Search strategies to obtain evidence for cancer classification

The accuracy of the WHO classification of tumours (WCT), and hence cancer diagnosis, depends on rigorous assessment of the scientific evidence available which describes the defining characteristics of each tumour type. The scale of the task is considerable, given the vast number of publications in the field. The WCT Evidence Gap Map project (WCT EVI MAP; https://wct-evi-map.iarc.who.int/) is designed to answer this need, using Evidence and Gap Map (EGM) methodology to inform the WCT, which is well suited to large datasets.

As examples to test the strategy, we chose sebaceous carcinoma and cholangiocarcinoma. As a rare tumour, sebaceous carcinoma has a relatively small literature, while the landscape for cholangiocarcinoma is more complex due to recent changes in the classification, the use of outdated terms, and many publications.

The search strategy uses two components - the first describing the tumour type, including alternative terminology, and the second the tumour characteristics (aetiology, pathogenesis, clinical features, epidemiology, histopathology, etc.), with synonyms, as used in the WCT.

The results emphasise how relatively small differences in the exact terms and fields searched can produce big differences in the results obtained. It is often important to use a wide range of alternative terms and to cover multiple variations of phrases (via techniques such as proximity searching) to ensure comprehensive results. Limiting results to those characteristics used to classify tumours and using date thresholds helps to improve the relevance of the literature identified.

The World Health Organization (WHO) classifies different types of cancers based on reviews of the scientific literature. Their WHO Classification of Tumours reference book series helps medical professionals make accurate diagnoses. However, reviewing all the available research comprehensively is an enormous task, given the number of studies. To help with this, the WHO launched the WCT Evidence Gap Map (WCT EVI MAP) project. A method called "Evidence and Gap Mapping" is used to organize and assess large amounts of research more efficiently. This method helps identify what is already known and where more research is needed. To test the approach, the researchers modelled their process by looking at two types of cancer: sebaceous carcinoma (a rare cancer with fewer studies) and cholangiocarcinoma (a more complex cancer with many studies and changing terminology). They created search strategies that combined a range of similar terms for the tumour types, along with terms describing their key characteristics, such as how these cancers develop, how they appear under the microscope, and how common they are. They found that small variations in search terms can lead to very different results, so it's important to use a well-worked out search technique using a wide variety of terms. At the same time, the researchers also found that focusing on the most relevant tumour features and using date filters made the search results more useful.