Association of Plasma Metabolites and Lipids with Kidney Function in Early ADPKD

Autosomal dominant polycystic kidney disease (ADPKD), characterized by gradual enlargement of numerous cysts in the kidneys over decades, affects one in 1000 individuals. The disease process begins before loss of estimated glomerular filtration rate (eGFR) occurs. The most common genetic cause of ADPKD are mutations in polycystin 1 (PKD1, ~75%); the second most common are mutations in polycystin 2 (PKD2, ~15%). A third  causative gene, GANAB, has recently been identified in ADPKD (0.3%) and autosomal dominant polycystic liver disease. When GANAB is mutated and PKD1 maturation is blocked, 5% to 10% of patients have no detectable mutation following DNA sequencing of their PKD1 and PKD2 genes.

Depending on which gene is mutated and the strength of the mutation, the course of ADPKD is variable. Environmental factors such as dietary sodium intake and smoking exposure also contribute to the variability of disease progression and severity. With the recognition that ADPKD is, at least in part, a metabolic disease based on the discovery that glucose, histidine, glutamine, and arginine metabolic pathways are reprogrammed, there is an opportunity to study such pathways to develop new therapies for this disease.

The Consortium for Radiologic Imaging Studies of Polycystic Kidney Disease study established height corrected total kidney volume (ht-TKV) as an accurate and reliable measure of renal cyst burden; it is also a strong predictor of which patients are at risk of progression to chronic kidney disease. The FDA has approved ht-TKV as a prognostic and clinical trial enrichment imaging biomarker for ADPKD. The HALT A clinical trial sought to test the hypothesis that disease progression would be slowed with rigorous blood pressure control (<110/75 mmHg) or combined angiotensin converting enzyme inhibitor and angiotensin receptor blocking therapy. Slowed progression was defined as the rate of increase in ht-TKV and the chronic slope of decline in eGFR.

Utilizing plasma from the baseline visit of participants in the randomized HALT-A clinical trial following cessation of all antihypertensive medications for at least 2 weeks, Kyoungmi Kim, PhD, and colleagues recently sought to correlate expression of specific metabolites with simultaneously measured eGFR and ht-TKV. Results were reported online in BMC Nephrology

Data from a HALT-A subset of 277 white participants who had known mutations in either the PKD1 or PKD2 genes were analyzed. To identify individual metabolites whose intensities are significantly correlated with eGFR and ht-TKV, association analyses were performed using linear regression with each metabolite signal level as the primary predictor variable and baseline eGFR and ht-TKV as the continuous outcomes of interest, while adjusting for covariates. Storey’s false discovery rate (FDR) q-values to correct for multiple testing were used to determine significance.

To assess for systematic shifts among the samples, the researchers investigated reproducibility and time-dependent variation using the repeated measurements for the 35 technical replicates for metabolomics and the 36 technical replicates for lipidomics.

Following accounting for covariates, 20 metabolites were significantly associated with eGFR at a P value of .05. Of those, there was significant correlation between 12 metabolites and eGFR at FDR q-value <.05. All of the significant metabolites exhibited negative associations with eGFR; the metabolite levels were increased with decreasing eGFR levels.

Significant associations were found between 10 metabolites and ht-TVK at a P value of .05; none retained their significance at FDR q-value <.05. Four metabolites (taurine, tyrosine, glutamic acid, and 2-hydroxybutanoic acid) were negatively correlated with ht-TKV; the others exhibited a positive correlation.

Analysis of the lipidome revealed significant associations between 11 lipids and eGFR at P<.05; no lipids remained significant at FDR q-value <.05. Associations were negative between acylcarnitines and triglycerides and eGFR and positive between phosphatidylcholines and diglyceride. Analysis of the relationship between the lipidome and ht-TVK revealed a significant association between five triglycerides and one sphingomyelin and ht-TKV; only two triglycerides remained significant at FDR q-value <.05. All six of these lipids showed positive associations with ht-TKV.

The researchers performed further analyses using the metabolites and lipids association with eGFR and ht-TVK as covariates to identify the metabolites associated only with eGFR or ht-TVK exclusively independent of each other. When controlled for ht-TVK as a covariate, four metabolites (pseudo-uridine, indole-3-lactqate, creatinine, and uric acid) were significantly associated with eGFR at FDR q-value <.05. The four metabolites were among the 12 previously identified as significant at FDR q-value <.05 without controlling for ht-TKV. In analyses of ht-TVK while controlling for eGFR as a covariate, only taurine was significant at P<.05; significance was lost at FDR q-value <.05.

The researchers cited some limitations to using non-targeted metabolomics to identify novel small molecular biomarkers that may have therapeutic impact, including the lack of structural information on some of the metabolites identified. In this study, one such metabolite (191801) was found to be highly correlated with eGFR even after controlling for ht-TVK as a covariate.

In conclusion, the researchers said, “This study identifies metabolic derangements in early ADPKD which may be prognostic for ADPKD progression. Such data will be useful for future studies designed to predict outcome of disease based on such early metabolic changes and will likely lead to new therapeutic paradigms for a disease with quite limited therapeutic options.”

Takeaway Points

  1. Researchers tested the hypothesis that specific plasma metabolites would correlate with known predictors of disease progression (estimated glomerular filtration rate [eGFR] and height corrected total kidney volume [ht-TVK] ) in patients with autosomal dominant polycystic kidney disease (ADPKD) in early stages of the disease.
  2. Twelve metabolites were significantly correlated with eGFR and two triglycerides significantly correlated with baseline ht-TKV.
  3. Identification of metabolic derangements in early ADPKD may aid in prediction of disease outcome and ultimately lead to new therapeutic paradigms for this patient population.