170 INFLUENCE OF FACTORS ASSOCIATED WITH EMBRYO RECOVERY AND POLICY OF TRANSFER ON PREGNANCY RATES AFTER TRANSFER OF BOVINE FRESH EMBRYOS

Abstract

Success of embryo transfer (ET) is related to many well identified factors such as quality and stage of embryos, or parity of recipients. However, there are no data on the effects of factors associated with embryo recovery and ET policy (proportion of embryos transferred as fresh or frozen) on pregnancy rates (PR) after ET. A retrospective study was designed to study how these factors affect PR achieved with fresh embryos (FE). A total of 3380 embryo recovery sessions (RS) realized by Embryotop (France) between 1995 and 2002 were studied, corresponding to 14,755 transfers with FE. Forty-four RS were made with only frozen embryos and removed from data set. A total of 1716 RS involved transfer from both frozen and FE, whereas in 1620 RS all transfers were made with FE. The organization of RS was assessed by the number of operators (OP) per RS and per donor cow, the number of embryos collected per RS and per OP. The operator at RS and ET was also noted. Embryo characteristics such as stage and quality (according to IETS criteria), paternal origin, and breed of donor cow were recorded together with the breed, date of ET and parity of recipients. The association between each variable and PR was analyzed by Chi-square. Only the significant factors were considered for further analysis (P < 0.05) and introduced with a stepwise procedure in a multivariate model of logistic regression to calculate adjusted odds ratio (OR). The number of RS was stable among years (463 to 493), whereas the number of embryos collected increased by 10% from 1995 to 2002. For 681 RS (20.4%), more than one donor cow were used for collection. Mean number of donor cows per RS and OP increased from 1.16 in 1995 to 1.34 in 2002. The mean number of transfers of FE per RS and OP was quite stable (3.82 in 1995 to 4.16 in 2002). This was explained by the increase in the number of frozen embryos per OP, averaging 1.99 in 1995 and 4.14 in 2002. The proportion of RS with only transfers of FE decreased by 20 points from 1995 (57%) to 2002 (36%). The number of transfers of FE per RS and OP was highly variable, ranging from 0 to 26. PR with FE averaged 52.1% and were significantly influenced by the number of donor cows per OP, operator of RS, embryo quality and paternal origin as well as recipient parity. Transfer of FE collected in RS with one donor cow resulted in higher PR than when more than one cow (53.0% v. 49.9%, OR = 1.12, P = 0.003). This could be partly explained by a higher mean number of transfers per RS and OP (1 donor cow = 3.7 ± 2.0 v. > 1 = 5.3 ± 3.1; P < 0.05) as well as a higher proportion of RS with only FE transferred (1 donor cow = 51.2% v. > 1 = 38.5% ; P < 0.05). Although the embryo quality was lower in RS combining ET of both frozen and FE than when only FE were transferred, these effects were not related to the lower quality of FE used in RS involving more than 1 donor cow (1 donor cow = 1.74 ± 0.50 v. > 1 = 1.75 ± 0.44; P > 0.05), but may be explained by a longer time between embryo recovery and ET. After adjustment for the usually well identified sources of variation of PR, the conditions of RS as well as ET policy (leading to the selection of the best embryos for transfer as frozen and to influence on the quantity of work per OP) can significantly influence PR. These sources of variation should be taken into account when analyzing PR results.