Pancreatic Cancer: Patterns in a Low- to Middle-Income Population, Zambia

  • Akwi W Asombang Division of Gastroenterology and Hepatology, University of Missouri-Columbia School of Medicine, Missouri
  • R Madsen Department of Statistics, University of Missouri-Columbia School of Medicine, Missouri 65203 USA
  • M Simuyandi Center of Infectious Disease Research in Zambia (CIDRZ), Lusaka, Zambia
  • G Phiri Cancer Disease Hospital (CDH), Lusaka, Zambia
  • M Bechtold Division of Gastroenterology and Hepatology, University of Missouri-Columbia School of Medicine, Missouri
  • J A Ibdah Division of Gastroenterology and Hepatology, University of Missouri-Columbia School of Medicine, Missouri
  • K Lishimpi Cancer Disease Hospital (CDH), Lusaka, Zambia
  • L Banda Cancer Disease Hospital (CDH), Lusaka, Zambia


In 2007, the Cancer Disease Hospital (CDH) was opened as the national referral center for patients diagnosed with cancer in Zambia.   Since inception of the CDH, there has been no systematic analysis of the disease burden and implication on healthcare delivery with regards to pancreatic cancer.   There are limited studies describing patterns of pancreatic cancer in a native African population. Data suggest African-Americans have a higher incidence and poorer prognosis of pancreatic cancer than non-African Americans. 


Our aim is to describe the demographic features (age, gender) of pancreatic cancer using the Cancer Disease Hospital (CDH) data base in a native African population and compare with the African-American cohort using the Surveillance, Epidemiology End Results (SEER) Program database.


This was a retrospective study of patients diagnosed with pancreatic cancer at the CDH in Zambia, Southern Africa between 2007 and 2014.  We entered the term “pancreatic cancer” into the CDH database, extracted patient medical records numbers, and manually located the records for review.  From each chart we extracted: age, gender, geographic origin, ethnicity, clinical features at presentations, location of tumor, stage at diagnosis and treatment. Data collection tool and master code sheet created a priori were used. Data was analyzed using statistical analysis software (SAS).  Descriptive statistics including means, medians as well as frequency distributions and cross-tabulations were used.    SEER database was used to compare subjects between Zambia and USA. Groups were compared using Chi-square tests and Wilcoxon Signed Rank test.  A p-value less than 0.05 was used as the level of significance.

Thirty-eight charts were identified in the CDH dataset, of which 27 were included in final analysis and 11 excluded (5 non-pancreatic cancer diagnosis, 6 not manually located). The mean age of diagnosis was 55.7 years in the native African population, compared to 66.7 years for the African-Americans in the SEER database, p < 0.0001.  There were 63.0% males (CDH) compared to 48.1% (SEER), p=0.121.  Further review of the CDH database revealed that the most common presenting symptom was abdominal pain (52.6%), mode of diagnosis surgical (83%, missing 3), histopathology adenocarcinoma (86%, missing 6), location head of pancreas (83%, missing 9) and stage 4 at diagnosis (100%, missing 3). 

Pancreatic cancer occurs at a younger age in Zambians when compared to the African American, USA population. There is no statistically significant difference in sex presentation between Zambian and USA black population.


Download data is not yet available.