ABC | Volume 111, Nº2, August 2018

Original Article Arq Bras Cardiol. 2018; 111(2):193-202 Miyazaki et al Videodensitometry assessment of AR to guide TAVI Figure 4 – Cumulative frequency curve of the rate of improvement or deterioration of aortic regurgitation by balloon post-dilatation. The formula of the relative delta VD-AR was “(VD-AR after BPD - VD-AR before BPD)/VD-AR before BPD”. Negative values indicate improvements of AR after BPD, whereas positive values stand for deterioration of AR after BPD. Using twenty-fifth percentile of the absolute delta VD-AR, arbitrarily we defined a relative delta of less than -20% as a significant improvement (blue), from -20 to 20% as no change (green), and 20% more as a significant deterioration (orange). Relative delta VD-AR Relative delta VDAR = (VDAR after PBD – VDAR before PBD) / VDAR before PBD Median [IQR] Range –46.2(–75.0,–17.6)% –100% – 40% n = 61 Significant improvement Significant deterioration No change –100 –80 –60 –40 –20 20 40 0 Relative delta VD-AR (%) Cumulative frequency 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Serial change of AR based on Sellers’ grade Before BPD, AR was visually classified as Sellers’ III in 36 patients (59%), and as Sellers’ II in 25 patients (41%). After post-dilatation, there were 3 (5%) cases with Seller’ III, 19 (31%) cases with Sellers’ II, 34 (56%) cases with Sellers’ I and 5 (8%) cases with Sellers’ 0. Out of 36 patients with Sellers’ III before BPD, 34 patients had their Sellers’ grade reduced (to Sellers’ II in 16, Sellers’ I in 17, and Sellers’ 0 in one patient. Out of 25 patients with Sellers’ II before BPD, PVL improved to Sellers’ I in 17 patients and to Sellers’ 0 in 4 patients, deteriorated to Sellers’ III in one patient, and remained unchanged (Sellers’ II) in three patients (Figure 5). Efficacy of BPD Before BPD, VD-AR > 17%, a value that has a prognostic significance in long-term follow-up, was observed in 47 patients (77%). Fourteen cases (23%) had a VD-AR ≤ 17%, eleven (11/14, 79%) were evaluated as Sellers’ II before BPD and 3 (3/14, 21%) as Sellers’ III. After BPD, VD-AR > 17% was observed in 19 patients (falling from 77% to 31% of subjects) – 3 patients (16%) in Seller’s III, 10 patients (53%) in Sellers’ II, and 6 patients in Sellers’ I (32%) (Figure 6). In addition, in these patients with VD-AR > 17%, 7 deaths (34%) occurred during follow-up period, whereas among 42 patients with VD-AR ≤ 17%, 8 patients (19%) died. Predilatationwas performed in 18 patients and had no impact on the reduction of AR assessed by VD-AR. VD-AR was 25.5% (19.5%‑36.0%) with predilatation and 23.0% (16.0%‑29.0%) without predilatation (p = 0.159) before PBD, and 16.5% (9.5%‑22.8%) with predilatation and 11.0% (5.0%‑17.0%) without predilatation (p = 0.106) after PBD. Normalized delta VD-AR was -44.5 (-60.1 – -13.0) with predilatation and -50.0(‑75.0 – -17.9) without predilatation (p = 0.569). Discussion This is the first study to report the value of VD-AR in assessing periprocedural changes in AR. In clinical practice, echocardiogram and aortography are the standard tools to define the device success. As mentioned in the Valve Academic Research Consortium-2 (VARC-2) consensus document, quantitative and semi-quantitative hemodynamic assessment are recommended to assess AR severity by echocardiogram and moderate-to-severe AR is defined as valve failure. 11, 20 Nombela-Franco et al. 8 reported serial changes using semi-quantitative grading based on echocardiogram and showed a reduction of at least 1 degree of AR in 71% of patients. To make a decision whether BPD is needed or not, echocardiogram is an important tool to evaluate the severity of AR. However, we must consider that with the increasing minimalist TAVI approach, the usage of TEE as a guidance of TAVI is becoming unfeasible. Moreover, low inter-observer agreement for the PVL 4-class grading (kappa 0.481) and the 7-class grading (kappa 0.517) has been reported, 21 making a more reliable technique necessary. 18 These facts support the value of aortography with VD assessment as the most practical and objective screening tool to determine the severity of PVL during the procedure. The technique has a median time of execution of 3 minutes. We have previously shown that a VD-AR > 17% correlates with increased mortality and with impaired cardiac reverse remodelling as determined by echocardiography after TAVI with excellent reproducibility. 13,14 This value (VD-AR > 17%) 198