In this study, 7 patients (8.6%) undergoing TF-TAVI for severe aortic stenosis developed AKI; 6 (7.4%) had stage 1 AKI and one (1.2%) had stage 3 AKI according to the VARC-2 criteria. No patient needed renal replacement therapy. Risk factors for AKI included the SCr, eGFR, CMV, and CMV × SCr/BW value. A CMV × SCr/BW value > 2.99 could be considered the threshold value for prediction of AKI during TF-TAVI and for intervention.
In 2012, the endpoint definitions in the VARC-2 consensus document were revised  to include a recommendation for the Acute Kidney Injury Network criteria to add urine output in the definition of AKI and the timing for diagnosis of postoperative AKI was extended from 72 hours to 7 days. A recent meta-analysis identified New York Heart Association functional class IV, previous CKD, requirement for red blood cell transfusion, previous peripheral artery disease, and a TA approach as strong risk factors for AKI after TAVI according to the VARC-2 definition . Furthermore, the rate of AKI was higher in patients who underwent TA-TAVI than in those who underwent TF-TAVI, as reported previously [34-38]. Another study identified major bleeding to be an important risk factor for AKI and to have a significant impact on outcomes .
Our rate of AKI following TAVI is consistent with the observations of Keles et al  and Konigstein et al , who found rates of 7.1% and 16.7% in 70 and 300 patients, respectively. The definition of AKI used (VARC-2 criteria) and the ratio of TF procedures (92.9% and 98%, respectively) in those studies were very similar to those in our study. Consistent with the previous research, we found that the baseline SCr, eGFR, CMV, and CMV × SCr/BW value predicted AKI. Elhmidi et al  and Seiffert et al  identified a correlation between baseline renal function and incidence of AKI after TAVI. Furthermore, Van Linden et al  reported administration of a greater amount of contrast medium to be an independent risk factor for AKI, while Yamamoto et al  identified a relationship between an increment in the dose of contrast medium and an increased prevalence of AKI in their series of 415 consecutive patients who underwent TF-TAVI.
The CMV × SCr/BW value has been established as a criterion for prevention of CIN, and a value >5.0 was shown to predict post-procedural AKI after percutaneous coronary intervention [11-14]. In our study, a CMV × SCr/BW value >2.99 was a risk factor for AKI after TAVI. The reason why our threshold CMV × SCr/BW value was smaller than in an earlier coronary angiography study  may lie in the difference in the diagnostic criteria used for AKI and CIN. The diagnostic criterion for CIN is an increase in SCr of >0.5 mg/dl or an increase of >25% from baseline in the 48–72 hours following administration of contrast medium . However, the definition of AKI in the VARC-2 document is based on the Acute Kidney Injury Network classification, i.e., only a slight increase in SCr of 0.3 mg/dl is needed to diagnose the onset of AKI in patients with normal renal function.
In this study, we excluded patients who underwent TA-TAVI and those who had massive bleeding or failure of circulatory dynamics intraoperatively. For the first time, it was possible to identify CMV as a risk factor for AKI after TAVI. Furthermore, although the CMV used was smaller than that in previous studies, the incidence of AKI was comparable. Therefore, the CMV should be determined by taking not only the absolute amount into consideration, but also renal function and BW. Therefore, renal function and body weight should be taken into account when determining the CMV in, for example, an elderly patient with a small body habitus, which is common in Asian populations. By using the CMV × SCr/BW formula, it is possible to determine the maximum amount of contrast medium that can be used on a case-by-case basis according to preoperative renal function and BW. By determining the maximum dose of the contrast agent, the risk of AKI can decreased by limiting the CMV, the dilution factor, and the type of contrast agent used. Our findings suggest that avoidance of major complications and reducing the CMV decreases the risk of AKI after TF-TAVI.
This study has several limitations. First, it was small and had a retrospective observational design, so our results may have been affected by unknown cofounders. Second, we did not apply diagnostic criteria based on urine output. Therefore, it is possible that the number of cases of AKI was underestimated. Third, the long-term renal function and outcomes in patients with AKI were not investigated. In a previous study, even a small increase in the baseline creatinine level after TAVI was associated with a worse outcome . The poor prognosis in our patients should encourage better patient selection and management for prevention of AKI.