The predictive value of the renal resistive index for contrast-induced nephropathy in patients with acute coronary syndrome

Percutaneous coronary intervention (PCI) is associated with contrast-induced nephropathy (CIN) rates that vary between 0 and 24% depending on a patient’s risk factors [1]. Despite the fact that most cases are transient and reversible, CIN is associated with significant morbidity and mortality. Therefore, prevention and early diagnosis of CIN are extremely important. Currently, the diagnosis of CIN still depends on serum creatinine monitoring, which is a late marker of acute kidney injury (AKI). Several new potential serum and urinary biomarkers of AKI have been reported. Nevertheless, as solely fascinating findings at the basic science level, they remain far from clinical utility [2].

The renal resistive index (RRI) is one of the most sensitive parameters in renal pathology and reflects alterations in renal plasma flow. The RRI has been proven clinically useful for assessing renovascular hypertension, determining the prognosis for renal allografts and calculating AKI risk in the intensive care unit [3, 4]. Contrast medium (CM)-induced alterations in renal microvascular haemodynamics are considered to be one of the mechanisms in the development of CIN [1]. However, few studies have focused on the relationship between RRI and CIN. Hetzel et al. revealed that RRIs rose transiently after the administration of CM and then decreased progressively to baseline values [5]. Thus, we hypothesized that RRI is a good marker for determining the risk of developing CIN. Additionally, we considered that persistently elevated RRIs after CM exposure could indicate the possibility of contrast-induced AKI (CI-AKI).

A total of 172 patients with ACS underwent PCI at our centre between October 2015 and January 2017; of these, 23 patients were excluded according to the exclusion criteria and 3 patients declined to sign the written informed consent (Fig. 1). Finally, 146 patients were enrolled and analysed in this prospective study, including 65 patients with STEMI (44.5%) and 81 with NSTE-ACS (55.5%). There were 31 women and 115 men with a mean age of 56.6 ± 12.4 years. The mean eGFR was 92.4 ± 28.0 mL/min/1.73 m², which included 15 patients with moderate and 4 with severe renal impairments. The percentage of patients with complicating conditions including hypertension, diabetes, hyperlipidaemia and hyperuricemia were 30.8, 47.9, 50.7 and 22.6% respectively. Fifty-six patients (38.4%) had multi-vessel disease (more than triple-vessel) and the left anterior descending branch (LAD) was the target vessel in 63 (43.2%). Transient peri-procedural hypotension occurred in 6 patients (4.1%) and none required IABP placement.

The mean pre-procedure SCr value was 84.5 ± 37.3 μmol/L, while at 48 h post-procedure, the mean SCr value rose to 90.4 ± 38.5 μmol/L. CIN developed in 31 patients (21.2%), but none required haemodialysis. There were no significant differences in the rates of diabetes, hypertension and hyperlipidaemia or differences in other pre-procedure parameters including age, body mass index, CM and SCr between the CIN and non-CIN groups. Nevertheless, patients with CIN were characterized by higher hydration volume, higher incidence of hyperuricemia, and lower eGFRs. Moreover, a trend toward lower LVEFs and the LAD being the more likely target vessel was observed in patients with CIN (Table 1).

RRI showed no significant differences before and after CM exposure in all of the patients, and also in the non-CIN group. Nevertheless, a much higher RRI value was observed at 1 h and 24 h after CM exposure in the patients with CIN (Table 2). An analysis of the CIN group showed that, compared to baseline, the RRI rose significantly at 1 h (0.71 ± 0.05 vs 0.68 ± 0.05, respectively; p = 0.030), but there was no significant elevation at 24 h (0.70 ± 0.05 vs 0.68 ± 0.05, respectively; p = 0.071) (Table 2).

The ROC curve analysis indicated that the pre-procedure RRI was a powerful predictor of CIN (area under the curve = 0.661, 95% confidence interval (CI) 0.549–0.772; p = 0.006). Next, we determined that the best cutoff value based on the Youden index was 0.69, with 67.7% sensitivity and 67% specificity (Fig. 1).

The selected variables and results of the multivariate logistic regression analysis are presented in Table 3. The multivariate logistic regression analysis revealed that a pre-procedure RRI ≥0.69 was positively associated with CIN (odds ratio (OR) = 4.445; 95% CI 1.806–10.937; p = 0.001). Other significant factors included hyperuricemia (OR 2.607; 95% CI 1.007–6.749; p = 0.048) and chronic kidney disease (CKD; OR 3.163; 95% CI 0.999–10.013; p = 0.050) (Table 3).

The present study was designed to identify the risk factors for CI-AKI relative to the RRI, which was a novel undertaking. Our results indicate that the renal vascular bed pathology contributes to an increase in the incidence of CI-AKI in patients with ACS who are referred for PCI.

RRI is calculated as (Vsyst-Vdiast)/Vsyst [8] and reflects the resistance in the renal vascular beds. RRI is affected by renal interstitial disease or vascular compliance as well as extra-renal factors such as blood pressure, heart rate and rhythm, age and significant aortic valve stenosis [9]. To date, it is one of the most sensitive parameters for kidney disease and can provide useful information about the pathophysiology of renal diseases in both native and transplanted kidneys [3]. However, there has been little focus on the relationship between RRI and CI-AKI. Wybraniec et al. firstly revealed that a high pre-procedure RRI value may be a useful and novel risk factor for CI-AKI [10]. However, there were some limitations in that study, such as the absence of patients with pre-existing renal function impairment, and only 9 patients with evidence of CI-AKI onset. In our study, we maintained the same viewpoint as Wybraniec and found that the RRI after contrast administration was higher in patients with CI-AKI and was likely to be higher at baseline. Normal RRI values in adults range from 0.47 to 0.70. The patients enrolled in our study could be considered to have ‘high-normal’ RRI values, which may mainly affectedly by age and vascular bed atherosclerotic, appropriately depending on these high rates of concomitant diseases entity (e.g., diabetes, CKD) [11]. Nevertheless, no significant differences were observed between the two groups with respect to extra-renal factors, and peri-procedural hypotension occurred in only a small number of patients. Therefore, a higher prone baseline RRI value in the CIN group may mainly indicate that the renal vascular resistance tended to be higher, either for renal interstitial disease or vascular compliance. Thus, patients with CIN had worse baseline renal function or more severe vascular bed atherosclerosis, which could explain why the RRIs after CM administration rose significantly in the CIN group and the patients with a relatively high baseline RRI were more prone to developing CI-AKI. Our study showed that patients with a high RRI (≥0.69) were more likely to develop CI-AKI.

In clinical practice, most cases of CI-AKI are reversible and seldom are these patients referred for haemodialysis. There is a similar alteration in RRI values following CM exposure. Hetzel et al. showed that RRI increased significantly within minutes after the administration of intravenous CM and then decreased progressively to baseline [5]. This could explain why RRI, after CM exposure in the non-CIN group, was not significantly different in comparison to the baseline standard value in our study. RRI rose transiently after CM exposure in general; nevertheless, this RRI alteration seemed to be prolonged and even lasted more than several hours in the CIN group. We think that the prolonged alteration might be relevant in patients with CI-AKI that have a delayed vascular response to CM [12]. Although subsequent changes in RRI were not recorded, the SCr returned to baseline and the absence of oliguria or anuria suggested that these changes were functional and reversible. This provides important evidence for the hypothesis that CI-AKI is related to vasoconstriction in the renal vessel beds, especially after CM exposure, which causes renal medullary hypoxia.

The multivariate logistic regression analysis showed that a high baseline RRI (≥0.69), hyperuricemia and CKD accurately predicted CI-AKI. LAD as a target vessel, LVEF, eGFR and 24-h hydration volumes were thought to be risk factors based solely on the univariate analysis, but these factors were excluded after the multivariate analysis. It is no doubt that patients with baseline CKD are high-risk for CI-AKI. Hyperuricemia has been demonstrated to be independently associated with an increased risk of CKD in cross-sectional studies [13]. In contrast to previous studies, our results excluded risk factors such as age, contrast volume, diabetes mellitus and a history of congestive heart failure (CHF) [14]. There might be several reasons why these differences were related to our unique study design. First, the mean age of the patients was 56.6 ± 12.4 years; therefore, they might have had diabetes for only a short time and atherosclerosis may not yet have developed. On the other hand, RRI was also thought to be closely related to atherosclerosis which may cause a confusing effect. [11]. Second, we enrolled a large number of patients with STEMI, most of whom required primary PCI and were administered relatively small volumes of CM (≤100 mL, in general). Finally, we did not discriminate between pre-existing CHF and new-onset CHF caused by acute myocardial infarction, which may have also been a confounder.

The present study had several limitations. First, it was a single-centre study, and we enrolled a relatively small number of patients with ACS. Second, we may have overestimated the true incidence of CI-AKI. Some patients may have easily met the diagnostic criteria for CIN (a relative rise of SCr ≥25%) after CM exposure, especially when the baseline SCr was relatively low. Moreover, SCr values could have been affected by haemodynamic changes in the setting of acute myocardial infarction. Third, although we were committed to minimizing errors (colour Doppler was performed in three different areas of the kidney and was focused on the arcuate/interlobular arteries), the RRI values may have been affected by the view angle and patient positioning. Thus, employing an invasive Doppler flow wire might be an ideal alternative method of measuring RRI [12].

The current study showed that a relatively high pre-procedural RRI appears to be an independent predictor of CIN in patients undergoing PCI and that preoperative RRI assessment facilitates stratification of risk for CI-AKI among patients referred for coronary angiography. A significantly increased RRI after CM administration is suggestive of developing CI-AKI; however, the absolute value of the increase is still unknown and may be our focus of study in the future. Moreover, an analysis of intrarenal arterial Doppler flow profiles constitutes a favourable means of investigating the pathophysiologic mechanisms of CI-AKI.