Randomised Controlled Trial Evaluating Prophylactic Intravenous Hydration for the Prevention of Contrast Induced Nephropathy
Randomised Controlled Trial Evaluating Prophylactic Intravenous Hydration for the Prevention of Contrast Induced Nephropathy
Contrast-induced nephropathy (CIN) is a side-effect of intravascular administration of iodinated contrast material. It is defined as an absolute (>44μmol/l) or relative (>25%) increase in serum creatinine from baseline values within 48-72 hours of iodinated contrast material administration, and usually resolves within two weeks. In some cases CIN has been associated with persistent renal failure, increased risk of dialysis, and mortality. It is not clear however, whether CIN is causally related to this increased risk or whether risk of morbidity and mortality is inherent in those at risk of CIN.
CIN itself is asymptomatic and no treatment for CIN exists. Therefore, the focus lies on its prevention. Prevention guidelines have been drawn up in most countries and been implemented in most radiological departments. In the Netherlands, currently two guidelines for the prevention of CIN coexist, issued by CBO (Centraal BegeleidingsOrgaan) and VMS (Veiligheids Management Systeem).
The prevention guidelines aim to increase patient safety by identifying patients that may be at risk of CIN (mostly patients with chronic renal insufficiency), and subsequently administering prophylactic intravenous hydration to the so identified patients, in order to prevent CIN (intravenous normal saline 4-12 hours before and 4-12 hours after exposure to iodinated contrast material).
Needless to say, the introduction of these guidelines has had a great impact on patient- and health care burden. In the Netherlands alone it is estimated that yearly 100.000 to 150.000 patients receive the prophylactic treatment, incurring a total cost of over 50 million Euro. Considering the steady yearly increase of contrast procedures and the ageing population, it is evident that, in future, these numbers shall only increase further.
The prophylactic treatment prescribed by the guidelines is based on a consensus of the opinion of experts in general agreement that the treatment is beneficial. However, the effectiveness of prophylactic hydration has never been adequately evaluated. Sufficiently large randomised trials comparing prophylactic intravenous hydration with a proper control group receiving no prophylactic treatment are not available, and baseline CIN incidences in untreated populations are unknown. Thus, it is not clear whether prophylactic hydration achieves its aim to prevent CIN.
In order to be able to take effective measures to the benefit of patient safety, it is important to distinguish between the mechanisms underlying CIN and the ensuing increased risk of morbidity and mortality: whether it be biological variation of serum creatinine, renal damage, or cholesterol embolism; whether any causality exists between these and iodinated contrast material; and whether prophylactic intravenous hydration can prevent these from occurring without incurring more risks than it removes. These, in short, are the aims of the AMACING study.
Studies evaluating the effect of various prophylactic treatments invariably focus on CIN instead of clinically relevant measures as primary outcome. CIN itself being asymptomatic, it is important to determine whether prophylactic treatment has a preventive effect on clinically relevant endpoints sometimes associated with CIN, such as dialysis and mortality. Consider for example: even if it transpires that prophylactic treatment reduces CIN incidence, it may be that intravenous hydration merely dilutes serum creatinine to such an extent that it masks CIN, having no protective effect on renal function. It has been shown that changes in volume status can influence serum creatinine levels, but a potential dilution effect of intravenous hydration has not been investigated to date.
On the other hand, it is important to realise that prophylactic intravenous hydration is not without risk. Patients may suffer mild to serious complications ranging from phlebitis to pulmonary oedema, the latter being potentially fatal. Those patients selected according to the guidelines for the risk of CIN - risk factors including poor renal function, age, diabetes and cardiac disease - are especially sensitive to complications of intravenous hydration. The risk of intravenous hydration in this population has not as yet been charted, and is not taken into account by guidelines for the prevention of CIN.
It stands to reason that a patient's pre-existing hydration status may be a determining factor for the net effect of intravenous hydration: from the same treatment dehydrated patients may enjoy benefits, whereas overhydrated patients may suffer complications. The importance of hydration status in determining the effects of prophylactic hydration, however, has not been investigated to date.
The mechanism by which prophylactic hydration may protect renal function from injury by iodinated contrast material is unclear, as the mechanism by which iodinated contrast material may induce CIN is unclear. Indeed, in patients with chronic renal insufficiency biological variation of serum creatinine in the absence of contrast material has been shown to be indistinguishable from CIN. The question has arisen in recent literature whether CIN is anything more than an asymptomatic increase in serum creatinine, lacking any prognostic negative impact, and not significantly different from that observed in controls not receiving iodinated contrast material.
Recent studies comparing patients with chronic renal deficiency receiving intravascular iodinated contrast material to those patients not receiving iodinated contrast material found no association between increase in serum creatinine (CIN) and contrast administration. Indeed, it has been suggested that renal damage after intra-arterial procedures is caused, not by contrast material, but by cholesterol embolism arising from the erosion of aortic atheromatous plaques by the catheter used in such procedures.
It is perhaps of importance to note that relatively recently, monomeric non-ionic low-osmolar iodinated contrast materials - with less toxic properties than 'traditional' contrast materials - have been introduced and are now widely used, perhaps altering the landscape of CIN.
Patient Risk Any risk incurred by participating in the AMACING study will be due to not receiving prophylactic intravenous hydration. The true risk incurred from foregoing prophylactic treatment is unknown; however, recent literature suggests that it is likely to be minimal.
The estimation of risk of CIN according to current guidelines is largely based on renal function and an eGFR threshold of <60 ml/min/1.73 m2 in combination with other risk factors is currently applied for prophylactic hydration according to Dutch guidelines. The incidence of chronic kidney disease stage 3 (eGFR 30-60 ml/min/1,73m2) in the Netherlands is 5.3%, of which at most an estimated third will have an eGFR <45 ml/min/1,73m2.
The European Society of Urogenital Radiology updated their CIN prevention guidelines in 2011 to indicate that intravenous prophylactic hydration is unnecessary in patients with an eGFR ≥45 ml/min/1.73m2 before intravenous contrast administration. Risk analyses revealed that intravenous contrast administration does not impose a nephrotoxic risk above an eGFR of 30ml/min/1.73m2. Indeed, zero incidence of dialysis and mortality is consistently reported after intravenous contrast administration, even in patients with severe renal insufficiency (eGFR<30) and in absence of prophylactic treatment. Since intravenous contrast administration procedures make up more than an estimated 70% of all contrast procedures this implies that prophylactic intravenous hydration is superfluous in the majority of patients currently receiving it. This patient population is also the larger proportion of patients to be included in our randomized controlled trial - an estimated 75% - who are therefore not thought to incur any risk from participation and not receiving prophylactic intravenous hydration.
Other patients we shall include may incur some risk of CIN: some have an eGFR between 30-44 ml/min/1.73m2, and some will be administered iodinated contrast material intra-arterially. The risk of CIN for the first group does not appear to be much elevated. A pooled overview of studies involving iodinated contrast material administration without prophylactic intravenous hydration yielded a CIN incidence of 3.9%: 83% of the patients diagnosed with CIN had an eGFR of ≥45 ml/min/1.73 m2. Baseline eGFR of the other CIN cases was not published.
As for the second group, although cases of long term adverse effects such as dialysis and mortality have rarely been reported following CIN after intravenous contrast administration, they have been reported after intra-arterial contrast administration. It has been thought, therefore, that intra-arterial administration led to more nefarious effects of iodinated contrast material than intravenous administration. This too, however, has been repeatedly put to question. A recent study found no difference in the risk of CIN after intra-arterial or intravenous contrast administration when an adjustment was made for patient related risk factors. Several studies found no increased risk of CIN after intra-arterial contrast administration as compared to intravenous administration, and one report even goes against all previous literature, reporting a higher risk of morbidity and mortality after intravenous contrast administration than after intra-arterial administration. In short, it is not clear whether an increased risk of dialysis and mortality arises from contrast administration and CIN or whether it is inherent in the patient population studied (i.e. a population requiring intra-arterial contrast procedures or requiring prophylactic hydration according to CIN prevention guidelines may conceivably have such an increased inherent risk). Indeed, chronic kidney disease, the main criterion in the guidelines for increased risk of CIN, in itself increases the risk of all-cause mortality, cardiovascular disease and progression to kidney failure.
Causality between an increase in serum creatinine after contrast administration (CIN) and adverse events has not been shown to exist. A recent meta-analysis by McDonald et al including a 157 140 contrast procedures showed no difference in incidences of CIN, dialysis, or death between patient groups receiving contrast material versus patients not receiving contrast material (7.2% CIN after contrast-enhanced scans versus 11.1% CIN after unenhanced scans in medium- to high-risk populations, suggesting that contrast material may not be causally related to CIN). More and more the opinion rises that, where it occurs, it is the risk inherent to specific populations that leads to higher morbidity/mortality incidence after CIN or specific administration routes, and that CIN is merely a marker for such populations instead of there being a causal relationship between CIN and morbidity/mortality, or even a causal relationship between all diagnosed CIN and intravascular contrast material administration.
On the other hand there is no evidence that prophylactic intravenous hydration has a protective effect on renal function. Almost all studies evaluating prophylactic intravenous hydration to date are uncontrolled trials or retrospective cohort analyses, often involving experimental additions to the standard administration of saline prescribed in the guidelines, and thus no conclusions on its effectiveness can be drawn.
In a recent Dutch study on CIN, 35 of 454 patients at risk of CIN according to current guidelines did not receive prophylactic intravenous hydration (for reasons unexplained); yet the incidence of CIN in this subgroup was not significantly higher than that found in the population having received prophylactic treatment (1/35 or 2.9%, versus 10/419 or 2.4%). Furthermore, in studies including patients receiving contrast material without prophylactic intravenous hydration (up to 94% of the patients did not undergo prophylactic intravenous hydration in some of these studies), and having severely diminished renal function (up to 49.3% of patients), low CIN incidences were seen (range: 1.3% - 5.2%; pooled incidence 3.6%), and zero long term adverse effects were reported.
Another issue is that prophylactic intravenous hydration is not without risk. Complications may occur, especially in those patients with cardiac and/or kidney disease, such as pulmonary oedema and/or cardiac failure which could lead to respiratory insufficiency. There is a considerable overlap between patients considered to be at risk of developing CIN and patients with at risk of complications from prophylactic intravenous hydration, and therefore this is a real concern in clinical practice. A study performed in a Dutch hospital using Dutch CIN prevention guidelines reported an incidence of serious complications of intravenous hydration of 1.4% in hydrated patients. The incidence of clinically relevant events after CIN is <1% when monomeric non-ionic low-osmolar iodinated contrast material is used, thus putting the appropriateness of the prophylactic treatment to question, and highlighting the importance of its evaluation against a proper control group not receiving intravenous hydration.
The AMACING study will be pivotal in deciding the future role of prophylactic intravenous hydration in routine clinical practice. Considering the potential benefits of the results of this study to a large population - who are perhaps burdened with unnecessary and sometimes harmful treatments - and the potential benefits to our health care system in terms of efficiency and costs, we believe the risk for all patients included in this RCT is acceptable. The incidence of CIN is low, CIN itself has no direct relevant clinical implications, and prophylactic intravenous hydration may have negative effects largely disregarded until now. Based on current evidence, therefore, we see no ethical barriers to performing the RCT in our study population. Importantly, we will not include patients that have an eGFR of < 30 ml/min/1.73m2 even though in all probability even these patients will not be at greater risk without intravenous hydration, and we will include only those procedures involving non-ionic low osmolality monomer contrast material.
Inclusion criteria
Exclusion criteria
No repeat inclusion will occur. We restrict our study to procedures where monomeric non-ionic low-osmolar iodinated contrast material is used.