Association of Blood Mitochondrial DNA Copy Number With Risk of Acute Kidney Injury After Cardiac Surgery

To the editor:

Acute kidney injury (AKI) is a common complication of cardiac surgery and its occurrence is associated with increased risk of all-cause mortality, adverse cardiovascular outcomes, and chronic kidney disease (CKD).¹ Mitochondria are essential for healthy function of the kidneys due to their roles in oxidative phosphorylation, reactive oxygen species generation, and programmed cell death.² Experimental evidence suggests that mitochondrial dysfunction is a key contributor to the pathophysiology of acute kidney injury (AKI) and that replenishment of mitochondrial stores is necessary for recovery from AKI,^3–5 but limited data exist in humans. Mitochondrial DNA copy number (mtDNA-CN) is an indirect marker of mitochondrial abundance that quantifies the number of mitochondrial genomes per cell. Recent studies have reported that higher blood mtDNA-CN is associated with reduced risks of chronic kidney disease (CKD), cardiovascular disease (CVD), and mortality,^6–9 but no prior study has investigated associations with AKI. We hypothesized that mitochondrial quantity, assessed by mtDNA-CN, might indicate resilience from acute ischemic stress.

This study evaluated associations of preoperative blood mtDNA-CN with risk of AKI after cardiac surgery and with long-term risk of CKD, CVD, and mortality among 628 adult participants of the Translational Research Investigating Biomarker Endpoints in AKI (TRIBE-AKI) Study.¹⁰ Detailed Methods are provided in Item S1. Briefly, mtDNA-CN was measured in preoperative buffy coat specimens by a monochrome multiplex real-time quantitative PCR assay and was standardized to nuclear DNA. The primary outcome was AKI defined as the receipt of renal replacement therapy or an increase of ≥50% in serum creatinine within the hospitalization for cardiac surgery. Secondary outcomes included: elevations in urine biomarkers of tubular damage (interleukin-18, IL-18; kidney injury molecule-1, KIM-1; monocyte chemoattractant protein-1, MCP-1; and chitinase-3-like protein 1, YKL-40) to their highest quintile on the first postoperative day; CKD incidence or progression; CVD events; and, death. Multivariable logistic regression models were used to evaluate mtDNA-CN associations with AKI risk and with the highest quintile of each biomarker outcome. Cause-specific proportional hazards models were used to examine associations of mtDNA-CN with CKD incidence or progression, CVD events, and mortality.

The mean participant age was 73±9 years, and the mean preoperative eGFR was 72±18 ml/min/1.73m² (Table S1). Approximately one-third of participants were women, and diabetes mellitus and hypertension were prevalent in 41% and 85% of participants, respectively. When participants were categorized by mtDNA-CN tertile, there were no differences in age, sex, presence of diabetes mellitus or hypertension, or history of CVD. However, those in the highest vs. lowest tertile had lower preoperative serum creatinine and higher eGFR.

Postoperative AKI occurred in 70 (11%) participants, with 7 (1%) individuals requiring renal replacement therapy. Compared to participants in the lowest mtDNA-CN tertile (Table S2), those in the highest tertile had a lower incidence of AKI (6% vs 16%; p=0.008), lower risk of sustained AKI of two days or longer (2% vs 8%; p=0.034), and shorter length of ICU stay (2.2 vs 3.0 days; p<0.001). In multivariable-adjusted models, each standard deviation higher mtDNA-CN was associated with an approximately 38% lower risk of postoperative AKI (Table 1), and those in the highest vs. lowest tertile had an approximately 66% lower AKI risk.

Table 1:

Association of preoperative mtDNA copy number with risk of AKI¹ after cardiac surgery among TRIBE-AKI participants (N=628)

	Overall	Tertile 1	Tertile 2	Tertile 3
No. at risk	628	209	210	209
No. AKI events	70	33	24	13
Unadjusted, OR (Q1, Q3)	0.64 (0.49, 0.84)	Ref	0.69 (0.39, 1.21)	0.35 (0.18, 0.69)
Multivariable Adjusted3, OR (Q1, Q3)	0.62 (0.47, 0.82)	---	0.67 (0.37, 1.21)	0.34 (0.17, 0.69)

Odds ratio are 95% CI.

¹AKI defined as >50% increase in serum creatinine or receipt of dialysis

²Odds ratio per standard deviation of mtDNA copy number

³Multivariable model adjusts for age, sex, pre-operative serum creatinine, diabetes mellitus, heart failure, and type of surgery.

Abbreviations: AKI, acute kidney injury

There were no meaningful correlations of mtDNA-CN with postoperative urine IL-18 (r=0.07; p=0.10), KIM-1 (r=−0.02; p=0.66), MCP-1 (r=−0.005; p=0.91), or YKL-40 (r=0.09; p=0.04), nor were there significant associations with the highest quintile for any of the biomarkers (Table S3). There were no statistically significant associations of mtDNA-CN with risks of CKD incidence or progression, CVD, or mortality (Table 2).

Table 2:

Association of preoperative mtDNA copy number with risk of CKD incidence or progression, CVD events, and mortality among TRIBE-AKI participants¹

	Overall	Tertile 1	Tertile 2	Tertile 3
CKD incidence or progression (N=597)
Event rate per 1000 person-years	84.0	94.0	92.4	67.8
Unadjusted, HR (Q1, Q3)	0.88 (0.75, 1.03)	Ref	0.99 (0.69, 1.43)	0.73 (0.50, 1.08)
Multivariable Adjusted3, HR (Q1, Q3)	0.91 (0.78, 1.08)	---	1.05 (0.72, 1.54)	0.78 (0.52, 1.17)
CVD events (N=592)
Event rate per 1000 person-years	36.0	37.1	32.7	38.3
Unadjusted, HR (Q1, Q3)	0.90 (0.73, 1.11)	Ref	0.88 (0.53, 1.46)	1.03 (0.63, 1.68)
Multivariable Adjusted3, HR (Q1, Q3)	0.96 (0.78, 1.19)	---	1.04 (0.62, 1.75)	1.20 (0.72, 1.98)
Mortality (N=592)
Event rate per 1000 person-years	36.2	42.3	38.5	28.3
Unadjusted, HR (Q1, Q3)	0.82 (0.67, 1.01)	Ref	0.92 (0.58, 1.45)	0.68 (0.41, 1.12)
Multivariable Adjusted3, HR (Q1, Q3)	0.89 (0.72, 1.09)	---	1.04 (0.65, 1.67)	0.83 (0.50, 1.40)

HR are 95% CI

¹Median follow up of 3.5 years (25^th percentile, 75^th percentile; 2.1, 4.6 years)

²Hazard ratio per standard deviation of mtDNA copy number. Cause-specific hazards models were used for the outcomes of CKD incidence or progression and CVD events, as the outcomes were censored at the competing event of death.

³Multivariable model adjusts for age, sex, pre-operative serum creatinine, diabetes mellitus, heart failure and type of surgery.

Abbreviations: CI, confidence interval; CKD, chronic kidney disease; CVD, cardiovascular disease; HR, hazard ratio.

In this cohort of adults undergoing cardiac surgery, we found that higher preoperative mtDNA-CN was associated with reduced risk of postoperative AKI. As a metabolically active organ, the kidney is fueled by ATP generated by the mitochondrial respiratory chain, and mitochondrial damage is an early feature of ischemic, septic, and toxin-associated AKI.³ In murine models, knockout of mitochondrial biogenesis leads to increased severity and duration of AKI,⁴ whereas its induction confers tolerance to renal ischemia.⁵ Our finding builds upon this experimental evidence that suggests mitochondrial abundance is a key determinant of kidney reserve and resistance to stress.

The low number of events is a limitation of our study, and external validation is warranted. Although the associations of mtDNA-CN with the composite CKD outcome, CVD, and mortality risk did not reach statistical significance, the directionality and point estimates are consistent with prior studies, which were larger and had a longer duration of follow-up.^6–9 Additional limitations include the absence of data on cell composition and albuminuria for adjustment in multivariable models and the lack of diversity within the study population.

In summary, a higher preoperative mtDNA copy number was associated with a lower risk of postoperative AKI among adults undergoing cardiac surgery. Future studies are needed to replicate this finding and to explore the diagnostic or therapeutic implications for individuals at risk for AKI.