作者：Martin Möckel 柏林 Charité 医科大学

    内容概述： 影像学诊断及治疗的使用在全球范围内逐渐扩大，导致了造影剂诱发肾病的发病率越来越高。造影剂肾病一般定义为造影剂使用48 小时内血清肌酐浓度比基础值升高25%或＞ 0.5mg/dl。在一般人群，造影剂肾病的发生是可变的并且通常较低。由于某些危险因素的存在，造影剂肾病的发生率增加，但是不同中心报道的发生率显著不同。对早已存在肾脏疾病的患者进行对比研究发现，对比剂肾病的发生率达到42%。对比剂肾病的发展与显著增加的院内和一年死亡率密切相关，特别是需要血液透析时。本文从实验证据的病理生理学背景、对比剂肾病的临床特点和对如何选择对照的临床数据争论等方面进行了阐述。

    1. Introduction

    Radiographic diagnostic and therapeutic procedures are increasingly being used worldwide and this has resulted in an increasing incidence of contrast media induced nephropathy (CIN). CIN is usually defined as an increase in serum creatinine (Scr) concentration of ＞25% above baseline or ＞0.5 mg/dl within 48 hours after radio contrast media (RCM) administration. The occurrence of CIN is variable and usually low in the general population. Depending on the presence of certain risk factors, the incidence of CIN increases, but the rates reported from different centers vary significantly. In patients with pre-existing kidney disease who undergo contrast studies, CIN occurs in up to 42% of cases. The development of CIN is associated with a significant increase in both in-hospital and one year mortality, especially if hemodialysis treatment is required.

    Extensive research has been dedicated to identifying risk factors for a decline in renal function after RCM application. Pre-existing chronic kidney disease (CKD), the presence of congestive heart failure, diabetes mellitus and RCM cumulative dose were shown to be the most important and independent risk factors. In patients who undergo coronary angiography and/or intervention one or more of these risk factors is quite often found.

    The impairment of renal function after RCM exposure is multi-factorial. Mainly two pathophysiological mechanisms which are related to the contast media have been proposed, RCM induced direct cytotoxic effects and RCM induced asoconstriction followed by either regional hypoxic damage or ischemia / reperfusion injury. These changes have mostly been adjucated to the high osmolality of contrast media, because in an early landmark experimental study in dogs, Katzberg and colleagues showed a biphasic response to intraarterial renal RCM administration: a short period of hyperperfusion, then vasoconstriction with hypoperfusion and restoration of normal flow within several minutes. The hypoperfusion was more pronounced with higher osmolarity of the contrast medium. Recently, another mechanism, which is also related to the physico-chemical properties of different agents has been proposed, namely rheological disturbances with consecutive renal medullary hypoxia due to increased viscosity in the arterioles of the kidney.

    2. Experimental evidence of the pathophysiological background

    Renal hemodynamic response

    Vasoconstriction that CM induced have been shown before more than three decades and described as biphasic response during first couple of minutes after CM administration. Increase in renal blood flow (RBF), followed by decrease was confirmed in few experimental studies, measuring RBF in renal arteries during renal angiography. Intravenous application of contrast showed different renal hemodynamic response. Human studies that measured RBF, using left renal vein thermodilution catheter, could not confirm fall in RBF after bolus of contrast needed for left ventricle angiography and relation between CIN, although biphasic pattern was occasionally observed during longer time frame (hours). Data from experimental studies further emphasized importance of CM induced disturbances in microcirculation as an important pathophysiologic mechanism. In the case of coronary angiography and left ventricle angiography, bolus of 25% of total amount of contrast medium with high or low osmolality directly enters renal arteries. This situation could be characterized by vasoconstriction of renal artery and distal branches, which could potentially lead to dominant hemodynamic participation in radio-contrast toxicity although acute vasoconstriction was never confirmed in man.

    Contrast media cell toxicity

    Contrast media renal cell toxicity has been considered as the other mechanism that causes CIN. Fall in oxygen partial
pressure (PO₂) in renal medulla caused by CM has been confirmed in experimental conditions. Renal medulla tubular
cells (TC) have been challenged after CM administration by increased higher energy demands needed for sodium reabsorption in distal tubule. This leads to decrease in PO₂, which could be further aggravated by dopamine caused redistribution of blood toward renal cortex. Experiments with use of non-ionic and ionic CM confirmed more frequent functional disturbances when ionic CM had been added to cultured TC. In recent paper Zager et al have shown direct CM damage to TC, apart from reperfusion injury, after ischemia caused by CM induced vasoconstriction and tubular obstruction. Experiment on mouse proximal TC clearly shown that CM induced plasma membrane damage characterized by loss of Na-K-ATPase, and caveolin and mitochondrial integrity impairment.

    Other experimental evidences for CM induced renal TC toxic injury have shown impaired viability, impairment in tight
junction associated proteins, decrease in content of energy molecules and K⁺, increase in TC Ca²⁺ content, reduced
activity of antioxidant enzymes, lipid peroxidation of proximal TC membranes and apoptosis. These results were summarized elsewhere.

    Effect of viscosity

    Recently, Seeliger and colleagues administered contrast media of equivalent iodine concentrations but differing
physiocochemical properties: the highosmolality iopromide was compared to the high-viscosity iodixanol. In addition,
the non-iodinated substances mannitol (equivalent osmolality to iopromide) and dextran (equivalent viscosity to
iodixanol) were also studied. Both types of contrast media transiently increased renal and hindquarter blood flow in
experimental rats. The high-osmolality agents iopromide and mannitol markedly increased urine production whereas iodixanol, which caused less diuresis, significantly enhanced urine viscosity. Only the high-viscosity agents iodixanol
and dextran decreased renal medullary blood flux, erythrocyte concentration, and PO₂. Moreover, iodixanol prolonged
the tubuloglomerular feedback response and increased plasma creatinine levels to a greater extent than iopromide
or dextran. The authors conclude that the viscosity of contrast media may play a significant role in contrastinduced
nephropathy.

    3. Clinical context of CIN:Risk factors and procedural characteristics

    The concept of CIN as a complication of radiocontrast media exposure only is challenged by several facts:

    ♦ Day-to-day variation in creatinine levels in one individual is very high and therefore spontaneous changes may mimick creatinine increase due to a specific cause.

    ♦ The rate of patients with worsening of renal function after hospital admission is very high and occurs also in those who did not underwent radiographic procedures.

    ♦ CIN-rate is higher after coronary procedure than after venous contrast application.

    ♦ Patients with complicated creatinine increase after coronary procedures are characterized by risk factors which influence independently the renal function as hypotension, shock and heart failure.

    Putting these points into perspective it means that CIN should better be translated into "cardiovascular intervention
nephropathy" and has much to do with renal perfusion changes during the procedure, which are most deleterious in
patients with pre-existing disturbances of renal auto-regulation due to several factors including typical medications
in atherosclerotic patients like ACEinhibitors and Angiotensin receptor blockers. Drop of blood pressure even without acute symptoms of the patient and maybe within physiological range maybe one predominant cause of worsening renal function in patients undergoing coronary procedure. This concept is referred to as "normotensive ischemic acute renal failure".

    The application of this concept in CIN can be done by looking in the details of a paper which analyses risk factors
for CIN in a large number of patients. In this landmark study a number of risk factors are summarized to build a score

    which predicts the probability of creatinine increase after coronary procedures. The top three factors which contribute
most to the development of CIN are hypotension, use of intra aortic balloon pump (in shock) and chronic heart failure (with reduced perfusion of the kidney). The amount of CM used ranks only number 8 in the end of the list. The table 1 shows the Mehran-Score in detail.

    Table 1.

    4. Clinical data on the controversy around the choice of modern contrast

    There has been a debate in the past decade as to whether the well-established class of the so-called low-osmolar
contrast media (LOCM) posses an inferior renal tolerance as compared to the so-called iso-osmolar contrast media (IOCM).The majority of the LOCM compounds consist of non-ionic monomers, with such well-known compounds as iopromide (Ultravist®, Bayer Schering Pharma), iohexol (Omnipaque®,GE Healthcase), iopamidol (Isovue®, Bracco), or ioversol (Optiray ®, Covidien). The LOCM are moderately hyper-osmolar compared to blood and posses a low viscosity even at higher iodine concentrations. The LOCM are remarkably well tolerated substances and have been administered in well over 1.000 million patients globally. The class of the IOCM was created in the quest to further reduce osmolality in the mid 1990ies.The reduction of osmolality to blood osmolality was achieved by creating a non-ionic dimeric structure, but at the price of a considerably increased viscosity at similar iodine concentration. Iodixanol (VisipaqueTM, GE Health) is the only representative of this class indicated for intra-vascular use on the market. The IOCM have a clinically proven significant higher rate of delayed skin reactions compared to the LOCM class.

    The NEPHRIC trial in early 2003 created the hypothesis that the IOCM may have a better renal tolerance compared to the established LOCM. The NEPHRIC study was a randomized, double-blind, controlled comparison of LOCM iohexol to IOCM iodixanol in 129 risk patients undergoing coronary intervention. The trial found a statistically significant lower incidence of surrogate CIN, employing the ≥0.5 mg/dl absolute increase definition, in the iodixanol group. There are a number of limitations of this study, which are largely due to the small sample size (imbalances in patient groups, some clearly favoring the iodixanol group), however, those limitations are typical for a hypothesis-generating study and as such no criticism on the study itself, but merely a reminder for the need of confirmatory trials. As of mid 2011 one has to conclude that the NEPHRIC study results have not been confirmed, quite to the contrary there are increasing indications that the choice of modern contrast is not the main factor to influence the risk of CIN.

    There are two studies, which claim or are claimed to support the NEPHRIC findings, however, this does not fully stand up to scrutiny. In the RECOVER study, a prospective, randomized, singlecenter study in 275 renally impaired patients undergoing coronary intervention, a higher rate of serum creatinine changes was found in the patients receiving the LOCM Ioxaglate compared to the IOCM iodixanol group. However, the difference was only significant with one, the most unspecific version, of the three investigated definitions of the surrogate marker CIN. The differences where not significant using the same definition as employed in NEPHRIC, putting the favourable conclusions drawn by the authors for iodixanol into question. Another single-center investigation conducted in China compared iodixanol and LOCM iopromide in 208 risk patients undergoing PCI and found mildly significant results in favor of iodixanol with one variation of the surrogate endpoint, but again not showing consistency across the results.

    Multi-center studies with larger patient numbers have found no differences in the rate of the surrogate marker CIN between the LOCM and the IOCM group. The VALOR study showed no statistical differences in the incidence of surrogate CIN in 259 risk patients in the patient group receiving the LOCM ioversol as compared to the IOCM iodixanol group. The CARE trial, a randomized, double-blind, controlled investigation of 414 renally impaired patients undergoing coronary intervention found consistently neutral results comparing the LOCM and the IOCM group. Similar neutral results were obtained in the ICON trial (LOCM ioxaglate) the Jürgens et al trial (LOCM iopromide) and the CONTRAST trial (LOCM ioversol). And finally, the Laskey et al. trial should be mentioned. This trial sponsored by the manufacturer of iodixanol is in design almost identical to the NEPHRIC study, except that it uses a different LOCM comparator and stringently employs current state of the art hydration prophylaxis. This investigation of 526 risk patients found no significant differences between the LOCM and the IOCM group regarding the incidence of the surrogate marker CIN, but notably in the per protocol set, the rate of what the authors call 'conservative CIN' (serum creatinine increases where likely noncontrast related causes were excluded) was almost two-fold higher (9.3%) in the iodixanol group as compared to the LOCM group (5.4%) (see figure). A recent meta-analysis of all published CIN trials confirms that there is no benefit of the IOCM over LOCM regarding rate of CIN in risk patients undergoing cardiac intervention.

    Figure 1. Comparison of the results of one small (n=129, Aspelin 2003) and the larger second study (n=526, Laskey 2009) on iodixanol versus LOCM (low osmolar contrast media). The difference in the incidence of CIN in the pilot study was not confirmed.

    Given the limitations of the surrogate definition of CIN studies employing patient outcomes should also be considered
in the evaluation of preventive strategies. Here the picture is also clear. There are to-date two large registry analyses comparing renal tolerance of LOCM and IOCM. Both of these largepatient registries find no benefit of IOCM iodixanol on the contrary, both registries find an about two-fold higher incidence of hard clinical outcomes such as clinically diagnosed renal failure or need for dialysis in the IOCM patient group. While the first analysis concluded that these differences are related to the type of contrast, the second register study found that a riskadjustment between the two groups level the differences to result in a neutral outcome in the comparison of IOCM and LOCM.

    From the currently published clinical evidence one can conclude that there is overwhelming and consistent evidence,
that iso-osmolar, high-viscous CM do not offer advantages over low-osmolar, low-viscous CM regarding renal safety.

    5. Current interpretation of the evidence

    Choice of contrast media

    Interventional cardiologists focus their interest on performing an excellent intervention in which CM is only one
necessary means. Therefore, a huge number of details need attention and CM is one of them. The topic of renal function
in cardiovascular intervention is addressed in the 2009 Focused Updates: ACC/AHA Guidelines for the Management of Patients With ST-Elevation Myocardial Infarction and in the actual Guidelines on myocardial revascularization of the European Society of Cardiology (ESC).

    In the American guidelines "in patients with chronic kidney disease undergoing angiography who are not undergoing
chronic dialysis, either an isoosmolar contrast medium or a low-molecular-weight contrast medium" is recommended.

    The European guidelines clearly state to use "LOCM or IOCM" and less than 350ml volume or ＜ 4ml 7kg of bodyweight.

    Preventive measures

    The actually recommended preventive measures consist of optimal medical therapy of the underlying atherosclerotic
disease including medication with statins, ACE inhibitors or sartans and β-blockers; hydration with isotonic saline; considering of N-Acetylcysteine and infusion of 0.84% sodium bicarbonate. A prophylactic hemofiltration can be considered in patients with severe chronic kidney disease.

    6. Perspectives in the perception of CIN

    Actual guidelines define chronic kidney disease (CKD) in terms of strata of the calculated glomerular filtration rate
(GFR). GFR is nowadays usually estimated by using the MDRDformula (eGFR) and categorized into three groups: Patients with mild (eGFR ≥ 60 and ＜ 90ml/min/1.73m²) or moderate (eGFR ≥30 and ＜ 60ml/min/1.73m²) chronic kidney disease; patients with severe chronic and end stage renal disease or in hemodialysis (eGFR ＜ 60ml/min/1.73m²). Recommendations are differentiated among these groups.

    Once research has been more focused on the importance of CKD, also acute renal injury has been more accurately defined.There are actually two important classification systems in use, called the Risk, Injury, Failure, Loss, and End-stage Kidney (RIFLE) and the Acute Kidney Injury Network (AKIN) classifications. Although it was clearly shown in many analyses
that CIN by the traditional definition as outlined above (introduction section) is related to outcome, there is a tendency to classify acute renal injury more accurate using the new classification. In the recently performed DIRECT study the primary endpoint was chosen according to the 'R' ('R' for risk) –criterion of RIFLE, which is a 50% increase of serum creatinine within 3 days after the procedure.

    As discussed in detail above, a more sophisticated view on the nature and pathophysiology of kidney injury in patients
undergoing interventional procedures, led to the discovery that the contrast media itself is just one factors in a multicause entity.

    7. Future of CIN and safe interventions in patients with renal impairment

    Patients with CKD are prone for complications and need special vigilance to minimize adverse outcomes and provide the basis for effective and safe cardiovascular interventions. The basic principles of safe interventions in CKD patients are summarized in table 2.

    Table 2: Safe interventions in patients with CKD