Pancreatic Tumors: a Multicentre Study of Multidetector Computed Tomography Findings and Histopathologic Correlation in Nairobi, Kenya.

Abstract

Background: Pancreatic tumours have variable morbidities and mortalities dependent on the histological subtype, tumour grade and stage. These influence management strategies. Triple phase Multi-Detector Computed Tomography (MDCT) imaging has a role in tumour detection, staging, prognostication and treatment response assessment. There are limited studies on pancreatic tumour imaging characteristics and histologic types in Kenya and in Africa as a whole. Knowledge of MDCT imaging characteristics of different tumour types encountered in Kenya will impact on patient management strategies. Objective: To analyse the MDCT imaging spectrum and histopathologic correlation of patients with pancreatic tumours at Kenyatta National Hospital, Plaza Imaging Solutions and German Medical Centre. Methodology: This was a cross sectional study done at the radiology departments of KNH, Plaza Imaging Solutions and German Medical Centre in Nairobi, Kenya. Thirty-nine consenting patients found to have pancreatic tumours on triple phase pancreatic protocol MDCT imaging and who obtained histopathology results after tumour tissue sampling were recruited. A structured data collection tool was used to document the demographic data and MDCT pancreatic tumour imaging characteristics of the study participants. The most likely tumour type from evaluation of the MDCT imaging characteristics was documented as well as the histopathological diagnosis. Data analysis was done using the Statistical Package for Social Scientists software (version 25). The results were presented in tables, pie carts, and bar charts. Results: A total of 39 participants were recruited into the study. The mean age of patients with pancreatic tumours was 58.4±13.9 years with 59% of them being female. Pancreatic ductal adenocarcinoma was the most prevalent tumour subtype at 87.2%. Pancreatic neuroendocrine tumour, solid pseudopapillary neoplasm and paraganglioma constituted 2.6% each while other pancreatic tumour subtypes were not represented. PDAs were mostly poorly circumscribed (91.2%), solid (85.3%), and located in the pancreatic head (55.9%). All the PDAs were hypovascular in the arterial and portovenous phases. Main pancreatic duct dilatation and distal pancreatic atrophy was seen in 61.8% and 58.8% of PDAs respectively. The double duct sign was only seen in 29.4%. Distant metastases were found in 47.1% of the tumours at presentation. The diagnostic accuracy of MDCT imaging for pancreatic ductal carcinoma was 94.9%. Sensitivity and specificity were 100% and 60% respectively, while the PPV and NPV were 94.4% and 100%.

Conclusion: This study has shown that pancreatic ductal adenocarcinoma is the most prevalent of the pancreatic tumour subtypes in Kenya. MDCT imaging using the pancreatic protocol had high accuracy consistent with studies conducted elsewhere and is therefore reliable in the diagnosis and exclusion of pancreatic ductal adenocarcinoma. Recommendation: We recommend larger multicentre studies to evaluate the less commonly occurring pancreatic tumour subtypes. Retrospective studies focusing on the less common subtypes can be done to determine MDCT imaging features and diagnostic accuracy.