Response to the Letter to the Editor Entitled “From Molecular Signal to Clinical Action in Cardiovascular-Kidney-Metabolic Disease 2”

The author replies:

We thank Wu et al.¹ for their response to our review² and agree with the adoption of their 3-tier hierarchy for biomarker evidence, which maps roughly onto the established triad of analytic validity, clinical validity, and clinical utility.³ Our refinement is that each tier makes a claim about a different object—an assay, a comparison, and a care pathway.

Tier 1, analytic validity, is a claim about the assay. Aptamer and antibody measures of the same protein diverge across ancestry groups, attributed partly to protein-altering variant frequency⁴— the portability concern raised for polygenic risk scores recurs in measurement.

Tier 2, incremental value, belongs not to a biomarker but to a comparison—a named reference model and end point, and a metric, whether discrimination or net benefit. Their chronic kidney disease-versus-estimated glomerular filtration rate-decline contrast¹ changes the end point, and the answer changes with it. Neither answer is a clinical utility claim.

Tier 3, clinical usage, is a claim about the care pathway. A biomarker-strategy trial cannot separate the biomarker from the threshold and action applied to it,⁵ so a positive result shows the pathway works, not which component made it work. A null result does not invalidate Tier 1 or Tier 2 and may say nothing about the pathway either. In GUIDE-IT, the arms barely diverged on NT-proBNP, so the strategy was never tested.⁶

We take Wu et al.’s point about metabolic dysfunction-associated steatotic liver disease. Our review used the diagnosis to modify cardiovascular-kidney-metabolic disease stage 1, letting presence alone carry a Tier-2 claim. Placement should follow the stepwise fibrosis pathway they cite; fibrosis determines how much weight the modifier carries, not whether it applies. Atherogenic lipids raise a separate Tier-2 question.

We agree that their 4 requirements are prerequisites, not refinements. An endotype is a classification, but its clinical claim, that it selects treatment, is a Tier 3 claim. Standardization and reporting are Tier 1-work; transportability, as they cite it, is Tier 2. The fourth, prospective utility testing, is that Tier 3 claim, which no cardiovascular-kidney-metabolic endotype has met. This list still names no incremental-value step: evidence that the endotype predicts outcome beyond cardiovascular-kidney-metabolic staging.

Cardiovascular-kidney-metabolic outcome trials already adjudicate endpoints under standardized follow-up. Serial biospecimen collection would settle Tier 2 against a named model and end point and identify the biomarker and threshold for a Tier-3 strategy trial.

Disclosures

The authors declare no competing interests.

Author Contributions

WRM drafted the manuscript. TM, DG, SS, and PAK reviewed and revised it. All authors read and approved the final version.