Can MR measurement of renal artery flow and renal volume predict the outcome of percutaneous transluminal renal angioplasty?

Purpose: Predicting therapeutic benefit from percutaneous transluminal rena l angioplasty (PTRA) in patients with renal artery stenosis (RAS) remains d ifficult. This study investigates whether magnetic resonance (MR)-based ren al artery flow measurements relative to renal parenchymal volume can predic t clinical outcome following PTRA. Methods: The data on 23 patients (13 men, 10 women; age range 47-82 years, mean age 64 years) were analyzed. The indication for treatment was hyperten sion (n = 18) or renal insufficiency (n = 5). Thirty-four cases of RAS were identified: bilateral disease was manifest in 11 and unilateral disease in 12 patients. The MR imaging protocol included a breath-hold, cardiac-gated cine phase-contrast sequence for renal flow measurement and a fast multipl anar spoiled gradient-echo sequence for renal volume measurement. MR measur ements were performed on the day prior to and the day following PTRA. Clini cal success was defined as (a) a reduction in diastolic blood pressure > 15 % or (b) a reduction in serum creatinine > 20%. Kidneys were categorized as normal volume or low volume. A renal flow index (RFI) was calculated by di viding the renal flow (ml/min) by the renal volume (cm(3)). Results: Clinical success was observed in 11 patients. Twelve patients did not benefit from angioplasty. Normal kidney volume was seen in 10 of 11 res ponders and in 8 of 12 nonresponders, resulting in a sensitivity of 91%, sp ecificity of 33%, a positive predictive value (PPV) of 56% and a negative p redictive value (NPV) of 80%. A RFI below a threshold of 1.5 ml/min/cm(3) p redicted successful outcome with 100% sensitivity, 33% specificity, 58% PPV , and 100% NPV. The combination of normal renal volume and a RFI below 1.5 ml/min/cm(3) identified PTRA responders with a sensitivity of 91%, a specif icity of 67%, a PPV of 71%, and a NPV of 89%. PTRA resulted in a greater in crease in renal flow in responders compared with nonresponders (p < 0.001). Conclusion: A combination of cine phase-contrast MR renal flow and parenchy mal volume measurements enables identification of patients benefiting from PTRA with a high sensitivity and NPV, but only moderate specificity and PPV .