Dr Kapur began by highlighting the impact of kidney injury on long-term outcomes in patients after acute MI (AMI). The study by Parikh et al. in 2008 showed that the long-term implications of mild acute kidney injury (AKI) in the setting of AMI are striking, with only about 20% of patients surviving to 10 years.1 Also, severe AKI was associated with high mortality as early as 30 days. Another study by Goldberg et al. showed that even transient AKI after AMI is associated with a high probability of death in the long term.2 The recent study by Chalikias et al. further emphasised the significantly higher mortality within 30 days in patients developing AKI after AMI than in those without AKI, which continued for more than 10 years.3
The underlying pathophysiological mechanisms in AKI is best validated in the heart failure model. The main determinants of decreased estimated glomerular filtration rate (eGFR) are a decrease in renal blood flow (RBF) and an increase in central and renal venous pressure.4 The latter can be caused by intravascular congestion. Owing to increased renal venous pressure, renal interstitial pressure rises, resulting in the collapse of renal tubules, thus decreasing the eGFR. This eventually leads to decreased urine output, sodium retention and kidney congestion. Decreased RBF and low blood pressure trigger renal autoregulation, preserving the glomerular filtration rate by increasing the filtration fraction by increased efferent vasoconstriction. The use of certain drugs in heart failure may inhibit this efferent vasoconstriction, which leads to an increase in RBF, but may also result in a reduction in the eGFR (pseudo-worsening renal function).
Since several haemodynamic parameters influence kidney function, Dr Kapur’s research team is interested in investigating the effect of acute mechanical circulatory support devices, such as Impella and veno-arterial extracorporeal membrane oxygenation (VA-ECMO) on renal blood flow and function. The study by Abadeer et al. showed that there is a 60% incidence of AKI in patients receiving VA-ECMO for cardiogenic shock, which also correlated with poor survival.5 Patients who had severe AKI after initiation of VA-ECMO had lower survival than those with non-severe AKI. Villa et al. postulated that AKI with VA-ECMO might be related to haemodynamic, hormonal and patient-specific variables, leading to a reduction in renal oxygen delivery and/or inflammatory damage.6
A recent study by Flaherty et al. was the first to demonstrate renal protection by Impella 2.5 compared to no support in high-risk percutaneous coronary intervention patients.7 Impella 2.5 support was independently associated with a significant reduction in the risk of developing AKI during high-risk percutaneous coronary intervention. Furthermore, unpublished results from Dr Westenfeld’s research group shows that peri-interventional serum creatinine levels are stable only with Impella compared to VA-ECMO. Previously, Møller-Helgestad et al. compared the RBF with intra-aortic balloon pump (IABP) versus Impella 2.5.8 The results suggest a signal of higher RBF with Impella 2.5 and no change with IABP support.
To understand haemodynamics in the kidney, researchers at Dr Kapur’s laboratory performed pressure and flow measurements in the renal artery, renal vein and inside the renal parenchyma in the large animal, closed-abdomen experimental setup. The renal module setup was validated in healthy adult pigs treated with sham, Impella or VA-ECMO, followed by the measurement of cardiac and kidney haemodynamics over 6 hours. Preliminary results suggest that in the sham control animals, cardiac output and stroke volume (SV) were maintained over 6 hours, whereas the systemic vascular resistance (SVR) gradually declined. Treatment with Impella almost paralleled sham control animals. In contrast, the cardiac output and SV of the native heart dropped following initiation of VA-ECMO, along with a significant drop in SVR within 1 hour, which persisted throughout the treatment duration.
Dr Kapur introduced the concept that the reduction in kidney function may also be related to the role of the kidney as a sensor of damage- and pathogen-associated molecular patterns, culminating in a massive decline of function in diseases, such as sepsis.9 AKI can be characterised using functional markers, such as creatinine and urine output, and biomarkers of kidney damage, such as neutrophil gelatinase-associated lipocalin (NGAL) and kidney injury molecule-1 (KIM-1).10 This raised the question if any insight into AKI using biomarkers of kidney damage can be gained from the pre-clinical models of AMI supported with Impella or VA-ECMO. To address this question, adult male swine were subjected to left anterior descending artery occlusion for 90 minutes, followed by either immediate reperfusion (IRI), ventricular unloading with Impella for 30 minutes prior to reperfusion while on support, VA-ECMO support for 30 minutes prior reperfusion while on support or sham-operated controls (n=4/group). Renal injury biomarkers, KIM-1 and NGAL, were measured in urine, and plasma sample was collected at different time points throughout the study. The results showed that the urinary KIM-1 levels were elevated in the IRI and VA-ECMO groups, but not the Impella group. No changes in plasma KIM-1 levels were observed in any group. Compared to baseline values, VA-ECMO increased urinary NGAL levels, but Impella did not. Compared to IRI, Impella reduced plasma NGAL levels after reperfusion.
The mechanism for increased urinary levels of KIM-1 with VA-ECMO and IRI was investigated. The protein expression for both the cytoplasmic tail fragment and the extracellular domain (ECD) of KIM-1 was measured by western blot. Compared to sham and Impella, IRI and VA-ECMO had reduced levels of the ECD, but higher levels of cytoplasmic tail of KIM-1 within the renal cortex. KIM-1 is a known substrate of metalloproteinases (MMP). Higher levels of MMP-9 expression were observed in the IRI and VA-ECMO groups than the sham and Impella groups, both in the renal cortex and left ventricular infarct zone, suggesting cross-talk between the left ventricle and renal cortex. These results also suggest that IRI and VA-ECMO activate MMP activity and promote the shedding of the ECD of KIM-1 from the renal cortex, whereas Impella does not.
In conclusion, the results indicate there may be cross-talk between the heart and the kidney after an AMI via inflammatory pathways, and Impella support prior to reperfusion attenuates this effect. It would be interesting to assess the levels of the biomarkers of kidney injury in patients treated with Impella versus no support in the ST-elevation MI Door-To-Unload pivotal study.