Background: Different disease courses within the same high-risk genotype

A score combining plasma proteins and clinical data from people with high-risk APOL1 genotypes showed potential to predict kidney disease progression risk. Chenyu Li and colleagues developed a risk score by adding clinical variables to a pattern of nine proteins. In an independent population-based cohort, they reported better performance than existing risk prediction tools.

In populations of African ancestry, high-risk APOL1 genotypes are strongly associated with kidney failure. Yet disease does not progress in everyone with these genotypes. The study set out to identify those more likely to experience progression within this genetically high-risk group.

The researchers looked to plasma proteins for clues that could support early prediction. A Nature Genetics Research Highlight described the approach as having potential for early risk screening and targeted intervention.

Key findings: APRS combines a protein signature with clinical variables

The researchers analyzed plasma proteins from 1,113 participants with high-risk APOL1 genotypes in the Penn Medicine BioBank. Of these, 851 had preserved kidney function at the time of sample collection, based on estimated glomerular filtration rate (eGFR). The analysis included people with both reduced and preserved kidney function.

Using elastic net Cox regression with cross-validation, the team derived a nine-protein signature. A protein signature is a pattern that combines information from multiple proteins. According to the Research Highlight, this signature predicted kidney disease progression risk independently of other clinical variables.

The researchers then combined the protein signature with clinical variables to create the APOL1 proteomic risk score (APRS). The nine-protein signature itself and APRS, which also includes clinical data, should be understood as distinct measures.

In an independent population-based cohort, APRS predicted progression to kidney disease-related outcomes better than the kidney failure risk equation (KFRE), the Chronic Renal Insufficiency Cohort (CRIC) proteomic score, and a chronic kidney disease polygenic risk score. The comparators included a prediction equation based on clinical data and scores based on proteomic or genetic information.

However, the available Research Highlight did not provide specific numerical measures of predictive performance, the prediction period, or the size of the validation cohort. The magnitude of the improvement and the detailed outcomes assessed cannot be established from this material alone.

Meaning and outlook: From risk screening to clinical use

The findings suggest that progression risk could be further stratified among people with high-risk APOL1 genotypes. Including participants whose kidney function was preserved is also consistent with the study's aim of early risk assessment. However, prediction performance specifically for that subgroup could not be confirmed from the available Research Highlight.

Predictive performance does not by itself establish treatment benefit. Whether giving a particular drug to people with high APRS scores reduces the need for dialysis or the occurrence of kidney failure, or whether changing care based on the score benefits patients, requires separate testing. Before considering practical use, the original study's outcomes and follow-up period, testing methods, performance metrics, and target population need to be examined.

This article is based on a Nature Genetics Research Highlight by Kyle Vogan introducing the original study in Nature Medicine. The full original paper and supplementary materials were not available, so this account focuses on the participants, model, and comparative results confirmed in the highlight. Source: Nature Genetics Research Highlight.