Anti-Mullerian hormone serum levels predict response to controlled

Document Sample
Anti-Mullerian hormone serum levels predict response to controlled Powered By Docstoc
					Anti-Mullerian hormone serum levels predict response to controlled
ovarian hyperstimulation in oocyte donors
R.Riggs1, S. Bocca1, L.Yin2 V. Bourque2, B.Leader2 S.Oehninger1,
L Stadtmauer1
1
 The Jones Institute for Reproductive Medicine/EVMS, Obstetrics and Gynecology, Norfolk,
U.S.A.;
2
 Repromedix Corporation, Repromedix Corporation, Woburn, U.S.A.;

         Introduction: Traditional markers for ovarian reserve such as FSH, age, antral follicle
count, estradiol (E2), and inhibin B (IB) are user-dependent or provide limited information.
                                      B




Emerging trials utilizing anti-Müllerian hormone (AMH) in infertility populations demonstrate
predictive ability but no study has assessed the predictive ability of AMH in a donor egg (fertile)
population. The objective of this study was to evaluate the predictive value of AMH serum levels
as a marker of ovarian responsiveness in a donor egg program.

         Materials and Methods: Frozen day 3 serum samples from consecutive oocyte donors
from March 2002 to August 2007 were evaluated for AMH, FSH, and IB hormone levels. A
                                                                        B




retrospective chart review was completed to extract outcome data. Patients with day 3 E2 levels
greater than 100 ng/mL were excluded. The majority of patients (96%) were down-regulated with
luteal leuprolide acetate and birth control pills. Remaining patients underwent a ganirelix acetate
protocol. Gonadotropins were initiated and adjusted based on ultrasound and E2 levels. When at
least two follicles were ≥18mm, hCG was administered and all follicles greater than 10mm were
transvaginally aspirated. Main outcomes were the total number of oocytes retrieved, the number
of mature oocytes, and peak serum E2 levels.

         Results: 168 cycles (78 patients) were evaluated. Cycle day 3 baseline characteristics
included an average age of 27±3 years, BMI of 24±4.4 kg/m2, E2 of 38±16 pg/mL, FSH of 6±1
IU/L, IB of 97±41 pg/mL and AMH of 3.4±2.5 ng/mL. Cycle outcomes were: number of mature
          B




(metaphase II) oocytes of 14±7, average number of total oocytes retrieved of 17±8, and peak E2
levels of 4551±2768 pg/mL.
         Correlation analysis demonstrated AMH had the highest correlation value with the total
number of oocytes retrieved (r=0.360, P≤0.01) relative to age (r=-0.047, P=.549), FSH (r=-0.305,
P≤0.01), and IB (r=0.242, P≤0.01). Correlation analysis demonstrated AMH and FSH had the
               B




highest correlation value with number of mature (metaphase II) oocytes retrieved (r=0.283,
P≤0.01; r=-0.291, P≤0.01, respectively) relative to age (r=-0.087, P=0.267), and IB (r=0.207,
                                                                                    B




P≤0.01). Lastly, correlation analysis demonstrated AMH had the highest correlation value with
peak E2 (r=0.316, P≤0.01) relative to age (r=0.022, P=0.783), FSH (r=-0.115, P=0.143), and IB   B




(r=0.158, P≤0.05).
         Receiver operating characteristic curve (ROC) analysis for poor response (≤ 6 total
oocytes) demonstrated that AMH had the largest area under the curve (0.852) relative to IB  B




(0.728), age (0.701), and FSH (0.555). Pair wise comparison of the AUC for AMH, FSH, IB, andB




age demonstrated a significant difference when compared to FSH but not IB or age. For the
                                                                                B




prediction of hyper-response (≥ 15 oocytes) AMH generated the largest AUC (0.680) relative to IB      B




(0.575), age (0.502), and FSH (0.636). Pair wise comparison of the AUC for AMH, FSH, IB, andB




age demonstrated a significant difference when compared to age and IB but not FSH. Using an
                                                                            B




AMH cut-off value of 1.45 ng/mL for poor response yielded a negative predictive value (NPV) of
98% and a positive predictive value (PPV) of 30%. Using an AMH cut-off value of 2.67 ng/mL for
hyper-response yielded a NPV of 96% and a PPV of 20%.

        Conclusions: AMH correlated best with the number of total oocytes retrieved and peak
estradiol levels compared to FSH, age, and IB. AMH and FSH correlated the best with the
                                              B




number of mature oocytes retrieved relative to age and IB. The predictive value of AMH for poor
                                                          B




response and hyper-response paralleled or exceeded conventional analyte testing using FSH, E2,
and IB. These data represent the first application of AMH as a predictive hormone in a fertile
      B




population and imply AMH may be successfully utilized as a screening test in an oocyte donation
population.