| PartⅠThe predictors of ovarian reserve and ovarian responseIt is very important to make an estimation of ovarian reserve and predict ovarian response before in vitro fertilization-embryo transfer and the controlled ovarian hyperstimulation.Ovarian seserve is defined as the the number and quality of the follicle left in the ovary at any given time. Physicians use some makers to estimate ovarian reserve,including ultrasound imaging(antrol follicle count, ovarian volume, mean ovarian diameter, ovarian vascular flow), endocrine markers and dynamic tests sush as clomiphene citrate challenge test (CCCT), exogenous FSH ORT (EFORT), GnRH stimulation test (G-test). Ovarian response is influenced by many factors, the most important factor is ovarian reserve. Until now, many professors studied the predictors of ovarian reserve, but there is still no suitble prediction model in demostic. In this study, we want to analysis the relationship between patient’s age, antrol follicle count(AFC), serum anti-mullerian hormone(AMH) level, follicle stimulation hormone(FSH) level with ovarian response, and then construct a prediction model for ovarian response. So we can use this model before the treatment of in vitro fertilization-embryo transfer to estimate ovarian reserve and predict ovarian response.OBJECTIVETo screen the optimum predictors of ovarian response by the comparison of patient’s age, antrol follicle count, serum anti-mullerian hormone level, follicle stimulation hormone level, follicle stimulation hormone/luteinizing hormone with ovarian response, and compare the diagnostic utility of them.MATERIALS AND METHODSA total of 331oocyte retrieved cycles in patients conducted in our center between Sep 1,2009 and Dec 31,2010 were analyzed retrospectively in this cohort study. We define ovarian low response as oocytes cellected less than 4 or the pick estradiol lower than 400 pg/ml, and ovarian high response as oocytes cellected more than 17 or the pick estradiol higher than 4000 pg/ml. Three groups were divided according to ovarian response:group normal response, group low response and group high response. Compare the baseline characteristic and the treatment of the patients from different groups using one-way ANOVA and Chi-Square test. Screen the optimum predictors of ovarian response by Bivariate correlate and Logistic Regression. Then compare the diagnostic utility of the predictors by receiver operating curve(ROC).RESULTSThe serum AMH level, AFC are lower, while patient’s age, serum FSH level and FSH/LH are higher in group ovarian low response than group normal response. And the serum AMH level, AFC are higher, while patient’s age, serum FSH level and FSH/LH are lower in group ovarian high response than group normal response, P value are all<0.001. Group low seaponse have higher total Gn dose, lower peak E2 level, follicles aspirated and the number of oocytes retrieved than group normal response. While group high seaponse have lowerer total Gn dose, higher peak E2 level, follicles aspirated and the number of oocytes retrieved than group normal response(P<0.05).By comparison the correlationships between the predictors and eggs collected x 1000 /number of ampoules of Gn, we know that serum AMH level, AFC, serum FSH level, FSH/LH all have significant correlationships with eggs collected x1000/number of ampoules of Gn,r value are 0.655,0.785,-0.541,and-0.521, P value are all< 0.001,.patient’s age is correlated moderatly with eggs collected x1000/number of ampoules of Gn r=-0.494, P<0.001, but patient’s age has nothing to do with it,r=-0.060,P>0.05.3.Using ROC curve for the a diagnose of low ovrian response, the area under(AUC) the curve for AFC, serum AMH level are 0.855 (0.787-0.924)and0.832(0.764-0.900) (P<0.05). The cutoff are AFC≤9, AMH≤1.88ng/ml. The ROC curve for the a diagnose of high ovrian response, AUC is 0.787 (0.728-0.847) for AFC and the cutoff is AFC≥15.4.For the prediction of ovarian low response, AFC, serum AMH level, FSH level are selected in the Logistic regression model, and the mutivariable model can improve the prediction power. But for the prediction of ovarian high response, only AFC is selected in the Logistic regression model, and the mutivariable model can not improve the prediction power.CONCLUSIONS1. AFC, serum AMH level, FSH level, FSH/LH, patient’s age are all predictors of ovarian reserve and ovarian response. 2. AFC is the best single predictor for the ovarian response, the prediction power is stronger than other predictors.3. For ovarian low response, mutivariable predict model can improve the prediction power. But for ovarian high response, mutivariable model can not improve the prediction power.PartⅡThe analyses of ovarian response and clinic outcome of patients with different serum AMH level and AFCOBJECTIVETo construct the prediction model that including the predictors AFC, serum AMH level and FSH level refer to the outcome of part I.MATERIALS AND METHODSA total of 331 oocyte retrieved cycles in patients conducted in our center between Sep 1,2009 and Dec 31,2010 were analyzed retrospectively in this cohort study. Four groups were were divided according to AFC and serum AMH level:group A 1: AFC<9 while AMH≤1.88ng/ml, group A2:AFC≤9 while AMH>1.88ng/ml, group B1:AFC>9 while AMH≤1.88ng/ml, group B2:AFC>9 while AMH>1.88ng/ml. Compare the clinical treatment andovarian response from different groups distinguished by serum FSH level.RESULTS1. Patient’s age and eggs collected are increasing from, serum FSH level, total Gn dose and the ratio of ovarian low response are deceasing group Al to group B2, P <0.001. Serum FSH level and age of group A1 are significantly higher than A2, the total Gn dose are both more than 4000 U, with no significant difference. The ratio of low response was decreasing from group A1 to B2, x2=94.400, P< 0.001.in group Al is 46.2%, higher than 15.0% in group A2.Age in group B1 is higher than in group B2, serum FSH level is not significantly different in the two groups. Egge collected in group B1 is significantly lower than B2, while total Gn dose is significantly higher than B2(P<0.05). The ratio of low response in group B1 is 9.1%, higher than 2.0% in group B2.Eggs collected are not different in goup A2 and B1, total Gn dose in group A2 is significantly higher than Bl(P<0.05). The ratio of low response in group A2 is 15.0%, higher than 9.1% in group B1.2. With a FSH lower than 10.9U/L, total Gn dose decreasing, while eggs collected x1000/number of ampoules of Gn increasing from group Al to B2,F value are 50.413,54.317 and 18;497,P<0.001. Eggs collected is the least in the four groups. Total Gn dose in group A1 and A2 are higher than 3800 U and group Al has the the highest ratio of low response in the four groups,χ2=53.675, P< 0.001.3. With a FSH level higher than 10.9U/L, eggs collected is the highest while total Gn dose is the lowest of group B2 in the four groups, F value are 4.275,4.992 and 13.939, P value were 0.046,0.005 and<0.001. The ratio of low response in group A1 and A2 are 57.1% and 45.5%, higher than the other two groups,χ2=15.856, P=0.015.CONCLUSIONS1. AFC is a better predictor than serum AMH level for ovarian response. 2. For patients whose serum FSH≤10.9 U/L, only when her AFC≤9 and serum AMH level≤1.88 ng/ml should be diagnosed as low ovarian reserve. For patients whose serum FSH>10.9 U/L, with an AFC≤9 or serum AMH level≤1.88 ng/ml should be diagnosed as low ovarian reserve. These patients would bear a high risk of getting low ovarian response. |
PDF Full Text Request