ABC | Volume 112, Nº3, March 2019

Original Article Kiyose et al A systematic review for valvular prostheses Arq Bras Cardiol. 2019; 112(3):292-301 Table 1 – Characteristics of included studies Trials Year of publication Total randomised Type of valves Number Randomised Patients Characteristics Local of prosthesis implantation Follow-up (m/y) Vallejo 1981 110 Bioprosthesis: Angell-Shiley 38 7% NYHA II; 27% NYHA III; 4% NYHA IV 66% Male, Mean age: 39.7 ± 11 .2 MVR Mean 24.13 ±11.16 m Mechanical prosthesis *: Bjork-Shiley 35 7% NYHA II; 24% NYHA III; 4% NYHA IV 69% Male, 40.7 ± 11.3 MVR Mean 31.61 ± 13.02 m Lillehei-Kaster 37 4% NYHA II; 30% NYHA III; 3% NYHA IV 76% Male, 41.9 ± 10.4 MVR Mean 30.4 ± 15.9 m Veterans Affairs (Hammermeister) 2000 575 Bioprosthesis: Hancock porcine 289 100% Male 67% AVR; 33% MVR Maximum 18 y Mechanical prosthesis: Bjork-Shiley 286 100% Male 69% AVR; 31% MVR Maximum 18 y Edinburgh (Oxenham, Bloomfield) 2003 533 Bioprosthesis: Hancock porcine Carpentier- Edwards 107 159 53% NYHA III or IV AF † 76% Female mitral valve 38% AVR, 50% MVR, 12% AVR+MVR Mean 20.4 y Mechanical prosthesis: Bjork-Shiley 267 57% NYHA III or IV 74% Female mitral valve 41% AVR, 48% MVR, 11% AVR+MVR Mean 20.4 y Stassano 2009 310 Bioprosthesis: Carpentier- Edwards SAV 93 75.5% NYHA III or IV Male 50.3% Age 63.5 ± 3.9 100% AVR Mean 106 ± 28 m Carpentier- Edwards Pericardial 62 Mechanical prosthesis: St. Jude Medical 107 76,8% NYHA III or IV Male 42,5% Age 64.0 ± 7.6 Carbomedics 48 *Tilting disc valve. 37.8% previous surgery in mitral valve with LK (p < 0.005); † 67% Bioprosthesis in atrial fibrillation. The need for reoperation was more frequent among patients who received biological prostheses than among those who received mechanical prostheses (RR = 3.60; 95% CI 2.44-5.32; I 2 = 0%). The effect was similar in patients who received prostheses in the aortic or mitral position or both simultaneously (Figure 3). The risk of bleeding was lower in patients treated with biological prostheses than in those treated with mechanical prostheses (RR=0.64; 95%CI 0.52-0.78; I 2  =0%). There was a trend toward a distinct effect between the subgroups according to the position of the implant, but that was not statistically significant (P for subgroup differences = 0.09) (Figure 3). It should be noted that the definitions of bleeding were not equal across studies. Vallejo et al. 13 considered only bleeding that required hospitalisation or that was a direct cause of death. 13 In their study, Bloomfield et al. 20 included all major (65%) and minor bleeding. 20 The Veterans Affairs study included clinically important bleeding. 17 Stassano et al. 21 made no reference to the magnitude of the bleeding. 21 There were no significant differences in the risk of endocarditis (RR = 1.21, 95% CI 0.78-1.88; I 2  = 4%) or systemic arterial embolism (RR = 0.93, 95% CI 0.66-1.31; I 2  = 31%) between the group that received bioprostheses and the group that received mechanical prostheses (Figure 4). Discussion This systematic review and meta-analysis of randomized studies involving patients requiring cardiac valve replacement revealed similar mortality between patients who underwent implantation of biological prostheses and those who underwent implantation of mechanical prostheses. There were also no differences regarding the risk of thromboembolism and endocarditis. However, the risk of bleeding was approximately one third lower among patients treated with biological prostheses than in those treated with mechanical prostheses. In contrast, the need for reoperation among patients treated with bioprostheses was more than three times greater that of patients treated with mechanical prostheses. 295