From July 2015 to July 2018, 64 patients with aortoiliac occlusive disease who were eligible for endovascular treatment and lacked contraindications to either iodine contrast or carbon dioxide (CO2) were randomized into the carbon dioxide CO2 group(32) or iodine contrast medium(ICM) group(32) were subjected to aortoiliac angioplasty. The primary outcome was the quality of image as the sole contrast agent used in interventions. The secondary outcomes were technical success rate and the safety of procedure.

Sampling The sample size was calculated by the intuitional review board statistician using open epi based on the RCT done by Mendes et al. Power of study as set at 80% and confidence interval set at 95%. The sample size was calculated to be 64 cases, 32 in each group.

Randomization All patients who qualified for the inclusion and exclusion criteria were offered to participate in the trial after explaining all the details involved with the potential benefits and risks of both contrast mediums. Patients who agreed were asked to sign an informed consent. After patient data input was done, randomization was computer generated in the vascular surgery department in Zagazig university hospitals. Computer generated random numbers were created with the use of randomly permuted blocks with two block sizes; after which they were secured in consecutive numbered envelopes and group allocation was independent of time and person delivering the treatment. Single blinding was used to recruit the patients. The patients were randomized into the following two arms: a CO2 group (treatment arm) and an ICM group (control arm) according to the contrast medium selected for the intervention. Group I included 32 patients who were randomized for using CO2 as the contrast medium. Group II: involved 32 patients who were randomized for using iodine contrast medium (ICM).

Postoperative assessment and follow-up The endovascular equipment used in each intervention and the volumes of contrast used were accurately documented for analysis. Immediately after the operation all patients in both arms received intravenous fluids pre and post intervention following a local guideline for renal protection.¬ Patients were discharged the next day and outpatient renal functions were repeated 3 days post procedure, 1 week and 1 month. We analyzed creatinine levels between the two groups during the pre- and post-operative periods. Any elevation of serum creatinine (Scr) of more than 25% or ≥0.5 mg/dl is defined as contrast induced nephropathy (CIN). Patients were followed up for 1 year. One patient from the CO2 group decided to withdraw from the trial during the follow up period and was excluded from the follow up analysis.

Imaging from all cases were analyzed on conclusion of the study. The images were reviewed by two independent observers (radiologists with > 5 years of experience in performing endovascular procedures) blinded to treatment arm. Image quality was assessed using Likert rating scale ranging from 1 (very poor) to 5 (very good) for each image.18 A score of 1 was defined as poor quality with loss of delineation of the vessel. A score of 2 was considered below average with poor delineation of the vessel, a score of 3 is for average quality image and medium delineation of the vessel, a score of 4 was considered good quality with clear delineation of the vessel and a score of 5 was considered very good quality with very clear delineation of the vessel.

Inclusion Criteria:

• Patients with aortoiliac arteries atherosclerotic disease (with arterial atherosclerotic disease classified as Trans-Atlantic Inter-Society Consensus (TASC) A, B and C, (classified by CT angiography)
• Patients with good distal runoff
• Patients suitable for either types of contrast ICM or CO2 (No history of allergy to contrast and normal kidney functions)

Exclusion Criteria:

• Patients with TASC D aortoiliac lesions
• Patients requiring femoral endartectomy
• Patient s with significant multilevel distal disease
• Patients suffering from severe chronic obstructive lung disease, chronic kidney failure, heart failure, or pregnancy.
• Patients younger than 18 years

Patients who agreed were asked to sign an informed consent. After patient data input was done, randomization was computer generated in the vascular surgery department in Zagazig university hospitals. Computer generated random numbers were created with the use of randomly permuted blocks with two block sizes; after which they were secured in consecutive numbered envelopes and group allocation was independent of time and person delivering the treatment. Single blinding was used to recruit the patients.