**9. Preliminary results**

From 2003 to 2009, 187 patients underwent Remodeling with subvalvular ring annuloplasty in 14 centers (24 surgeons). Preoperative AI≥grade 2 was present in 67.9% (127), and bicuspid valve in 20.8% (39). Need for cusp repair was assessed according to 3 strategies: Group 1: gross visual estimation (74 patients), Group 2: alignment of cusp free edges (62 patients), Group 3: two-steps approach associating alignment of cusp free edges with effective height resuspension (51 patients). A composite outcome was defined as recurrence of aortic insufficiency ≥ grade 2 and/or reoperation.

Valve sparing was successful in all but 2 cases. Operative mortality was 3.2% (6). Ring produced a significant annular base reduction from 27.3±0.6 to 20.5±1.7 mm (p<0.01) without significant mean transvalvular gradient (7.2±2.9 mmHg). Treatment of cusp lesion was most frequently performed in Group 3 (70.6%, versus 20.3% Group 1 and 30.6% Group 2, p<0.001). Nine patients required reoperation during follow-up (31.4±4.5 months (1-80)). At one year, no patients in Group 3 presented with composite outcome events (versus 28.1% Group 1, 15% Group 2, p<0.001). Residual aortic insufficiency and tricuspid anatomy were independent risk factors for composite outcome in Group 1 and 2. Annulus diameter, Marfan syndrome and cusp repair had no effect on aortic insufficiency recurrence or reoperation. Repair of bicuspid valves showed better results than tricuspid valves anatomy. This might be partly explained by higher rate of cusp repair in bicuspid patients (72% of patients versus 28% of patients for tricuspid valves). Apart from high rate of cusp repair, better results in Group 3 also could reflect a learning curve or other confounding factors, since these patients were operated later (Lansac et al., 2010a, 2010b).
