Ass.llt_d.repr()d~~tiY

Transcription

1 Ass.llt_d.repr()d~~tiY.'.'.'..techn"lo9Y FERTILITY AND STERILITY VoL 63, No.3, March 1995 Copyright e 1995 American Society for Reproductive Medicine Printed on acid-free paper in U. S. A. The effect of endometriosis, its stage and activity, and of autoantibodies on in vitro fertilization and embryo transfer success rates* W. Paul Dmowski, M.D., Ph.D.t Nasiruddin Rana, M.D., M.P.H. Janina Michalowska, Ph.D. Jan Friberg, M.D. Cynthia Papierniak, M.S. Albert EI-Roeiy, M.D. Oakbrook Fertility Center, Oak Brook and Family Fertility Center at Grant Hospital, Chicago, Illinois Objectives: To analyze IVF cycle parameters, including pregnancy rates (PR), in women with and without endometriosis and to evaluate the effect of the stage and activity of endometriosis and of autoantibodies. Design: A retrospective analysis of 237 consecutive IVF cycles (193 patients), 119 in women with and 118 without endometriosis. The endometriosis group was further subdivided according to the stage and activity of the disease and autoantibody positivity. Setting: Hospital-based and freestanding IVF programs with the same IVF team. Patients: One hundred ninety-three women of reproductive age undergoing IVF; 84 had prior diagnosis of endometriosis, and 19 had other indications for IVF. Within the endometriosis group, 4 did and 44 did not have evidence of active disease. Autoantibodies were measured in 5 patients. Interventions: The IVF protocol was standard with GnRH agonist administered from the midluteal phase of the preceding cycle. Variables included the method of ET and the use of corticosteroids. Main Outcome Measures: Number of follicles produced, number of eggs retrieved, fertilization rates, number of embryos transferred, and PR per transfer. Results: There was no difference between groups in the response to stimulation, number of oocytes retrieved, number fertilized, and number cleaved. The overall PR was 27% per transfer; it was similar in women with and without endometriosis (29% and 25%, respectively). There was also no difference in PR according to the stage or activity of the disease. However, PR in autoantibodypositive and -negative patients were significantly different (22.9% and 45.7%, respectively). Among autoantibody-positive patients treated with corticosteroids, 8 of 1 conceived. Conclusions: This study confirms previous reports that IVF success rates are comparable in women with and without endometriosis regardless of the activity and stage of the disease. However, our study also indicates that autoantibodies may affect adversely implantation of embryos and that this effect can be overcome by administration of corticosteroids. Fertil Steril 1995;63: Key Words: Endometriosis, autoantibodies, IVF success rates, corticosteroids The association between endometriosis and infertility is unclear and frequently questioned, especially for limited (stage I and II) disease (1). A Received January 25, 1994; revised and accepted September 23,1994. Presented at the Conjoint Meeting of The American Fertility Society and The Canadian Fertility and Andrology Society, Montreal, Quebec, Canada, October 11 to 14, variety of mechanisms through which limited endometriosis may lower fertility have been postulated and reviewed recently (2). It has been suggested that some of the factors that could explain low fe- t Reprints requests: W. Paul Dmowski, M.D., Ph.D., Family Fertility Center, Grant Hospital, 55 West Webster Avenue, Chicago, Illinois 6614 (FAX: ). Vol. 63, No.3, March 1995 Dmowski et al. Effect of autoantibodies on IVF success rates 555

2 r cundity and increased frequency of spontaneous abortions (SABs) in endometriosis are as follows: [1] ovulatory dysfunction; [2] endocrine dysfunction; [3] luteinization without follicular rupture and ovum release (luteinized unruptured follicle syndrome); [4] interference with ovum pickup by the fimbriae; [5] increased sperm phagocytosis; [6] decreased fertilization; [7] defective tubal gamete-embryo transport; and [8] early implantation failure caused by hormonal or autoimmune phenomena. Correction of these factors according to numerous studies was associated with higher pregnancy rates (PRs) (3). Yet, other reports suggest that fecundity in endometriosis may not be decreased and that various treatment methods have no effect on fertility (4, 5). During recent years, endometriosis, including limited disease, became an indication for IVF procedures. Two initial reports from the major IVF programs in Melbourne and Norfolk indicated that IVF success rates in women with treated endometriosis and in those with other diagnostic entities were comparable (6, 7). There was no difference in the response to stimulation, number of oocytes retrieved, fertilization rates, number of embryos transferred, or PRs, whether endometriosis or tubal disease was the indication for IVF. This was not surprising, conside ring that IVF and ET procedures can correct a variety of factors, such as ovulatory dysfunction, abnormal ovum pick-up, abnormal fertilization, and altered embryo transport, which have been implicated in low fecundity in endometriosis. Furthermore, hormonal stimulation during the IVF cycle and luteal phase support after ET, a common part of the IVF protocol, can correct follicular or luteal phase dysfunctions that also are considered as contributory to infertility in endometriosis. Subsequently, however, Wardle et a1. (8) reported a significant decrease in the fertilization rate in vitro in women with endometriosis, resulting in lower (although not significantly, because of small numbers) PRs. The authors attributed this finding to the fact that their patients had untreated endometriosis, diagnosed at the time of laparoscopic oocyte aspiration. This observation was supported by a subsequent report based on a larger number of patients by Mills et a1. (9). They reported a significantly lower fertilization rate (48%) in 67 women with endometriosis, as compared with 65% fertilization rate in 198 women with tubal disease or unexplained infertility. However, several other studies observed comparable fertilization rates in endometriosis and in nonmale factor infertility and no relationship between the activity or stage of endometriosis at the time of IVF and the percentage of oocytes fertilized (1-13). Pregnancy rates were comparable in women with and without endometriosis according to some reports (9, 13) or were lower in advanced (stage III and IV) disease, according to others (1, 12). This latter finding prompted some to speculate that the presence of endometriosis may have an adverse effect on implantation and that autoimmune changes may play a role in this phenomenon (14). The purpose ofthis retrospective analysis was to compare IVF parameters including PRs in patients with and without endometriosis and to evaluate the effect of the stage and activity of the disease and of circulating autoantibodies on the outcome of the IVF cycle in our programs. Patients MATERIALS AND METHODS We reviewed medical records of all consecutive patients undergoing IVF procedures in our programs between January 1, 1991 and March 31, The IVF protocol was standard. Gonadotropin-releasing hormone analogue (leuprolide acetate or nafarelin acetate) was administered from the midluteal phase ofthe preceding cycle. Stimulation with hmg began when there was no sonographic evidence of ovarian follicular activity and serum E2 level was < 5 pg/ml; it was continued until E2 levels reached between 384 and 4,38 pg/ml and at least two follicles 18 mm or larger were present. At that point, 1, IU of hcg was administered and was followed 36 hours later by ultrasound-guided transvaginal oocyte aspiration. Cleaving embryos in a two- to eight-cell stage were transferred transcervically into the uterus (ET) or laparoscopically into the fallopian tubes (TET). Embryo transfer was performed if both fallopian tubes were diseased or absent or for economic reasons. Tubal embryo transfer was performed if at least one fallopian tube was normal and patent. Luteal phase support consisted of three 1,5 IU hcg injections every other day from the day of transfer. Progesterone supplementation in a form of 1M injections or lozenges was used if serum P:E2 ratio was < 8 (15). In some patients, IVF regimen was supplemented with cortico steroids (prednisone 1 to 15 mg/d during the entire IVF cycle [Michalowska J, Friberg J, Rana N, Rotman C, Dmowski WP, abstract] or methyl prednisone 96 mg/d for 5 days beginning at ET) 556 Dmowski et a1. Effect of autoantibodies on IVF success rates Fertility and Sterility

3 (16). Corticosteroids were used without prior knowledge of the patients' autoantibody status by individual clinicians to improve IVF success rates based on prior literature reports (17). After January 1992, some patients were randomized to corticosteroids as a part of the specific protocol. A total of 237 consecutive IVF cycles (193 cases) and 18 cryopreserved-thawed ET cycles were performed during the 27-month period. The data for 237 consecutive IVF cycles were analyzed. There were a total of 56 clinical pregnancies, 8 of which ended in SABs and one was ectopic. All patients had at least one laparoscopy before the IVF procedure, and all had undergone extensive infertility evaluation and treatment. On the basis of the laparoscopic findings, they were divided into groups with and without prior diagnosis of endometriosis. In the group without endometriosis, indications for IVF included tubal disease (39%), pelvic adhesions (28%), male factor (25%), unexplained infertility (13%), ovarian dysfunction (17%), and other factors (5%). Male factor, defined as minimum of three semen analyses, which did not meet World Health Organization criteria for normal semen, was also present in 17%, tubal involvement in 2%, and pelvic adhesions in 25% of couples in the endometriosis group. The male factor was severe enough to prevent fertilization in 11 % of cycles in the endometriosis and in 15% of cycles in the no endometriosis groups. These patients did not go through ET procedure and are excluded from the PR analysis. Patients with endometriosis were subdivided into groups with active and inactive disease. Active endometriosis was considered if there was sonographic evidence of endometriomas during IVF stimulation, and/or if endometriomas were aspirated during oocyte retrieval, and/or if active endometriosis was observed during laparoscopic TET. The severity of the disease was staged according to the revised AFS classification (1). For the purpose of data analysis, the most advanced stage of endometriosis, as well as the most recent stage of the disease, was considered. A subset of 5 patients with endometriosis had serum autoantibodies measured within 1 months of the IVF attempt. These were immunoglobulin (Ig)G, IgM, and IgA isotype antibodies to cardiolipin, phosphatidylserine (PS), phosphatidylethanolamine (PE), phosphatidylglycerol (PG), phosphatidylinositol (PI), phosphatidic acid (PA), histone 2A (H2A) and 2B (H2B) fractions, singlestranded DNA (ssdna) and double-stranded DNA (dsdna). Assay The autoantibodies were measured using a modification of the previously described ELISA technique (14). Briefly, ELISA plates were coated with cardiolipin, PS, PE, PG, PI, PA, H2A, H2B, ssdna, or dsdna and incubated at 4 C overnight. Antigen-coated plates were protein blocked using 1% donor bovine calf serum (Irvine Scientific, Santa Ana, CA) in phosphate-buffered saline (PBS) and washed. The positive standard was obtained from a woman with systemic lupus erythematosus who had measurable antibodies to each of the antigens for the three isotypes measured. A seven-point dilution curve of the positive standard serum was used to generate the standard curve for each plate. Standards and patients' sera were diluted in 1% donor bovine calf serum-pbs, added to precoated plates, and incubated overnight at 4 C. Plates were washed and incubated with enzyme-labeled, affinity-purified anti-human IgG, IgM, and IgA antibodies. All plates were washed, and the substrate dye 2,2'-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid (Sigma Chemical Co., St. Louis, MO) in.5 M citrate buffer with H 2 2 was added. Plates were read at optical density (OD) 45. For each plate the maximal standard (the minimal dilution of the positive standard serum) was assigned the arbitrary value of 1 units per ml. The values of the remaining six dilutions were calculated according to their respective OD. Curve fitting of the seven points standard curves was performed using an exponential regression formula. Serum values were calculated similarly. Control sera were obtained from 32 women of reproductive age who had no autoimmune diseases, no infertility problems, and who during laparoscopic sterilization were negative for endometriosis. The values of autoantibodies in this group were not distributed normally; thus, all data were log transformed. Antibody values that exceeded the mean + 2 SD were defined as positive. Statistical Analysis For standard comparison of rates and percent- 2 ages between groups, the Pearson X statistic was used and was confirmed with Fisher's exact test. Fr comparison of continuous variables, t-test was used when two groups were compared, and analysis of variance F-test was used when more than two groups were compared. Because the chance of an abnormal autoantibody result increases in proportion to the number of tests performed, we calcu- Vol. 63, No.3, March 1995 Dmowski et al. Effect of autoantibodies on IVF success rates 557

4 r lated the expected number of patients with abnormal autoantibodies based on the binomial distribution when outcomes are equally likely (18). We took into account the total number of autoantibodies tested, the number and probability of abnormal antibodies' and the number of patients in the study. Contributing factors that may affect autoantibody production such as age, gravidity, endometriosis stage and activity were included in a regression adjustment analysis. Based on these calculations, the patient was defined as autoantibody-positive when three or more autoantibodies were present. Such an event could occur by chance in only 3.2% of patients. RESULTS Table 1 Analysis of 237 Consecutive IVF Cycles in Women With or Without Endometriosis, With or Without Evidence of Active Endometriosis, and Autoantibody-Positive or -Negative Patients' characteristics and IVF cycle data for women with or without endometriosis, with or without evidence of active disease, and for those autoantibody-positive or -negative are demonstrated in Table 1. Women with endometriosis had somewhat shorter infertility, but there were no differences in any of the other parameters with exception of the PRs between the groups. Pregnancy rates per transfer were similar in women with and without endometriosis (29% and 25%, respectively) and within the endometriosis group between patients with and without evidence of active disease (31 % and 28%, respectively). Pregnancy rates were, however, significantly lower in the autoantibodypositive than in the autoantibody-negative group (22.9% and 45.7%, respectively, P <.5). When the data were analyzed according to the stage of endometriosis, the results were comparable whether the most advanced or most recent stage of the disease was considered. As illustrated in Table 2, there was no difference in any of the parameters analyzed. Pregnancy rates, although lower in patients with stage IV endometriosis, were also not significantly different. Progesterone support during the late luteal phase and early gestation was used liberally, but there was no difference between the groups in the frequency or dose of P used. Of 5 patients with endometriosis who were tested for autoantibodies, 25 had to 2 antibodies and were considered autoantibody-negative, whereas the remaining 25 had between 3 and 18 autoantibodies and were considered autoantibodypositive. The most commonly elevated were IgM to PS, followed by IgG to ssdna, IgM to ssdna, IgM to H2A, and IgA to PE. There was no difference in the distribution of autoantibody-positive patients according to the stage of endometriosis. Among 25 women autoantibody-positive, 2 had evidence, and 5 had no evidence of active disease. In the autoantibody-negative group, 9 did and 16 did not have evidence of active disease. The distribution of pregnancies according to the number of autoantibodies and autoantibody-positivity is demonstrated in Table 3. It is interesting to note that all patients with three or more autoanti- Endometriosis All cases Active Endometriosis Autoantibody- Autoantibody- Endometriosis No endometriosis Yes No positive negative No. of cycles/cases 119/84 118/19 52/4 67/44 38/25 35/25 Age (y) (mean ± SD) 34.4 ± ± ± ± ± ± 5. Length of infertility (y) (mean ± SD) 4.6 ± 2.8* 5.5 ± 2.9' 5.1 ± ± ± ± 3. No. of follicles 2: 14 mm (mean ± SD) 12. ± ± ± ± ± ± 7.8 No. of eggs retrieved (mean ± SD) 8.2 ± ± ± ± ± ± 6.5 Fertilization rate (%) No. of embryos transferred (mean ± SD) 3.7 ± ± ± ± ± ± 2.3 No. of transfers No. of pregnancies PR per transfer (%) * 45.7* No. of abortions No. of ectopics * Difference statistically significant at P < Dmowski et al. Effect of autoantibodies on IVF success rates Fertility and Sterility

5 Table 2 Analysis of IVF Cycles in Women With Endometriosis According to the Most Advanced Stage of the Disease Stage I No. of cycles/cases 49/37 Age (y) (mean ± SD) 34.7 ± 4.7 Length of years infertility (mean ± SD) 5.2 ± 3.7 No. offollicles ~ 14 mm (mean ± SD) 12.3 ± 7.7 No. of eggs retrieved (mean ± SD) 8.5 ± 6.3 Fertilization rate (%) 61 No. embryos transferred (mean ± SD) 3.5 ± 1.9 No. of transfers 42 No. of pregnancies 11 PR per transfer (%) 26 Stage II Stage III Stage IV 4/22 21/19 9/ ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± bodies who conceived were on corticosteroid treatment, as were two of six who conceived with two autoantibodies present and four of seven with none. Table 4 demonstrates the effect of corticosteroids on IVF PRs in patients autoantibody-positive and -negative. Corticosteroids were administered during 1 IVF cycles to patients autoantibody-positive. Eight ofthese conceived (8%). This contrasts with no pregnancies during 25 IVF cycles in autoantibody-positive patients in a "no corticosteroids" group (P <.1). By comparison, in a group of autoantibody-negative patients with endometriosis, conception occurred during 7 of 15 corticosteroid and 9 of 2 noncorticosteroid supplemented cycles, for almost identical PRs of 46.7% and 45%, respectively. To exclude the effect ofthe method of transfer as a confounding factor, we attempted to evaluate sep- arately TET and ET PRs according to the autoantibody positivity (Table 5). In the TET group, PRs in autoantibody-positive and -negative patients were, respectively, 29.4% and 59.1 % (P =.65), barely missing statistical significance. In the ET group, there was the same trend, but the numbers were smaller and the differences not significant. DISCUSSION It is agreed generally that PRs resulting from IVF-ET procedures depend on multiple variables. Age of the patient, indication for IVF, diseases of the reproductive system, stimulation protocol, quality of the gametes, and type of IVF procedure playa significant role. Thus, comparisons of PRs between different programs and even within the same program between patients with different diagnostic Table 3 Distribution of IVF Pregnancies According to the Number of Autoantibodies and Autoantibody Positivity No. of autoantibodies No. IVF cycles No. pregnant Percent pregnant Patients autoantibody-negative (4)* 3 6 (2) Total o to (6) 45.7t Patients autoantibody-positive:j: (3) 1 (1) 2 (2) 2 (2) Total 3 to (8) 22.9t * Values in parentheses are number of patients who received corticosteroids during the IVF cycle. t Values significantly different at P <.5 :j: Autoantibody positivity was defined as the presence of three or more autoantibodies. Vol. 63, No.3, March 1995 Dmowski et al. Effect of autoantibodies on IVF success rates 559

6 r Table 4 Effect of Corticosteroids on IVF PRs No. IVF cycles No. pregnant Percent pregnant Probability Patients autoantibody-positive Corticosteroids during IVF cycle No corticosteroids Patients autoantibody-negative Corticosteroids during IVF cycle No corticosteroids <.1 NS* * NS, not significant. entities are difficult. In performing the analysis of our data, we attempted to account for each identifiable variable and evaluated its possible contributory role. All our patients were managed according to the same stimulation protocol, their age was comparable between the diagnostic groups, and they were under the care of the same IVF team. The contribution ofthe relative male factor into the endometriosis and the no endometriosis groups differed slightly, but severe male factor resulting in a lack of fertilization was similar and was excluded from the analysis of PRs because no transfers were performed. In agreement with previous reports (1-13), our patients with endometriosis and those with other diagnostic entities responded in a similar way to ovarian stimulation. They produced a similar number of follicles, a comparable number of - cytes was retrieved, there was a similar fertilization rate, and a comparable number of embryos was transferred. These parameters were also similar when women with and without evidence of active disease were compared or when endometriosis was staged according to the AFS classification. Also, in agreement with the previously published data (1, 13), our PRs per transfer were comparable in patients with and without endometriosis and regardless of whether there was evidence or not of the active disease. There is no question that our group of women "without evidence of active disease" is imperfect. Many of the patients in this category probably had stage I or II disease that was not evi- dent sonographically or was missed at the time of laparoscopic TET because of ovarian enlargement, resulting in a difficult laparoscopic examination. On the other hand, the group with active endometriosis consisted exclusively of patients with active disease, that is, endometriomas were aspirated at the time of oocyte retrieval or active endometriotic lesions were observed laparoscopically at the time of ET. In contrast to the earlier reports (8), but in agreement with more recent studies (13), there was no decrease in the fertilization or PRs in women with active endometriosis as compared with no endometriosis controls. Numerically, the PRs in patients with and without endometriosis were higher than in some reports (8, 11, 12) but comparable with others (9, 1, 13). In agreement with Inoue et al. (13), we did not find significant differences between PRs according to the stage of the disease, although there was a trend to lower rates in stage IV. Yovich and Matson (12) and Chillik et al. (1) reported a decline in PRs with increasing stage of endometriosis. Autoantibodies to phospho lids, histones, and polynucleotides have been observed previously in approximately 6% of women with endometriosis (19). In this study, 5% of patients with endometriosis were considered autoantibody-positive. Furthermore, and in contrast to our prior report (2), we did not observe higher frequency of autoantibodies in limited, as compared with advanced endometriosis. There was also no apparent relationship be- Table 5 Effect of Autoantibodies and the Method of ET on IVF PRs No. IVF cycles No. pregnant Percent pregnant Probability TET Patients autoantibody-positive Patients autoantibody-negative Transcervical uterine ET Patients autoantibody-positive Patients autoantibody-negative NS* * NS, not significant. 56 Dmowski et al. Effect of autoantibodies on IVF success rates Fertility and Sterility

7 tween the presence of autoantibodies and activity of endometriosis. Active disease was observed in 2 of 25 autoantibody-positive and in 9 of 25 autoantibody-negative patients. We previously reported that abnormal autoantibodies are suppressed with danazol, but not with GnRH agonists, and that they seem to be related to infertility (2). EI-Roeiy and associates (14) reported increased prevalence of these autoantibodies in serum and in follicular fluids of women undergoing IVF. However, there have been no studies to date, correlating success of the IVF procedures with the presence or absence of autoantibodies. The present study is the first to draw attention to the effect of circulating autoantibodies on the success of IVF procedures in women with endometriosis. As indicated in Table 1, PRs in autoantibodynegative patients were more than twice higher than in autoantibody-positive patients. We considered as autoantibody-positive only those patients who had three or more autoantibodies because 1 or 2 autoantibodies out of 3 tested could be elevated by chance in about 13% of patients. The likelihood of three or more being elevated by chance has been calculated as only 3.2% (18). Prospective studies to test the effect of autoantibodies on IVF success rates are currently underway in our programs. A significantly higher PR in autoantibody-positive patients treated with corticosteroids lends further credibility to our observations. Nevertheless, the study has all the limitations of the retrospective protocol, and the results should be considered as preliminary until confirmed by other programs. It is interesting to note that 1 of 17 conceptions in patients with one or more autoantibody occurred when corticosteroids were used during the IVF cycle (Table 3). The role of corticosteroids in autoantibody suppression is well recognized. Corticosteroid treatment was one of the variables in our program, and we attempted to evaluate its role as a confounding factor. As indicated in Table 4, autoantibody-positive patients undergoing IVF were much more likely to conceive if they were taking corticosteroids during the IVF cycle. This observation is in agreement with the previous reports, indicating higher PRs with the use of glucocorticoids during the IVF cycle (Michalowska J, Friberg J, Rana N, Rotman C, Dmowski WP, abstract) (16, 17). There was no difference in PRs when corticosteroids were or were not used in autoantibodynegative patients. It is possible that suppression of abnormal autoantibodies by corticosteroids facilitated implantation and pregnancy. Another variable in our study was the method of ET. Higher PRs with tubal as compared with uterine ETs previously have been reported and attributed to a more favorable tubal environment (21). Our observations also indicate that in patients with endometriosis, the PR with TET is significantly higher than with ET (Dmowski WP, Michalowska J, Rana N, unpublished data). Therefore, to exclude the possibility of a confounding effect of the method of transfer, we analyzed separately TET and ET PRs in autoantibody-positive and autoantibody-negative patients. With each method of transfer, there was a trend to a higher PR in the autoantibody-negative group, but the numbers were small and there was no statistical significance. In conclusion, the results of this study suggest that autoimmune phenomena may play a role in IVF success rates in women with endometriosis. The mechanism responsible for lower PRs in autoantibody-positive patients most likely involves interference with implantation because there was no difference in the number of oocytes retrieved, fertilization rates, or number of embryos transferred between the groups. Weed and Arguembourg (22), who several years ago reported IgG and complement deposits in the uterine endometrium of women with endometriosis, postulated that antigen-antibody reaction in the uterine cavity may interfere with implantation and/or result in the increased frequency of SABs in this disease. Abnormal autoantibodies also have been observed in patients with recurrent miscarriages and subchorionic hematomas (23, 24). It is interesting to note that in this study several of the autoantibodypositive patients who conceived had first trimester bleeding and/or sonographic evidence of subchorionic hemorrhage, although the pregnancy continued. REFERENCES 1. The American Fertility Society. Revised American Fertility Society classification of endometriosis: Fertil Steril 1985;43: Halme J. Surrey ES. Endometriosis and infertility: the mechanisms involved. In: Chadha D, Buttram V Jr, editors. Current concepts in endometriosis. N ew York: Alan R. Liss, 199: Damewood MD. Endometriosis and infertility: physiologic mechanisms and contemporary therapeutic considerations, including GnRH agonists. Semin Reprod Endocrinol 1993:11: Rodriguez-Escudero FJ, Neyro JL, Corcostegui B, Benito Vol. 63, No.3, March 1995 Dmowski et al. Effect of autoantibodies on IVF success rates 561

8 JA. Does minimal endometriosis reduce fecundity? Fertil Steril 1988;5: Hughes EG, Fedorkow DM, Collins JA. A quantitative overview of controlled trials in endometriosis-associated infertility. Fertil Steril1993;59: Mahadevan MM, Trounson AO, Leeton JF. The relationship of tubal blockage, infertility of unknown cause, suspected male infertility, and endometriosis to success of in vitro fertilization and embryo transfer. Fertil Steril 1983;4: Jones HW, Acosta AA, Andrews MC, Garcia JE, Seegar Jones G, Mayer J, et al. Three years of in vitro fertilization at Norfolk. Fertil SteriI1984;42: Wardle PG, McLaughlin EA, McDermott A, Mitchell JD, Ray BD, Hull MGR. Endometriosis and ovulatory disorder: reduced fertilization in vitro compared with tubal and unexplained infertility. Lancet 1985;2: Mills MS, Eddowes HA, Cahill DJ, Fahy UM, Abuzeid MIM, McDermott A, et al. A prospective controlled study of in vitro fertilization, gamete intra-fallopian transfer and intrauterine insemination combined with superovulation. Hum Reprod 1992;7: Chillik CF, Acosta AA, Garcia JE, Perera S, Van Uem J F, Rosenwaks Z, et al. The role of in vitro fertilization in infertile patients with endometriosis. Fertil Steril1985;44: Matson PL, Yovich JL. The treatment of infertility associated with endometriosis by in vitro fertilization. Fertil Steril 1986;46: Y ovich JL, Matson PL. The influence of infertility etiology on the outcome ofivf-et and GIFT treatments. Int J Fertil 199;35: Inoue M, Kobayashi Y, Honda I, Awaji H, Fujii A. The impact of endometriosis on the reproductive outcome of infertile patients. Am J Obstet Gynecol 1992;167: EI-Roeiy A, Gleicher N, Friberg J, Contino E, Dudkiewicz A. Correlation between peripheral blood and follicular fluid autoantibodies and impact on in vitro fertilization. Obstet Gynecol 1987;7: Maclin VM, Radwanska E, Binor Z, Dmowski WP. Progesterone:estradiollevels at implantation in ongoing pregnancies, abortions and nonconception cycles resulting from ovulation induction. Fertil Steril 199;54: Polak de Fried E, Blanco L, Lancuba S, Asch RH. Improvement of clinical pregnancy rate and implantation rate of in-vitro fertilization-embryo transfer patients by using methylprednisone. Hum Reprod 1993;8: Kemeter P & Feichtinger W. Prednisolon verbessert die schwangerschafts-rate der ivf. Fertilitat 1986;2: Ingeltinger JA, Mostella F, Thibodeau LA, Ware JH. Binomial distribution when outcomes are equally likely. In: Ingeltinger JA, editor. Biostatistics in clinical medicine. New York: McMillan, 1987: Gleicher N, EI-Roeiy A, Contino E, Friberg J. Autoantibodies in reproductive failure: Is endometriosis an autoimmune disease? Obstet Gynecol 1987;7: EI-Roeiy A, Dmowski WP, Gleicher N, Radwanska E, Harlow L, Binor Z, et al. Danazol but not GnRH agonists suppresses autoantibodies in endometriosis. Fertil Steril 1988;5: Asch RH. Uterine versus tubal embryo transfer in the human. Comparative analysis of implantation, pregnancy, and live-birth rates. Ann NY Acad Sci 1991;626: Weed JC, Arguembourg PC. Endometriosis: can it produce an autoimmune response resulting in infertility? Clin Obstet Gynecol 198;23: Parke AL, Wilson D, Maier D. The prevalence of antiphospholipid antibodies in women with recurrent spontaneous abortion, women with successful pregnancies, and women who have never been pregnant. Arthritis Rheum 1991;34: Baxi LV, Pearlstone MM. Subchorionic hematomas and the presence of autoantibodies. Am J Obstet Gynecol 1991;78: Dmowski et al. Effect of autoantibodies on IVF success rates Fertility and Sterility