Article Text

Download PDFPDF

Contrast-induced nephropathy in primary percutaneous coronary intervention
  1. Sean Gallagher,
  2. Charles Knight
  1. Department of Cardiology, London Chest Hospital, London, UK
  1. Correspondence to Dr Charles Knight, Department of Cardiology, London Chest Hospital, Bonner Road, London E2 9JX, UK; charles.knight{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

It is clear that the development of contrast-induced nephropathy (CIN) is both common and prognostically important in patients undergoing interventional cardiovascular procedures. CIN is now established as the third most common cause of hospital-acquired acute kidney injury (after surgery and hypotension).1 The expansion in the use of iodinated contrast media in both diagnostic and interventional procedures, in combination with an increasingly elderly infirm patient population, means that the incidence of CIN is likely to grow rapidly.

Several studies have shown that CIN is associated with increased morbidity and mortality, extended length of hospital stay and increased hospital costs.2–5 However, the key question remains whether CIN is just a marker of adverse cardiovascular, renal and procedural factors of no direct pathophysiological importance or a direct cause of increased mortality and morbidity. This is a difficult question to answer, but the paper by Wi et al published in this issue of Heart6 adds to the growing body of evidence that suggests CIN is a protagonist rather than a spectator in the adverse outcomes, and hence that we should increase our efforts to make the administration of iodinated contrast media safer for our patients.


CIN is impairment of renal function, either of new onset or an exacerbation of pre-existing renal dysfunction, following the administration of iodinated contrast media. Using the now increasingly standard definition (an absolute increase in serum creatinine of 44.4 μmol/l (0.5 mg/dl) and/or a 25% increase in serum creatinine concentration from baseline within 72 h of the administration of contrast media7), registry data suggest that the incidence of CIN in a ‘low-risk’ general population is approximately 3%,4 but the incidence can be >30% in those at highest risk of developing CIN.5 Numerous potential patient-related factors associated with the development of CIN have been described, but the most consistent independent risk predictor for the development of CIN is undoubtedly pre-existing kidney disease, potentially unrecognised in elderly patients with seemingly ‘normal’ serum creatinine levels but with estimated glomerular filtration rates <60 ml/min. The development and validation of the Mehran scoring system8 which considers several other prominent predictors of the development of CIN—namely, patient age, hypotension, presence of congestive heart failure, intra-aortic balloon pump use, pre-procedure serum creatinine, diabetes, anaemia and contrast volume—allows for improved recognition of procedural CIN risk and the institution of prophylactic measures, most commonly pre-procedure hydration to avoid hypovolaemia.


The development of CIN following percutaneous coronary intervention (PCI) clearly affects prognosis. Even small reductions in renal function following contrast exposure without obvious immediate clinical sequelae are prognostically important. Rich et al showed that patients with creatinine elevations ≥44 μmol/l within 48 h of cardiac catheterisation had an increased risk of in-hospital mortality compared with patients without post-procedural creatinine elevation (14.3% vs 6.2%).9 Weisbord et al retrospectively evaluated the incidence and clinical implications of CIN in 27 608 patients who underwent coronary angiography. Patients who developed small absolute (0.25–0.5 mg/dl) and small relative (25–50%) increases in serum creatinine following coronary angiography had an increased risk of in-hospital mortality, even after adjustment for comorbidity, compared with patients who did not have increases in post-procedure serum creatinine levels (risk-adjusted ORs for in-hospital mortality of 1.83 and 1.39, respectively). Larger increases in serum creatinine, particularly larger relative changes in serum creatinine (100% elevation from baseline), were associated with an even greater risk of in-hospital mortality (risk-adjusted OR 3.57).10 The prognostic implications of CIN are not confined to the short term; for example, a large retrospective analysis of more than 7500 patients undergoing procedures involving the administration of iodinated contrast media found that the in-hospital mortality rate for patients developing CIN was 22% compared with only 1.4% for patients without CIN. At 5 years this increased risk persisted, with a mortality rate of 44.6% in the patients who had previously developed CIN compared with 14.5% in patients without CIN.4 McCullough et al confirmed these observations by showing that patients who developed CIN had an in-hospital mortality rate of 7.1% compared with mortality rates of 1.1% in patients without CIN. This report also highlighted the grave consequences of CIN requiring renal replacement therapy, with a reported in-hospital mortality rate of 35.7% in this group. Again the early mortality risk associated with CIN persisted, with patients requiring renal replacement therapy having a mortality rate of 81.2% at 2 years.2

Marker or cause?

The crucial question as to whether CIN has a causal relationship with subsequent mortality or whether the development of CIN is simply a reflection of comorbidity within a high-risk population remains unresolved. Levy et al attempted to determine whether CIN was independently associated with mortality by comparing subjects who developed CIN with a control group matched for comorbidity who did not develop CIN. In-hospital mortality in the CIN group was 34% compared with 7% in the matched control group. After adjusting for comorbid conditions, the risk of mortality was 5.5-fold higher in patients who developed CIN.3

In their paper published in Heart, Wi et al report on the incidence and prognostic implications of CIN in patients treated with emergency PCI for acute myocardial infarction (AMI). This retrospective analysis of 1041 patients with AMI (515 (49.5%) patients with ST-segment elevation myocardial infarction and 526 (50.5%) patients with non-ST-segment elevation myocardial infarction) is the largest study to date of CIN in PCI-treated patients with AMI. This is also the first report in such a cohort that includes extended clinical follow-up data (available for a mean duration of 22.8±15.9 months).

The incidence of CIN observed was 14.2%, which is in line with previous reports in similar AMI cohorts.11–13 Patients undergoing emergency PCI in the setting of AMI represent a particularly high-risk population for the development of CIN. Potential contributing factors for the development of CIN in this setting include the presence of hypotension (or even cardiogenic shock), large volume contrast media use for technically demanding emergency procedures and lack of pre-procedure hydration. The development of CIN following emergency PCI for AMI was independently associated with a significant increase in both in-hospital mortality (14.2% vs 2.5%, p<0.001) and long-term mortality (2-year mortality rate 24.9% vs 6.3%, p<0.001) compared with patients who did not develop renal dysfunction after PCI. Crucially, the authors show that CIN is not a transient self-limiting event. Of the 148 patients who developed CIN, 80 (54.1%) showed partial or complete recovery of renal function (defined as a return of serum creatinine level to <25% or 0.5 mg/dl (44.2 μmol/l) above the baseline level at 1 month after PCI), but renal function failed to improve at all within 1 month in 68 patients (45.9%). It has often been assumed that the apparent renal impairment seen with the administration of contrast media is of no long-term importance because it resolves fully. In this study renal dysfunction persisted in nearly half of the patients with important prognostic effects; patients with persistent renal dysfunction after exposure to contrast media had higher all-cause mortality rates during follow-up than patients with transient renal dysfunction (34.6% vs 16.7%, p=0.012). Transient renal dysfunction following exposure to contrast media was still associated with an increase in mortality compared with patients who did not develop renal impairment (16.7% vs 6.3%, p=0.001), although not to the degree seen in patients with persisting renal dysfunction after PCI. The ‘transient’ group will presumably be made up of some patients with complete recovery and some with incomplete recovery and will therefore be, to some extent, a hybrid, explaining its intermediate prognosis. This study therefore concurs with the Dartmouth Dynamic Registry14 that CIN may, to some extent, be a permanent rather than a transient phenomenon and extends the previous findings to note a prognostic effect from persistent renal dysfunction. Once CIN has been demonstrated—at least in some patients—to cause irreversible kidney injury, the thesis that it is not directly related to mortality and morbidity becomes harder to sustain.

As with much research in this field, this report is limited by the retrospective single-centre design used. Several further limitations of the study should also be recognised. First, the incidence of CIN may be an underestimate as renal function was evaluated 48 h post-procedure, before it had reached a peak, rather than the standard 72 h post-procedure. Second, daily fluxes in serum creatinine occur due to hydration and nutritional status and single measurements may not be representative of true renal function. This is especially relevant to single measures of renal function 1  month post-procedure which were used to diagnose persisting renal dysfunction. Finally, it should be emphasised that there are a multitude of reasons for renal injury after PCI and AMI, and some events classified as CIN may be due to other causes—for instance, transient hypotension or atheroemboli complicating the procedure which were imperceptible at the time. It is therefore not possible to state with certainty the proportion of kidney injury in this study that is directly attributable to the administration of contrast media.


Interest in the phenomenon of CIN is necessarily reduced by a lack of effective treatments. Although various prophylactic therapies have been shown to be effective in small individual studies, consistent evidence of benefit has been difficult to demonstrate conclusively for most individual therapies. To date, the only universally accepted prophylactic therapy for CIN is peri-procedural intravenous hydration, but this is often logistically challenging in patients treated with emergency angiography and PCI. However, the pre-procedural recognition of the risk of CIN is the key to improving patient outcomes. It is important that the Mehran score predicts CIN, and thus subsequent mortality, within an AMI cohort.15 In high-risk patients hypovolaemia may be aggressively corrected (if left ventricular function allows) and contrast volume minimised. Operators should limit the volume of contrast media to the lowest possible. Potential additional therapies including N-acetylcysteine may be beneficial but, before these therapies become widely adopted, adequately powered prospective studies must be undertaken. Reports such as that by Wi et al—which suggest a prognostically important and, in some patients, permanent kidney injury resulting from primary PCI for AMI—must stimulate such studies and the development of safer contrast agents.


View Abstract


  • Competing interests None.

  • Provenance and peer review Commissioned; internally peer reviewed.

Linked Articles