Non-invasive urinary heteroplasmy screening outperforms blood testing for certain mitochondrial DNA variants in...

Mitochondrial diseases present diagnostic challenges due to variations in heteroplasmy levels of mitochondrial DNA (mtDNA) in different tissues. Current diagnostic approaches primarily rely on blood testing, which may miss pathogenic variants present at higher levels in other accessible tissues.

We analyzed tissue samples from 164 individuals (125 probands and 39 family members) with genetically confirmed mitochondrial diseases using droplet digital PCR (ddPCR). We quantified heteroplasmy levels in five tissue types: blood, urine, cardiac muscle, skeletal muscle, and kidney tissue.

Among the 108 blood samples and 87 urine samples from patients carrying the m.3243A>G variant, urine samples demonstrated significantly higher heteroplasmy levels than blood samples (median: 76.6% [interquartile range (IQR): 50.8-91.8%] vs. 20.9% [IQR: 11.3-41.4%], p < 0.001). In a paired analysis of 80 patients with the m.3243 A > G variant who provided both blood and urine samples, urine heteroplasmy exceeded blood levels in all cases (median difference: 44.6% [interquartile range (IQR): 26.9-56.7%], p < 0.001). Moderate positive correlations were observed between blood and urine heteroplasmy (all variants: r = 0.68, n = 94 pairs; m.3243 A > G: r = 0.66, n = 80 pairs; both p < 0.001). Using an exploratory 10% heteroplasmy threshold based on prior reports for m.3243A>G, blood classified 24 of the 94 paired cases (25%, or one in four patients) as <10% while urine was ≥10%. Cardiac tissue exhibited the highest heteroplasmy levels (mean: 86% ±8% in m.3243A>G patients), though this likely reflects selection bias, as cardiac biopsies were obtained only from patients with cardiac involvement.

Urine specimens demonstrate higher heteroplasmy levels and more often place individuals above an exploratory 10% heteroplasmy threshold compared with blood specimens for certain mitochondrial DNA (mtDNA) variants, most notably m.3243A>G. Using this exploratory threshold, 24 of 94 paired blood-urine samples (25%) had blood heteroplasmy <10% while urine heteroplasmy was ≥10%, supporting the use of urine as an important noninvasive complementary specimen, and in many cases a preferred specimen, for suspected m.3243A>G. However, blood testing remains appropriate for variants that typically show high blood heteroplasmy and a relatively high threshold for clinical expressivity, such as m.8993 T > G. These findings support the implementation of urine-based screening protocols for suspected m.3243A>G cases, which may reduce underestimation of variant load and improve diagnostic evaluation.