S5 - E17.4 - Previewing EASL Congress 2024: Is Big Data The Future Of MASLD?

Surfing the MASH Tsunami

This conversation revolves largely around two Big Data presentations Hannes Hagstrom's team is sharing at the EASL Congress. Every other panelist has at least one meaningful comment to make about the future of Big Data in MASLD.
Hannes describes two big data studies of his own, "very much Big Data, but it's not omics." One examined 860,000 residents of Stockholm who took a liver test between 2010-2022 seeking to estimate liver and cardiovascular risk. It relied on a tool from Stefano Romeo's group called the Fibrotic Mass Index. The results demonstrated the prediction of liver-related and cardiovascular-related outcomes.

The other study relied on the same dataset to build a prediction model for liver-related outcomes in the general population. Instead of relying on a dedicated test like FIB-4, which has false results on both sides, this assimilates five simple parameters (ALT, AST, GGT, ender, age) and proves to be far more accurate at risk assessment than FIB-4. The study team will also present a small, web-based calculator "where people can go to estimate the risk." The tricky part, Hannes notes, is where to set a tolerable risk threshold.

Jörn, thinking about the 860,000 sample, asks, "How much data is big enough?" Hannes notes that if the sample is big enough, we need not worry about statistical tests because we have true population-level data.
Michelle considers the overarching implications of Hannes's comments: "I think a risk-based view of this disease makes a lot of sense." Aleksander and Roger concur that using a risk-based screening approach is preferable, in large part because it should allow a far larger share of resources to be devoted to those with significant disease.