Authors’ Reply: Dialysis Dose and Flux as Potential Effect Modifiers of Sphingolipid–Mortality Associations

We appreciate the thoughtful Letter to the Editor from Zhao et al.1 and thank them for recognizing the strengths of our JASN study.2 We appreciate their question regarding whether the associations between sphingolipids and mortality remained consistent across treatment arms (dialysis dose and membrane flux). We agree that the design of the hemodialysis (HEMO) trial offers a unique opportunity to investigate for effect modification by intervention.3 To address this question, we performed additional analyses testing the intervention arms (dialysis dose [low, Kt/V goal of 1.05 versus high, Kt/V goal 1.45] and dialyzer flux [low versus high]) as potential effect modifiers of the association between sphingolipids and mortality. After correction for multiple hypothesis testing across 16 sphingolipids and two potential effect modifiers (as described in the main paper),2 neither dose nor flux significantly modified the association between sphingolipids and all-cause (corrected P values > 0.29) or cardiovascular (corrected P values > 0.05) death. We additionally found no significant effect modification of the association between sphingolipid subclasses (long-chain ceramides, the sum of ceramides 16:0 and 18:0; long-chain sphingomyelins, the sum of sphingomyelins 14:0, 16:0, and 18:0; and very long ceramides and sphingomyelins, the sums of 20:0, 22:0, and 24:0) and all-cause (corrected P values > 0.31) or cardiovascular (corrected P values > 0.09) death by dialysis dose or dialyzer flux. In summary, we did not appreciate significant effect modification by dialysis dose or dialyzer flux of the association between sphingolipids and all-cause and cardiovascular mortality. It is important to note that current dialysis practice does not perfectly mirror the intervention of the HEMO trial; high-flux dialyzers are now standard of care, with goal single-pool Kt/V targets typically set around 1.4 per session, very close to the high-dose arm in HEMO.4 Similarly, our present analyses may be inadequate to fully address the question at hand—we performed a post hoc analysis of a subset of the HEMO trial, whose primary outcomes were not sphingolipids. Although dialysis is unlikely to directly clear sphingolipids from circulation given their transport in lipoproteins, future studies are needed to understand the effects of dialysis on sphingolipids.5 In conclusion, we are grateful for this thoughtful engagement with our work, as well as the suggestion of additional analyses that support the robustness of our work noting associations between sphingolipids and all-cause and cardiovascular mortality in patients with kidney failure treated with maintenance hemodialysis.