Introduction

Timely treatment with primary percutaneous coronary intervention (primary PCI) is the recommended treatment for patients with ST-segment elevation myocardial infarction (STEMI) [1]. The prognosis for patients with STEMI has improved during the last decades, and the 1-year mortality is now around 11% [2]. However, some patients may still have an unfavorable outcome. Acute Kidney Injury (AKI) following PCI is a common complication that is observed in >15% of STEMI patients treated with primary PCI [3,4]. The incidence of AKI is higher following a primary PCI than after elective PCI suggestively owing to the use of more contrast agent during the procedure [5,6]. AKI may occur in patients with both preexisting renal disease and in patients with normal baseline renal function [7]. In STEMI patients the development of AKI following primary PCI has previously been related to an increased in-hospital stay and 3-year mortality [8,9]. Most previous studies on the incidence, risk factors and prognostic importance of AKI in patients with STEMI treated with primary PCI have included patients from randomized studies with various in- and exclusion criteria. Thus, these studies do not represent the full spectrum of patients with STEMI and there are no data on the importance of AKI in an all-comer STEMI population and on the mortality beyond 3-years. Therefore, the aims of the present study were to:

1. Report the incidence of AKI among an all-comer STEMI population undergoing primary PCI
2. Identify the risk factors for the development of AKI after primary PCI in STEMI patients
3. Report the long-term effects of AKI on mortality in a non-selected all-comer STEMI population.

Methods

Results

Baseline characteristics

A total of 4239 patients were treated with primary PCI between November 1, 2009 and December 31, 2014 of whom 4002 (94%) had available creatinine measurements allowing for assessment of AKI and thus inclusion in this study. A total of 765 patients (19.1%) developed AKI within 72 hours after primary PCI. Differences in the baseline characteristics between patients with and without AKI are summarized in Table 1. In brief, patients with AKI were older, more often female, and with lower left ventricular ejection fraction (LVEF), lower estimated glomerular filtration rate at admission (eGFR) as well as higher contrast volume/eGFR ratio. Patient with AKI also had a higher frequency of cardiogenic shock and pre-admission cardiac arrest. Both time from symptom onset to primary PCI and duration of the procedure were longer for patients that developed AKI.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Baseline characteristics of AKI vs. non-AKI.

https://doi.org/10.1371/journal.pone.0226625.t001

Independent predictors of AKI

In the multivariate backward elimination logistic model, the independent predictors for the development of AKI were: Peak TnT, symptom-to-PCI time, female sex and contrast volume to eGFR ratio, whereas age and smoking were statistically insignificant in the last step of the analysis. The analysis of independent predictors is summarized in Table 2. Since, LVEF at admission was only available in 24.8% of the patients it was excluded from the multivariable analyses.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Multivariate analysis of the significant risk factors for AKI.

https://doi.org/10.1371/journal.pone.0226625.t002

Mortality and AKI

Throughout the follow-up period of 3.4 years (range 0–5.52 years) a total of 527 (13.2%) patients died. The presence of AKI was statistically significant associated with higher mortality rates compared to patients without AKI during the entire follow-up period and at each landmark analysis, Table 3 and Fig 1. A landmark analysis did also show that the occurrence of AKI was related to mortality from 30 days to end of follow-up (HR 1.70 (95% 1.31–2.23)). Moreover, AKI remained independently associated with long-term mortality in multivariate Cox regression models adjusting for variables found to be associated with mortality, Table 4. Other important factors associated with long-term mortality were age, previous CABG, history of stroke, history of heart failure, cardiogenic shock at admission, resuscitated cardiac arrest, peak TnT and symptom-to-PCI.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Survival.

https://doi.org/10.1371/journal.pone.0226625.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Mortality.

https://doi.org/10.1371/journal.pone.0226625.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Multivariate analysis of the significant risk factors affecting mortality.

https://doi.org/10.1371/journal.pone.0226625.t004

Discussion

In an unselected all-comer population AKI occurred in one in five patients within the first 72 hours after primary PCI and was strongly related to both short and long-term prognosis. In previous studies the incidence of AKI following primary PCI in STEMI varied between 15–30% depending on study population, design and definition of AKI [3,4,8]. In comparison the incidence of AKI in the general population after contrast media consuming procedures (not restricted to PCI) is estimated to be only 1–2% [10]. There is no global consensus on the definition of AKI and in the present study we used the definition recommended in the ESUR guidelines which is an increase of 25% in serum creatinine from baseline creatinine within 72 hours from contrast media administration [4].

In the “Kidney Disease Improving Global Outcomes” criteria (KDIGO) there are 3 stages of AKI, stage 1 using 1,5–1,9 increase from baseline creatinine and stage 3 using more 3 times increase from baseline creatinine, and we translate this as a severity score, therefore 25% increase could presumably be the first point in developing the AKI and gradually getting worse with increasing creatinine levels [11].

The higher mortality rates for AKI patients is a consistent finding across studies [12]. A sub-study from the HORIZONS-AMI study found a mortality rate of 8.0% at 30 days and 16.2% after 3 years of follow up, as compared to 0.9% and 4.5% respectively for patients with and without AKI [8]. In a similar study a sevenfold increase in mortality (23.3% vs. 3.2%) was seen at 1 year [9]. In the present study AKI was related to a 39% relative increase in all-cause mortality through a median observation time of 3.4 years.

Thus, our results confirm previous findings, but add important data regarding long-term mortality and demonstrate the clinical relevance of AKI in an all-comer STEMI population including patients with out-of-hospital cardiac arrest and cardiogenic shock [13]. The reason for poorer outcome following AKI is still unknown. It may be due to preexisting illness, mainly of either renal or atherosclerotic character [14]. Another contributing factor may be damage to other organs during acute renal failure. Previous studies found severely diminished coronary vascular tone, reserve and reactivity after acute renal failure [11].

The present study is a true all-comer population including patients with an inherent risk of nephropathy such as patients with cardiogenic shock and cardiac arrest. These patients are almost always excluded from randomized clinical trials. In the recently published CULPRIT-SHOCK trial, the need for renal replacement therapy were as high as 11.6–16.4% among the study participants, underlining that nephropathy is a common complication in patients with acute myocardial infarction and shock [15]. Cardiogenic shock at admission and cardiac arrest before the procedure were not independently associated with AKI in the present study, possibly explained by the low frequency of both cardiac arrest and cardiogenic shock, and the fact that the most critically ill patients die before additional blood work-up can be performed.

Previously other risk factors for AKI have been reported such as old age, diabetic nephropathy, time from onset of symptoms to primary PCI, preexisting renal disease, congestive heart failure, contrast volume and length of the procedure [10,16][6,8,17],[18]. And in a recent study with 1656 STEMI patients both congestive heart failure and reduced left ventricular ejection fraction were identified as individual predictors of AKI [19].

In the present study age, female sex, time to treatment, peak TnT level and the contrast volume to eGFR ratio were all independent predictors for the development of AKI. Contrast volume, preexisting diabetes and procedure length were not independently associated with AKI in the present population may be explained by several factors. Earlier studies reported much higher average use of contrast media volume among patients developing AKI(153–378 ml) compared to the 111–114 ml used in the present study [6,8,17]. Although some studies have identified diabetes as an independent risk factor other studies have not [6,7,17,20]. One study indicates diabetes only being a risk factor combined with renal insufficiency but not per se [21]. Similar circumstances could explain the reason that the contrast to eGFR ratio is an independent predictor of the development of AKI while contrast volume is not.

That diabetes mellitus was not associated with AKI in the present study may also be due to the low prevalence of diabetes in the study population in combination with the relatively small amount of contrast used. A recent study found admission hyperglycemia to be an independent risk factor for AKI following primary PCI [22].

The association between time to treatment and AKI could be explained by the changes in renal hemodynamics that occurs during an acute myocardial infarction [18]. Also, longer ischemia times may lead to larger infarcts and subsequent heart failure, which also explains why infarct size determined by peak troponin T was associated with AKI.

Controversies still remain over the effectiveness of interventions to prevent AKI, e.g. hydrating with intravenous fluid (IV) and administering drugs like N-Acetylcysteine or Fenoldopam Mesylate [23,24,25,26]. Although most studies state no benefit from these interventions, a few randomized trials have found hydration with IV normal saline to be consistently effective. Mueller et al performed a large prospective randomized study comparing isotonic IV saline (0.9%) with half-isotonic IV saline (0.45%) and concluded that isotonic saline (0.9%) was superior in terms of reducing AKI. Unstable cardiac patients undergoing primary interventions were the most likely to benefit from isotonic hydration [27]. Starting the hydration earlier and continuing longer are considered best practice. However, no single regimen of IV saline has yet been established in practice [28]. A recent single-center retrospective study suggested that a micro axial percutaneous left ventricular assist device may protect against acquired kidney injury during high risk PCI. The majority of the patients in that study had acute coronary syndromes but only 13% were true STEMI[29]. Finally, lowering the patient's blood glucose levels before procedure have been suggested by some authors [22].

A recent review summarizes the pathophysiology and highlights preventative strategies for patients undergoing angiographic procedures. It also confirms that diabetes mellitus has not been shown to be an independent risk factor. It states that contrast volume > 350ml or repeated administration within 72 hours was shown to be associated with an increased risk. The review states that developing AKI is correlated with accelerated progression of an underlying chronic kidney disease. The review concludes that additional studies is needed to effectively determine the true toxic effect and to evaluate the survival benefit of preventing AKI [30].

Conclusion

In STEMI patients treated with primary PCI one in five experiences acute kidney injury, which was related to substantial increase in both short- and long-term mortality.