Presence of hydrosalpinx correlated to endometrial inflammatory response in vivo
2
the antral follicle count can be used as an accurate marker of ovarian reserve. Supported by: None. Monday, October 14, 2002 11:35 A.M. O-8 Inhibition of EMP2 expression prevents mouse blastocyst attachment in vitro and implantation in vivo. Molina B. Dayal, Madhuri Wadehra, Teri Ord, Jonathan Braun, Carmen J. Williams. Univ of Pennsylvania, Phila- delphia, PA; UCLA, Los Angeles, CA. Objective: Understanding the molecular mechanisms of implantation is critical for advancing contraceptive and infertility therapies. We have pre- viously shown that epithelial membrane protein (EMP2), a tetraspan pro- tein, is expressed in the uterus and has peak expression on the uterine epithelial surface at the time of implantation. The goal of this investigation was to determine if altering uterine EMP2 expression prevents blastocyst attachment and implantation. Design: Laboratory investigation. Materials/Methods: Ribozyme-mediated mRNA destruction was chosen as the method for inhibiting EMP2 protein expression both in vitro and in vivo. The ribozyme plasmid targeting EMP2 (RIBO-pMSCV) or control plasmid (pMSCV) were stably transfected into Hec1A (human endometrial carcinoma) cell lines. The cells were grown to confluence in 24 well plates, and single, hatched mouse blastocysts were placed into each well. Blasto- cyst attachment to the cells was examined three days later using light microscopy. In vivo implantation experiments were performed using gona- dotropin-stimulated, mated female mice. One day after mating, one uterine horn was transfected with RIBO-pMSCV, and the contralateral horn was transfected with pMSCV. Mice were euthanized 5 days post-coitum (normal time of implantation) for analysis of EMP2 expression by immunohisto- chemistry (IHC), and 8 days post-coitum for visual counting of implantation sites. Results: The RIBO-pMSCV Hec1A cells expressed significantly less EMP2 protein than the pMSCV Hec1A cells as shown by immunoblotting and immunocytochemistry experiments. Only 3/68 (4%) of blastocysts attached to the RIBO-pMSCV Hec1A cells, while 31/74 (42%) attached to the control, pMSCV Hec1A cells (p-value 0.001). In vivo transfections showed that inhibition of EMP2 expression occurred in a dose-dependent manner. IHC of uterine horns exposed to 20 mcg RIBO-pMSCV resulted in near complete loss of EMP2 expression on the epithelial surface by day 5 post-coitum. By day 8 post-coitum, RIBO-pMSCV-transfected horns had significantly fewer implantation sites than control horns in the same mouse (all p-values 0.01) at all doses examined (see Table). Mean Number of Implantation Sites. Transfection Dose 0 mcg (n 16) 5 mcg (n 10) 10 mcg (n 10) 20 mcg (n 10) 30 mcg (n 3) pMSCV 9.8 7.8 5.5 5.8 10.3 RIBO-pMSCV 9.8 3.7 1.6 0 0 # of implantation sites in pMSCV treated horns and RIBO-pMSCV treated horns are significantly different at all doses (all p-values 0.01). Conclusions: These studies demonstrate that EMP2 expression appears to be required for implantation in the mouse. The molecular mechanism by which EMP2 expression impacts implantation is unknown. EMP2 itself may serve as an adhesion molecule that interacts with the blastocyst. Alterna- tively, EMP2 may regulate expression of other molecules on the surface of the uterine epithelium critical for implantation. Supported by: Mellon Foundation, NIH R03 HD40376 (CJW); Jonsson Comprehensive Cancer Center, UCLA Gene Therapy Program, Ruzic Foundation (MW,JB). ART: CLINICAL ART I Monday, October 14, 2002 2:00 P.M. O-9 Do time from retrieval to transfer into IVF media and blood in follic- ular aspirates affect oocyte fertilization? H. Lee Higdon III, Martin M. Crane IV, Dawn W. Blackhurst, Jane E. Johnson, William R. Boone. Greenville Hosp System, Greenville, SC. Objective: Time from oocyte retrieval until placement in IVF culture media can vary substantially, sometimes by an order of magnitude, depend- ing on the number of oocytes retrieved. In some cases, we have observed blood in the follicular aspirate. The purpose of this study is to determine if elapsed time, presence of blood, or maternal age affect the likelihood that an oocyte will fertilize. Design: A retrospective analysis of assisted reproductive technology (ART) data from a cohort of 82 women presenting for IVF to a hospital based reproductive endocrinology practice in Greenville, SC. Materials/Methods: Oocytes were retrieved by standard methods using a double lumen needle. For each oocyte retrieved, time elapsed between retrieval and placement in culture media, presence of blood in the follicular aspirate, and age of the patient were noted. Logistic regression was used to estimate the odds of fertilization, adjusting for maternal age, time, blood, and a time by blood interaction. Two statistical methods were employed: a conventional analysis using SAS software, which treated each oocyte as an independent observation, and SUDAAN software which adjusted for the effect of clustering within patient. Results: There were 82 patients who contributed 1093 oocytes (median 12, range 1 to 38). In models with time elapsed and blood alone, the coefficient (standard error; SE) for time was 0.04 (0.01) and statistically significant in the SAS analyses (P 0.003), but was 0.04 (0.03) in the SUDAAN analysis (P 0.21). After adjusting for maternal age, the corresponding coefficients (SE) were 0.04 (0.01) and 0.04 (0.03) respec- tively. The coefficient (SE) for age in the SAS model was negative 0.03 (0.02; P 0.09), but was negative 0.03 (0.05) in the SUDAAN model (P 0.49). Neither presence of blood nor an interaction between time elapsed and blood were statistically significant in any of the analyses. Conclusions: Time elapsed, presence of blood, and age did not predict oocyte fertilization in these data although time and age would have appeared to be statistically significant using the conventional analytical approach of independent observations. Adjusting for clustering within patient adds an extra component to the variance which increases the standard error and thus, reduces the statistical power to detect a given difference at a fixed sample size. This analysis emphasizes the importance of appropriate statistical methods when analyzing IVF and other clustered reproductive data. Supported by: None. Monday, October 14, 2002 2:15 P.M. O-10 Presence of hydrosalpinx correlated to endometrial inflammatory re- sponse in vivo. Valerie Wells, Tamara Kalir, Tanmoy Mukherjee, Alan Barry Copperman. Mount Sinai Medical Ctr, New York, NY. Objective: The toxic effects of hydrosalpinx fluid on in vitro murine embryogenesis has been demonstrated by our group in the past and verified by others. It is possible that the deleterious effect of hydrosalpinx fluid may not be limited to direct embryotoxicity, and may, in fact, be the result of an induced inflammatory response within the endometrial cavity. By evaluat- ing inflammatory mediators and by analyzing the number and type of inflammatory cells in the endometrium of patients with hydrosalpinges, we aim to determine whether there is an endometrial contribution to the compromised reproductive outcome in these patients. Design: A retrospective case control study was performed after obtaining IRB approval. The Mount Sinai Medical Center CoPath database was used to identify hysterectomy samples from 1998 –2001 Materials/Methods: 30 cases were identified that demonstrated hydrosal- pinges or salpingitis and 30 age matched controls were identified with normal tubal architecture. Archived H&E slides were review to confirm the S4 Abstracts Vol. 78, No. 3, Suppl. 1, September 2002
-
Upload
valerie-wells -
Category
Documents
-
view
212 -
download
0
Transcript of Presence of hydrosalpinx correlated to endometrial inflammatory response in vivo
the antral follicle count can be used as an accurate marker of ovarianreserve.
Supported by: None.
Monday, October 14, 200211:35 A.M.
O-8
Inhibition of EMP2 expression prevents mouse blastocyst attachment invitro and implantation in vivo. Molina B. Dayal, Madhuri Wadehra, TeriOrd, Jonathan Braun, Carmen J. Williams. Univ of Pennsylvania, Phila-delphia, PA; UCLA, Los Angeles, CA.
Objective: Understanding the molecular mechanisms of implantation iscritical for advancing contraceptive and infertility therapies. We have pre-viously shown that epithelial membrane protein (EMP2), a tetraspan pro-tein, is expressed in the uterus and has peak expression on the uterineepithelial surface at the time of implantation. The goal of this investigationwas to determine if altering uterine EMP2 expression prevents blastocystattachment and implantation.
Design: Laboratory investigation.Materials/Methods: Ribozyme-mediated mRNA destruction was chosen
as the method for inhibiting EMP2 protein expression both in vitro and invivo. The ribozyme plasmid targeting EMP2 (RIBO-pMSCV) or controlplasmid (pMSCV) were stably transfected into Hec1A (human endometrialcarcinoma) cell lines. The cells were grown to confluence in 24 well plates,and single, hatched mouse blastocysts were placed into each well. Blasto-cyst attachment to the cells was examined three days later using lightmicroscopy. In vivo implantation experiments were performed using gona-dotropin-stimulated, mated female mice. One day after mating, one uterinehorn was transfected with RIBO-pMSCV, and the contralateral horn wastransfected with pMSCV. Mice were euthanized 5 days post-coitum (normaltime of implantation) for analysis of EMP2 expression by immunohisto-chemistry (IHC), and 8 days post-coitum for visual counting of implantationsites.
Results: The RIBO-pMSCV Hec1A cells expressed significantly lessEMP2 protein than the pMSCV Hec1A cells as shown by immunoblottingand immunocytochemistry experiments. Only 3/68 (4%) of blastocystsattached to the RIBO-pMSCV Hec1A cells, while 31/74 (42%) attached tothe control, pMSCV Hec1A cells (p-value �0.001). In vivo transfectionsshowed that inhibition of EMP2 expression occurred in a dose-dependentmanner. IHC of uterine horns exposed to 20 mcg RIBO-pMSCV resulted innear complete loss of EMP2 expression on the epithelial surface by day 5post-coitum. By day 8 post-coitum, RIBO-pMSCV-transfected horns hadsignificantly fewer implantation sites than control horns in the same mouse(all p-values �0.01) at all doses examined (see Table).
Mean Number of Implantation Sites.
Transfection Dose0 mcg
(n � 16)5 mcg
(n � 10)10 mcg
(n � 10)20 mcg
(n � 10)30 mcg(n � 3)
pMSCV 9.8 7.8 5.5 5.8 10.3RIBO-pMSCV 9.8 3.7 1.6 0 0
# of implantation sites in pMSCV treated horns and RIBO-pMSCV treatedhorns are significantly different at all doses (all p-values � 0.01).
Conclusions: These studies demonstrate that EMP2 expression appears tobe required for implantation in the mouse. The molecular mechanism bywhich EMP2 expression impacts implantation is unknown. EMP2 itself mayserve as an adhesion molecule that interacts with the blastocyst. Alterna-tively, EMP2 may regulate expression of other molecules on the surface ofthe uterine epithelium critical for implantation.
Supported by: Mellon Foundation, NIH R03 HD40376 (CJW); JonssonComprehensive Cancer Center, UCLA Gene Therapy Program, RuzicFoundation (MW,JB).
ART: CLINICAL ART I
Monday, October 14, 20022:00 P.M.
O-9
Do time from retrieval to transfer into IVF media and blood in follic-ular aspirates affect oocyte fertilization? H. Lee Higdon III, Martin M.Crane IV, Dawn W. Blackhurst, Jane E. Johnson, William R. Boone.Greenville Hosp System, Greenville, SC.
Objective: Time from oocyte retrieval until placement in IVF culturemedia can vary substantially, sometimes by an order of magnitude, depend-ing on the number of oocytes retrieved. In some cases, we have observedblood in the follicular aspirate. The purpose of this study is to determine ifelapsed time, presence of blood, or maternal age affect the likelihood that anoocyte will fertilize.
Design: A retrospective analysis of assisted reproductive technology(ART) data from a cohort of 82 women presenting for IVF to a hospitalbased reproductive endocrinology practice in Greenville, SC.
Materials/Methods: Oocytes were retrieved by standard methods using adouble lumen needle. For each oocyte retrieved, time elapsed betweenretrieval and placement in culture media, presence of blood in the follicularaspirate, and age of the patient were noted. Logistic regression was used toestimate the odds of fertilization, adjusting for maternal age, time, blood,and a time by blood interaction. Two statistical methods were employed: aconventional analysis using SAS software, which treated each oocyte as anindependent observation, and SUDAAN software which adjusted for theeffect of clustering within patient.
Results: There were 82 patients who contributed 1093 oocytes (median �12, range 1 to 38). In models with time elapsed and blood alone, thecoefficient (standard error; SE) for time was 0.04 (0.01) and statisticallysignificant in the SAS analyses (P � 0.003), but was 0.04 (0.03) in theSUDAAN analysis (P � 0.21). After adjusting for maternal age, thecorresponding coefficients (SE) were 0.04 (0.01) and 0.04 (0.03) respec-tively. The coefficient (SE) for age in the SAS model was negative 0.03(0.02; P � 0.09), but was negative 0.03 (0.05) in the SUDAAN model (P �0.49). Neither presence of blood nor an interaction between time elapsedand blood were statistically significant in any of the analyses.
Conclusions: Time elapsed, presence of blood, and age did not predictoocyte fertilization in these data although time and age would have appearedto be statistically significant using the conventional analytical approach ofindependent observations. Adjusting for clustering within patient adds anextra component to the variance which increases the standard error and thus,reduces the statistical power to detect a given difference at a fixed samplesize. This analysis emphasizes the importance of appropriate statisticalmethods when analyzing IVF and other clustered reproductive data.
Supported by: None.
Monday, October 14, 20022:15 P.M.
O-10
Presence of hydrosalpinx correlated to endometrial inflammatory re-sponse in vivo. Valerie Wells, Tamara Kalir, Tanmoy Mukherjee, AlanBarry Copperman. Mount Sinai Medical Ctr, New York, NY.
Objective: The toxic effects of hydrosalpinx fluid on in vitro murineembryogenesis has been demonstrated by our group in the past and verifiedby others. It is possible that the deleterious effect of hydrosalpinx fluid maynot be limited to direct embryotoxicity, and may, in fact, be the result of aninduced inflammatory response within the endometrial cavity. By evaluat-ing inflammatory mediators and by analyzing the number and type ofinflammatory cells in the endometrium of patients with hydrosalpinges, weaim to determine whether there is an endometrial contribution to thecompromised reproductive outcome in these patients.
Design: A retrospective case control study was performed after obtainingIRB approval. The Mount Sinai Medical Center CoPath database was usedto identify hysterectomy samples from 1998–2001
Materials/Methods: 30 cases were identified that demonstrated hydrosal-pinges or salpingitis and 30 age matched controls were identified withnormal tubal architecture. Archived H&E slides were review to confirm the
S4 Abstracts Vol. 78, No. 3, Suppl. 1, September 2002
original pathologic diagnosis of the fallopian tube and the endometrium.Two observers performed the analysis of slides for leukocytes. Five repre-sentative adjacent non-overlapping high power fields were examined foreach slide of tube and endometrium. Specimens were then labeled with ahuman monoclonal antibody using standard immunohistochemical tech-niques. Stained slides were then evaluated using light microscopy. Positivesamples showed granular membrane or cytoplasmic staining. Samples wereevaluated for the extent of staining and the intensity of staining (none, low,high) according to standard methods.
Results: Examination of the H&E slides with hydrosalpinx for inflam-matory cells demonstrated a statistically significant increase in the overallnumber of cells among the cases (median 70, range 1–525) vs controls(median 22.5, range 1–89; p � 0.005). There was a statistically significantincrease in the number of neutrophils (p �0.0001); lymphocytes (p �0.002); plasma cells (p �0.0001); basophils (p � 0.01); macrophages (p �0.006). There was no difference in the number of eosinophils. Examinationof the H&E slides of the enodmetrium demonstrated a statistically signifi-cant increase in the overall number of inflammatory cells in cases (median36, range 7–559) versus controls (median 16, range 0–119, p � 0.03). Therewas a statistically significant difference in the number of neutrophils (p �0.0001) and basophils (p � 0.02). There was no difference in the numbersof lymphocytes, plasma cells, eosinophils, or macrophages. Immunohisto-chemical staining for IL-2 demonstrated 7.4% of controls versus 65% ofcases with high intensity staining. 67% of controls versus 20% of casesdemonstrated low intensity staining. 15% of cases versus 26% of controlsdemonstrated no staining.
Conclusions: We have demonstrated for the first time that there is adefined, identifiable, local response in the endometrium to hydrosalpingealfluid. This response consists of statistically significant elevations of leuko-cytes in the endometrium. There is a trend toward increased levels of theinflammatory cytokine IL-2 in the endometrium among the patients withhydrosalpinges. An inflammatory endometrial response may be an indepen-dent contributor to the decreased reproductive outcome seen in patients withhydrosalpinges.
Supported by: self-funded.
Monday, October 14, 20022:30 P.M.
O-11
Increased hyaluronan concentration in the embryo transfer mediumresults in a significant increase in human embryo implantation rate.William Schoolcraft, Michelle Lane, John Stevens, David K. Gardner.Colorado Ctr for Reproductive Medicine, Englewood, CO.
Objective: The levels of hyaluronan, a glycosaminoglycan, in the femalereproductive tract increase up to the time of implantation. Previous studiesusing the mouse model have revealed that the presence of hyaluronan in theembryo transfer medium significantly increases resultant implantation rates.The aim of this study was therefore to determine the effect of hyaluronanpresent in human embryo transfer medium on subsequent implantation rate(fetal heart beat, FHB).
Design: Prospective randomized clinical trial.Materials/Methods: IVF patients and oocyte donors were recruited into a
prospective randomized trial following informed consent from January 2001to February 2002. There were no differences between the two groups ofpatients with respect to age, day 3 FSH, number of oocytes, the number ofpronucleate embryos or the number of embryos transferred. Patients wereallocated into either arm of the trial using a computer generated random-ization sheet. Following fertilization, embryos were cultured for 48h inmedium G1(GIII Series) in an environment of 6%CO2 and 5%O2 in groupsof up to 4 in 50 l drops. At the time of transfer on day 3 embryos were eithertransferred in medium G2 (GIII Series, containing 0.125mg/ml hyaluronan)(control), or in medium G2 containing 0.5mg/ml hyaluronan (EmbryoGlue).Embryo transfer was performed using ultrasound guidance and a Wallacecatheter employing a continuous column of transfer medium. Embryos weretransferred in 30 l of medium.
Results: The number of embryos transferred was 3.9 for both groups ofIVF patients, and was 3.3 and 3.2 for the control and EmbryoGlue groupsfor oocyte donors. There was a significant increase in implantation rate(FHB) in the EmbryoGlue arm of the trial for both IVF patients (29.0%, n �73 patients) and oocyte donors (45.8%, n � 18 patients) compared to their
control groups (21.7% and 22.3% respectively; 68 and 16 patients) (p�0.05). However, although there was a trend to higher pregnancy rates inthe EmbryoGlue groups (IVF, 58.9%; oocyte donors, 83.3%) compared tothe controls (IVF, 48.5%; oocyte donors, 62.5%), the differences were notsignificant.
Conclusions: This study has shown that the composition of the embryotransfer medium has a direct effect on implantation rate in both IVF patientsand oocyte donors having an embryo transfer on day 3. A plausible mech-anism for the beneficial effect of an increased hyaluronan concentration inthe transfer medium is its increased ability to mix with uterine fluid due toits similar consistency. This may therefore result in fewer embryos beingexpelled following the embryo transfer procedure. The use of a mediumwith an increased hyaluronan concentration should lead to the transfer offewer embryos per patient.
Supported by: Vitrolife.
Monday, October 14, 20022:45 P.M.
O-12
Assessment of the predictive value of follicular fluid cytokines andreactive oxygen species in IVF cycles. Mohamed A. Bedaiwy, KurtMiller, Jeffery M. Goldberg, David R. Nelson, Ashok Agarwal, TommasoFalcone. Cleveland Clin Fdn, Cleveland, OH.
Objective: There is an increasing evidence supporting the role of aregulated cytokine network in the ovulatory process. In humans, it has beenshown that follicular fluid (FF) exerts chemotactic activity toward granu-locytes and that activity is related to the outcome of in vitro fertilization(IVF) treatment. Reactive oxygen species (ROS) in FF have been shown tobe a potential marker of pregnancy in IVF patients. The objective of thisstudy was to investigate the role of FF cytokines and ROS in the periovu-latory FF during IVF cycles.
Design: A prospective studyMaterials/Methods: Follicular fluid from 112 women was obtained while
they underwent oocyte retrieval for IVF. We measured the concentrations of5 cytokines (IL-1, IL-2, IL-12, IL-13, and TNF-) and ROS in FF andcompared the levels among women who became pregnant and those whodid not. Thirty-one endometriosis patients, 15 with idiopathic infertility, 21with tubal-factor infertility, 15 with ovarian-factor infertility, and 30 pa-tients with male-factor infertility were included.
Variable
PregnantCycles
(n � 52)
Non-pregnantCycles
(n � 60) P-values
Age (years) 33.61 � 3.55 34.17 � 4.25 NSDuration of stimulation
(days)9.56 � 1.53 9.55 � 1.79 NS
No. FSH ampouls used 35.30 � 11.41 38.02 � 11.74 NSNo. of oocytes
retrieved12.27 � 6.88 11.87 � 7.11 NS
Fertilization rate 0.66 � 0.21 0.64 � 0.22 NSNo. of frozen embryos
(2 pn)1.72 � 3.73 0.97 � 0.32 NS
No. of frozen embryos(blastocysts)
1.23 � 1.94 0.70 � 1.53 NS
ROS (cpm) 22.83 � 26.67 6.74 � 8.57 0.0400IL-1B (pg/mL) 0.66 � 0.90 0.50 � 0.65 NSIL-12 (pg/mL) 3.87 � 10.70 24.45 � 20.15 0.002
NS - not significant
Results: Interleukin-2, IL-13 and TNF-were not detected in the periovu-latory FF in all patients. Fifty-two patients achieved pregnancy where 60 didnot. Both pregnant and non-pregnant patients were comparable regardingage, parity, ovarian stimulation parameters, fertilization rates, and embryofreezing rates (table). Levels of FF IL-1 were not significantly differentbetween pregnant and non-pregnant cycles. Levels of FF ROS were signif-icantly higher in pregnant compared to non-pregnant cycles (p � 0.04).Levels of FF IL-12 were significantly lower in pregnant compared tonon-pregnant cycles (p � 0.0002).
Conclusions: Higher levels of IL-12 and lower levels of ROS in the
FERTILITY & STERILITY� S5
