Why PLSD?

About 40 years ago the Leeds Castle Polyposis Group was established with the aim of finding the gene(s) causing familial adenomatous polyposis (FAP). The APC gene was identified. In parallel, interest grew in Inherited colorectal cancers that were not associated with adenomatous polyps and that were denoted as Hereditary Non-Polyposis Colon Cancer (HNPCC). An international collaborative group (ICG-HNPCC) developed clinical criteria to identify and to locate the causative genes for the Lynch syndromes I and II. This identified the four causative genes mismatch repair (MMR) genes, MSH2, MLH1, MSH6 and PMS2, which were demonstrated to cause inherited microsatellite instable (MSI) tumours. Because evidence for an adenoma to carcinoma sequence had been established for sporadic and FAP-associated colorectal cancer, it was assumed that colonoscopy to remove adenomatous polyps would prevent CRC in the Lynch syndromes, and colonoscopy every third year was advocated. CRC, however, continued to occur despite colonoscopy. More frequent colonoscopy was advocated based on the assumption that in the Lynch syndromes, deficient MMR caused transition of adenoma precursors to CRC to occur more quickly - the “accelerated adenoma-carcinoma hypothesis”. However, CRC continued to occur in the carriers despite shorter intervals between colonoscopies.

The European members of the former ICG-HNPCC and others interested in the field continued to meet annually in Mallorca (the Mallorca Group) to undertake collaborative research and develop clinical guidelines for FAP and the Lynch syndromes. By 2012 it was clear that there was something wrong with the hypothesis underlying the advice that colonoscopy would prevent CRC in the Lynch syndromes, but there was insufficient information to understand why. To obtain the information needed, the core members of the Mallorca group agreed to compile their combined empirically observed data and the Prospective Lynch Syndromes Database (PLSD) was born. The initial plan was to make three reports comparing cancer incidences in carriers of pathogenic MMR variants with and without cancer before or at inclusion. Following this, it was decided to further develop PLSD and invite anyone with appropriate data to join. The Mallorca Group was enlarged and incorporated as the European Hereditary Tumour Group (EHTG), taking legal responsibility for the PLSD. As more and more international collaborators joined PLSD the number of cases and follow-up years tripled to the current PLSD version 5, enabling more detailed reports and more precise estimation of cancer risks in the Lynch syndromes.

The challenge in designing the PLSD was a consequence of the four genes having different penetrance and expression, incidences of cancers that vary by age and sex, the occurrence of cancers in different organs and the putative relationships between the cancer incidences. In order to consider whether the assumptions (paradigms) we had believed in were true or false, the selection criteria for inclusion and the algorithm to analyse the data had to include no assumptions. Including such assumptions would inevitably lead to their return as false results. The challenge demanded a complex database and query structure. To obtain meaningful results, sufficient numbers of follow-up years in each age group for each gene for and both sexes were needed. The only way to do this was to establish a state-of-art relational database with no limitations on how to store and analyse the data. In short, the complexity of the database and query system required reflected the complexity of the problem to be addressed, with results that were functions of number of follow-up years in each stratum of interest.

There are four groups of PLSD reports: 1) Descriptions of the theoretical construction of the database and how the inherent epidemiological problems are handled, 2) cumulative incidences of cancers in selected strata and survival after prospective cancer diagnosis, and 3) hypothetical-deductive tests to validate paradigms that have become accepted but that have yet to be proved true. 4) The last group of reports attempt to interpret the combined findings.

The www.plsd.eu calculates cancer incidences as discrete events by the number of follow-up years in each 5-years age cohort for each selected group of carriers of pathogenic MMR variants. In contrast to the published reports where all results reflect calculations starting from 25 years of age, the interactive website may be tuned to start from any age to indicate future cancer risk in any organ in a carrier from his/her current age onwards. The www.plsd.eu is an original free-standing PLSD report displaying results that are not available or calculable from any other public source.