Strategies to reduce CIN incidence first suggest identifying patients at risk and finding for them, when possible, alternative examination tools. They can be identified through the use of questionnaires, the collection of complete anamnesis and the evaluation of glycaemia and, most of all, the assessment of renal function before the CM administration [10] .

The glomerular filtration rate (GFR) through either MDRD or Cockroft-Gault formulas correlates with the renal function better than the serous creatinine [19] - [21] .

The GFR estimate should be obtained within 3 months before a radiological exam with CM in stable day hospital patients, and within 7 days in the hospitalized ones. The measurement should be the latest possible in patients likely to be affected by a recent modification of the renal function (haemodynamically unstable patients, recent chemotherapy, recent use of nephrotoxic drugs, etc.) [21] .

When it is not possible to obtain GFR in due time, for example in Emergency room, the presence of multiple risk factors might indicate patients susceptible of glomerular filtration reduction and therefore renal damage induced by CM [21] .

Patients with a GFR ≥ 60 ml/min have an extremely low CIN risk and, generally, do not require preventive measures or follow-up. Using preventive measures when the GFR is <60 ml/min is considered worthwhile. In particular, several studies suggest that GFR cut-off for CIN risk is 40 - 45 ml/min and efforts to reduce the CIN risk should be concentrated on patients with GFR < 45 ml/min, with a specific attention to patients with a serious renal dysfunction (GFR < 30 ml/min) and additional risk factors [22] [23] .

Drugs that may exacerbate CIN can be divided into three categories:

1) FANS, aminoglycosides, cyclosporine, tacrolimus and amphotericin, which act with a direct nephrotoxic mechanism and which is better to suspend at least 24 hours before CM administration [24] .

2) Some antihypertensive drugs, among which ACE-inhibitors, angiotensin II receptor antagonists and aliskiren interfere with the renin-angiotensin system and decrease the renal perfusion, particularly when it is based on the renin-angiotensin system (in case of hypovolemia, dehydration, heart failure or renal artery stenosis). It has been recently demonstrated that these drugs represent a risk factor for CIN in patients undergoing coronarography [25] . Nevertheless, there is no evidence that the suspension of these drugs is useful for preventing CIN; hence, to date, it is not recommended their routine interruption [21] .

3) Metformin, which itself do not increase the CIN risk, may generate lactic acidosis in case of worsening of renal function and it is cause of acute cell damage. For this reason, it is recommended metformin interruption 48 hours before the CM administration in case of pre-existing renal failure (GFR < 60 ml/min) and restart 48 hours later, after evaluation of renal function [26] . The European Society of Urogenital Radiology (ESUR) adopts a conservative approach and recommends to continue the therapy with metformin until the CM injection in patients with normal serous creatinine levels and to suspend it 48 hours before the injection in patients with a compromised renal function [21] .

Obviously, the risk of CIN onset might be completely avoided obtaining necessary diagnostic data without intravascular administration of iodinated CM [26] . However, in some cases the use of CM is essential but it could be used an inferior dose as the renal toxicity of the iodinated CM is dose-dependent [27] .

The CIN prevalence is connected with the CM volume administered, and it has been observed that the lowest rates of CIN risk affect patients receiving less than 100-140 ml of CM. CM administered in volumes higher than 5 ml/Kg is strongly predictive of a serious acute renal insufficiency, which needs a dialysis treatment [28] . Recent studies found that by using only 50% of normal dose of CM it is possible to acquire appropriate diagnostic images of peripheral arteries, but reducing CIN development in patients at risk [29] .

Moreover, Leheti et al. showed how complications after endovascular aneurism repair appear to be acceptably imaged using only half dose of CM in patients with weight < 90 Kg or BMI < 35 Kg/m² [30] . It has been also demonstrated a significant increase of the CIN risk among patients who received a second CM dose within 48 hours after the administration of the first [31] - [33] .

It should be reduced the CM volume to the bare minimum, avoiding repeated injections by 72 hours. Moreover, it should be used the minimum amount of iodinated CM which allows satisfying image quality, as it could be often diluted with physiological solution without compromising the image quality [33] .

One of the advantages of the use of low-osmolar instead of high-osmolar CM is the reduction of CIN incidence in patients at risk [34] . The mechanisms that determine the CIN risk reduction with low-osmolar CM have not been clarified yet. Possible explanations concern the reduced osmolarity itself, the different ionization, and/or other chemical-physical properties [35] . High-osmolar CM are related to more systemic adverse events, including CIN, than low- or iso-osmolar CM. Therefore, the use of CM with higher osmolarity should be avoided in patients with chronic renal insufficiency. In particular, iso-osmolar CM has been demonstrated to be related to low nephrotoxicity, therefore, it has been widely recommended for patients with renal failure [36] .

Hydration is the only accepted prophylactic strategy for CIN and it is strongly recommended by the Guidelines, as it is effective in reducing the risk and severity of the nephropathy [37] - [39] . All the patients considered at CIN risk should receive hydration. The crystalloids generally used are saline isotonic solution and bicarbonate (154 sodium bicarbonate mmol per 0.85 liter of dextrose at 5%), which are low-cost and harmless for the patients. The possible physiopathological explanation about the hydration efficacy in reducing the CIN risk could be related to the fact that, if correctly completed, it increases the intravascular volume and induces diuresis. Consequently, there is CM dilution in the renal tubules resulting in reduction of its contact time with the tubular epithelium inside the kidney. Moreover, diuresis increase leads to vasodilatation at the level of renal medulla, which is a region more vulnerable to the CM action, probably increasing the prostacyclin production. Furthermore, the volume expansion suppresses the renin-angiotensin system and the anti-diuretic hormone (ADH) production with a vasoconstrictor effect [21] . Particularly for patients with left ventricular dysfunction is important to improve renal blood flow to prevent CIN, as recently demonstrated by Kawatani et al. in a group of patients undergoing endovascular stent graft positioning [40] . Both intravenous and oral hydration have been proposed, however, to the best of our knowledge, there are not strong evidences to consider the oral administration as effective as the intravenous infusion in patients at CIN risk [35] [41] . The most widespread intravenous hydration regime are: 2 ml/Kg/h 2 hours before CM administration and 1 ml/Kg in the following 6 hours; or 1 ml/Kg/h 12 hours before and 12 hours after the procedure [42] .

Pharmacological prophylaxis for preventing CIN would represent the best result to achieve, however, no medication offers a certain efficacy. In particular, several studies showed the possible role of N-Acetylcysteine (NAC) to prevent CIN due to its double role as vasodilator and antioxidant or because its effect on urine alkalinisation [43] - [46] . Numerous protocols have been proposed for NAC use. Actually, CIN incidence has been significantly reduced with oral administration of 600 mg of NAC 24 hours before CM injection instead of hydration alone [47] . Another study proposed intravenous NAC 7 hours instead of 20 min before CM injection to prevent CIN after coron- arography [45] . In a randomized study, Li et al. compared pre-medication with probucol and hydration. They demonstrated that oral intake of 500 mg of probucol twice a day for 3 days before and after coronary intervention procedure was associated with lower serum creatinine levels than the hydration group [48] . Another study showed that administration of 3 g of ascorbic acid before and 2 g after the procedure markedly reduced the CIN incidence [49] . However, high doses of NAC seem to be more effective than ascorbic acid [50] . Lee et al. attested that a short pre-treatment of 2 mmol/L of NAC for 15 min before CM injection and a supplementary dose of NAC 12 hours after reduced CIN more than pre-treatment with probucol or ascorbic acid [51] . Nevertheless, data about the efficacy of NAC and other drugs in reducing the incidence of CIN are divergent, therefore, usefulness remain unproven and their use cannot be recommended [52] [53] .

The sense of haemodialysis in patients at high risk for CIN is the early removal of the CM from the blood [54] . Several studies have been published in order to determine if haemodialysis post-administration of CM reduces CIN rate [55] - [57] . A systematic literature review has highlighted that haemodialysis, even if performed after CM administration, is not effective in reducing CIN rate, probably due to the very early CIN development after CM administration [55] . Moreover, haemodialysis simultaneous with CM administration has not been effective in reducing the damage. In addition, haemodialysis is related to risks connected to the procedure itself, and it can be in some cases nephrotoxic because of phlogosis activation and volume depletion [56] .

The hemofiltration is a continuous form of renal replacement therapy. When solutes and water are removed from blood, fluids are substituted by big volumes of isotonic fluid, which help maintaining the hemodynamic stability [58] . With this procedure, most of the CM is removed from blood with the hemofiltration while the isotonic fluid dilutes the remaining CM [59] .

The hemofiltration is an expensive procedure, which is performed in intensive care. Even though it may be effective for highly selected patients at extremely high risk of CIN, evidence of its efficacy remains poor [60] .