified from patients with nonmalignant conditions and matched 1:1 for gender and age ± 10 years. Risk factors were assessed through interviews. Logistic regression was used to estimate age-specific odds ratios (ORs) of EC for exposures across age groups (30 to 44 years, 45 to 59 years, and ≥ 60 years) and for interactions with age.

Results

A total of 473 cases and 473 controls were enrolled. Median ages were 59 years (range, 30 to 91 years) for cases and 55 years (range, 31 to 88 years) for controls. Among cases, 102 patients (22%) were age 30 to 44 years, 144 patients (30%) were age 45 to 59 years, and 227 patients (48%) were age ≥ 60 years. High household income was protective for those age 30 to 44 years (OR, 0.08; 95% CI, 0.01 to 0.69) and 45 to 59 years (OR, 0.13; 95% CI, 0.04 to 0.45), but not for those age ≥ 60 years (effect modification P = .047). Family history of EC was associated with a higher risk of EC among those age 45 to 59 years (OR, 3.8; 95% CI, 1.02 to 14.47) and age ≥ 60 years (OR, 6.63; 95% CI, 1.50 to 29.37), with no effect among those age 30 to 44 years (effect modification P = .019). Second-hand smoke and infrequent teeth cleaning were also associated with early-onset EC, but did not differ significantly across age groups. Additional factors associated with EC risk across all ages were firewood use (cooking), kerosene use (lighting), work on a maize farm, and in-home grain and nut storage. Protective factors were the regular use of medication, surrogates of high socioeconomic status (TV, radio, refrigerator, indoor toilet, and electricity), and charcoal or gas cooking.

Conclusion

Multiple exposures were identified as risk factors for early-onset EC in TZ. In age-stratified analyses, household income, second-hand smoke, and poor dental hygiene emerged as possible risk factors, whereas family history of EC had strong associations among the older but not the young age group. Our results suggest that environmental factors may underlie the high incidence of young patients with EC in TZ.

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

No COIs from the authors.

Predictors of Radiation Therapy Incompletion Among Patients With Esophageal Cancer in Tanzania

Abstract 96

Purpose

The incidence of esophageal cancer in East Africa is disproportionately high. In Tanzania, radiation therapy (RT) is routinely offered for definitive and palliative therapy; however, many patients do not complete RT or die shortly thereafter. The current study aimed to characterize RT treatment patterns in Tanzania and identify predictive factors for RT incompletion.

Methods

We performed a retrospective chart abstraction for patients with esophageal cancer who were treated with RT at a national referral cancer center in Tanzania from 2011 to 2013. Definitive intent was defined as RT prescriptions with at least 20 fractions with concurrent chemotherapy. Other fractionation regimens were considered palliative. Wilcoxon rank-sum tests, χ2 tests, and logistic regression models were used to identify factors that are associated with palliative or definitive RT incompletion.

Results

A total of 300 patients—202 male and 98 female patients—were identified with a median age of 60 years (interquartile range [IQR], 48 to 70 years). Nearly 100% (299 of 300) of patients reported dysphagia to solids, and 54% (155 of 288) reported dysphagia to liquids. Median duration of symptoms before presentation was 4 months (IQR, 2 to 6 months), and median time from diagnosis to RT was 1.2 months (IQR, 0.8 to 1.9 months). Overall, 23% were unable to complete RT as a result of death or clinical decompensation. Palliative treatment was administered to 149 patients, and 26% did not complete RT. Definitive treatment was administered to 151 patients, and 20% did not complete RT ( P = .24). Patients younger than age 60 years were less likely to complete palliative RT (odds ratio [OR], 2.4; P = .02). Tobacco use (OR, 2.7; P = .04) and RT initiation within 30 days of diagnosis (OR, 3.5; P = .004) were associated with incomplete definitive RT.

Conclusion

In Tanzania, approximately 23% of patients die or decompensate before completing esophageal RT. Patients younger than age 60 years were less likely to complete palliative RT. Tobacco use and RT initiation within 30 days of diagnosis was associated with definitive treatment incompletion, perhaps reflecting differences in cancer biology or pace of disease. Additional understanding of how these factors contribute to RT incompletion may inform supportive care resource allocation and patient selection for esophageal RT in Tanzania and similar resource-limited settings.

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

No COIs were provided by the authors.

Traditional and Commercial Alcohols as Risk Factors for Esophageal Cancer in Tanzania

PURPOSE

Esophageal cancer (EC) is one of the most common cancers in East Africa; however, risk factors that underly the high incidence in this region are not well understood. We aimed to investigate the association between exposure to specific alcohol subtypes and EC in Tanzania.

METHODS

We performed a secondary analysis of data from a hospital-based, case-control study conducted at Muhimbili National Hospital and Ocean Road Cancer Institute in Dar es Salaam, Tanzania. Cases of EC were identified between 2014 and 2016. Hospital controls were identified from patients with nonmalignant conditions and matched 1:1 for gender and age (± 10 years). Interviews were conducted using a survey with culturally relevant and context-specific questions on alcohol and smoking exposure. Conditional logistic regression analyses were applied to evaluate specific associations of potential exposures and EC.

RESULTS

A total of 473 EC cases and 473 controls were enrolled. Alcohol consumption was reported in 61% and 64% of cases and controls, respectively. Neither ‘current use of alcohol (any type)’, nor ‘ever use’ were associated with EC; however, local brew consumption was associated with increased EC risk (odds ratio [OR], 2.01; 95% CI, 1.53 to 2.66; P < .01). Increased risk was observed with consumption of gongo (OR, 2.91), komoni (OR, 2.41), wanzuki (OR, 2.40), kindi (OR, 3.13), and kangara (OR, 2.86). Smoking increased EC risk; however, it did not significantly modify the association between EC and alcohol subtypes.

CONCLUSION

This is the first case-control study to report a detailed analysis of alcohol exposures as a potential risk factor for EC in Tanzania. Although combined measures of alcohol use were not found to be associated with EC, several types of locally brewed alcohols emerged as risk factors. Additional research is needed to investigate these findings and examine the carcinogenic role of ingredients and/or contaminants, as well as any interactions with other putative risk factors in this region.