The third strike

“What is man, when you come to think upon him, but a minutely set, ingenious machine for turning, with infinite artfulness, the red wine of Shiraz into urine.” —Isak Dinesen, 1934 It has long been recognized that intravascular administration of iodinated contrast material may be associated with acute nephrotoxicity (CAN). Until recently, the interventionalist’s impression has been that, when it occurs, CAN is most often a temporary phenomenon that self-corrects over 1 or 2 weeks. Only occasionally is there a need for short-term dialysis and very infrequently is a state of end-stage renal disease reached. The disorder has been defined in terms of an absolute or percent of baseline increase in serum creatinine level (e.g., an increase in creatinine level of 0.5 mg/dl or 25% of baseline). Those with normal or mildly elevated baseline creatinines who go on to meet these criteria postcontrast are usually asymptomatic, rarely have sequelae, and remain, in many cases, unrecognized or unappreciated since creatinine does not peak until 3–5 days after exposure, by which time most patients are home. Certainly, patients with high baseline creatinines ( 2.5 mg/dl) are at greatest risk of CAN and its adverse consequences, thus requiring closer monitoring. One of the more perplexing aspects of CAN, in view of the relatively benign description given above, is the number of reports suggesting an extremely poor prognosis associated with it. Levy et al. [1] identified 183 in-patients who developed CAN after contrast exposure (1% of all in-patients exposed to contrast). The mortality during hospitalization of this group was 34% compared to a 7% mortality in an ageand baseline creatininematched group of contrast recipients who did not develop CAN. They concluded that acute renal failure seemed to increase the risk of developing nonrenal complications that led to death. More recently, a number of studies examining patients undergoing percutaneous coronary intervention arrive at similar results. Rihal et al. [2] found a 3.3% incidence of acute renal failure in their PCI registry of 7,586 patients. Twenty-two percent of those with CAN died during the index hospitalization as opposed to only 1.4% of those without acute renal failure. Hospital survivors with CAN had a 5-year mortality of 45% compared with a 15% mortality rate in those not developing acute renal failure. In patients with baseline creatinines 1.8 mg/dL undergoing PCI, Gruberg et al. [3] found that 37% developed CAN with an in-hospital mortality of 15% compared to a 5% incidence of death in those who did not develop acute nephrotoxicity. Twentythree percent of the subgroup of CAN patients requiring dialysis died. One-year mortality rates in these patients with preexistent renal dysfunction were 45%, 35%, and 19% in patients requiring dialysis, developing CAN without the need for dialysis, and not developing CAN, respectively. The presence of chronic renal insufficiency is also a predictor of poor prognosis in PCI patients. One study found that despite equally high angioplasty success rates, the in-hospital and 1-year mortalities were 6.8% and 49%, 4.2% and 26%, and 0.9% and 6% for patients with preexistent end-stage renal disease, chronic renal failure, and normal renal function, respectively [4]. A report from the Mayo Clinic group [5] found that 1-year mortality was directly related to preprocedural creatinine clearance, essentially doubling with each decrement of 20 ml/min in the latter from 70 ml/min. Patients on dialysis had approximately the same 1-year mortality ( 20%) as those with creatinine clearance 30 ml/min. Interestingly, there appears to be a benefit of bypass surgery over PCI in dialysis patients that does not extend to those with chronic renal failure [6]. The most significant risk factor for the development of CAN is preexistent renal dysfunction. From all of the above, it seems obvious that preexisting chronic renal failure as well as procedure-related acute renal dysfunction associated with PCI dramatically increases in-hospital and long-term mortality. The reason(s) for this is unclear, but may include one or more of the following: metabolic disturbances associated with renal dysfunction that might lead to bleeding, thrombosis, electrolyte disorders, drug toxicity, and accelerated restenosis; differences in the vascular substrate that might alter the response to PCI; or that CAN is just a marker for comorbid conditions such as poor left ventricular function, peripheral vascular disease, diabetes, hypotension, hypovolemia, etc., that account for the grave prognosis. While it