Several structurally unrelated scaffolds of the Rho kinase inhibitor were designed using pharmacophore information obtained from the results of a high-throughput screening and structural information from a homology model of Rho kinase. A docking simulation using the ligand-binding pocket of the Rho kinase model helped to comprehensively understand and to predict the structure–activity relationship of the inhibitors. This understanding was useful for developing new Rho kinase inhibitors of higher potency and selectivity. We identified several potent platforms for developing the Rho kinase inhibitors, namely, pyridine, 1H-indazole, isoquinoline, and phthalimide.

PROBLEM: Cytokines are involved in implantation success and failure. We envisage that they could be similarly involved in pre-eclampsia (PE).MATERIALS AND METHODS: First, we review the primipaternity and primiparity concepts and then why natural killer (NK) cells are involved in implantation. We stress that the common event in all PE is vascular remodelling.RESULTS AND CONCLUSION: We conclude that PE could involve cytokine and/or NK dysfunctions, and propose a working hypothesis.

Objective: To document in the endometrium the correlation among the interleukin (IL)-12, -15, and -18 mRNA and the correlation between cytokine levels, vascular status, and endometrial natural killer (NK) cell count in the context of recurrent implantation failure.Design: A pilot study.Setting: Department of Reproductive Immunology.Patient(s): Women who failed to become pregnant after repeated IVF-embryo transfer and fertile control subjects.Intervention(s): Ultrasonic evaluation and endometrial biopsy in luteal phase.Main Outcome Measure(s): Uterine artery Doppler, count of uterine CD56 bright cells/field, and quantification by real-time polymerase chain reaction (PCR) to monitor IL-12 family (IL-12p40, IL-12p35, EBI3, IL-23), the IL-18 system (IL-18, IL-18R, IL18BP), and the IL-15 mRNA ratio.Result(s): The uterine artery Doppler and the CD56 bright cell counts were significantly different in fertile and infertile patients. The mean uterine artery pulsatility index correlated significantly negatively with the IL-18/actin ratio suggesting a defect of the cytokine-dependent vascular remodeling pathway. The number of uterine CD56 bright cells was significantly correlated with the IL-15/actin and IL-18/IL-18BP ratios. Thus, IL-18 and IL-15 seems to be involved in the local recruitment and the activation of uterine natural killer (uNK) cells. IL-18 was itself correlated with IL-15 and IL-12, suggesting a local control of uNK cells activation.Conclusion(s): The assessment of the tripod IL-12/-15/-18 shows distinct immune-related mechanisms that are involved in the broader context of inadequate uterine receptivity. (c) 2005 by American Society for Reproductive Medicine.

In this paper, we briefly survey the history of concepts in reproductive immunology from antibody-mediated tolerance to the ‘fetal allograft’ to the current concept of an embryo ‘bathing in a sea of cytokines’. We then review the paradigm that ‘allopregnancy is a Th2 phenomenon’ and some of the evidence gained in animals and humans supporting it. We continue by discussing the light it sheds on immunologically caused recurrent abortion, and the present status of the concepts. We next show the limits of the Th1/Th2 paradigm by reviewing the role of ‘inflammatory’ cytokines in implantation (as first seen with leukemia inhibitory factor). We go on to discuss recent data showing that interferon-γ is not solely a ‘bad guy’, e.g. abortifacient as the paradigm would predict, but is needed at low doses for the vascular development and transformation of uterine spiral arteries required for implantation and successful pregnancy. We conclude by discussing the emerging role of NK and IL-12, IL-15, IL-18 tripods and other cytokines in local angiogenesis and tissue remodelling, a series of new data bringing us well beyond the Th1/Th2 paradigm in pregnancy which, in this context, appears now obsolete and an oversimplification, although it has indeed been useful at first. Rather, step-specific events have to be considered and a key role is seen in local tissue remodelling, in which immune cytokines play an important role while not always being secreted by immune cells.

It has recently been shown that uNK (uterine natural killer) cells may have a very important role in the uterine receptivity process and could be involved in implantation failure. We thus determine whether the uterine Natural killer (uNK) cell count is hormonally influenced and might help define which patients for whom unstimulated IVF-ET may be successful. On day 21 of two consecutive cycles, we performed an endometrial biopsy to assess the uterine NK cells count by anti-CD56 immunostaining. The first cycle took place under vaginal estrogen-progestin replacement treatment, the second with no hormone treatment. All the patients included failed to become pregnant after repeated ovarian hyperstimulation followed by IVF-ET. The main outcome measure was the pregnancy rate after the first cycle of unstimulated IVF-ET. We observed a significant reduction (r=0.54; p=0.04) in the CD56 bright cell count during the natural cycle compared with the artificial cycle. 33% (5/15) patients became pregnant during the first cycle of IVF-ET. Those who became pregnant had significantly more CD56 bright cells (p=0.02) at the initial assessment during hormone treatment than those who did not become pregnant. These preliminary data suggest that patients with a high uterine NK cell count during estrogen-progestin treatment may benefit from unstimulated IVF-ET.

In this paper, we briefly survey the history of concepts in reproductive immunology from antibody-mediated tolerance to the "fetal allograft" to the current concept of an embryo "bathing in a sea of cytokines". We then review the paradigm that "allopregnancy is a Th2 phenomenon" and some of the evidence gained in animals and humans supporting it. We continue by discussing the light it sheds on immunologically caused recurrent abortion, and the present status of the concepts. We next show the limits of the Th1/Th2 paradigm by reviewing the role of "inflammatory" cytokines in implantation (as first seen with leukemia inhibitory factor). We go on to discuss recent data showing that interferon-gamma is not solely a "bad guy", e.g. abortifacient as the paradigm would predict, but is needed at low doses for the vascular development and transformation of uterine spiral arteries required for implantation and successful pregnancy. We conclude by discussing the emerging role of NK and IL-12, IL-15, IL-18 tripods and other cytokines in local angiogenesis and tissue remodelling, a series of new data bringing us well beyond the Th1/Th2 paradigm in pregnancy which, in this context, appears now obsolete and an oversimplification, although it has indeed been useful at first. Rather, step-specific events have to be considered and a key role is seen in local tissue remodelling, in which immune cytokines play an important role while not always being secreted by immune cells.

BACKGROUND: Most implantation failures after successful in vitro fertilization-embryo transfer (IVF-ET) result from inadequate uterine receptivity. There is currently no way to predict this receptivity. METHODS: We investigated whether the detection of interleukin-(IL)18 by ELISA in uterine luminal secretions might predict implantation failure. Secretions of 133 patients enrolled in our IVF-ET program were sampled by uterine flushing immediately before oocyte retrieval. We assessed the following outcomes: pregnancy rate, multiple pregnancy rate, and implantation rate per embryo transferred. RESULTS: Interleukin-18 was detected in the flushing fluid of 38 patients (28.6%). Although the two groups were comparable for all other characteristics (age, etiology, ovarian reserve, number of embryos transferred, quality of embryos), all outcome variables differed significantly. The pregnancy rate was 37.9% in the IL-18 -ve group and 15% in the IL-18 + ve group, the multiple pregnancy rate 27.7% and 0%, and the implantation rate per embryo transferred 19.4% and 6.7% (all comparisons, P=0.02). Only embryos meeting good quality criteria were transferred to 65 patients: 50 IL-18 -ve and 15 IL-18 +ve. The pregnancy rate was 51% for the IL-18 -ve group and 20% for the IL-18 +ve group, the multiple pregnancy rate 36% and 0.0%, respectively, and the implantation rate 29% and 8.3% (P=0.02). CONCLUSION: This non-invasive and simple method predicted inadequate uterine receptivity, independent of embryo quality.

Il n’est pas possible de réduire l’interaction materno-fœtale à une simple tolérance maternelle vis-à-vis d’un tissu étranger. Il s’agit aussi d’une interaction cytokinique complexe qui dirige et contrôle, non seulement la réaction immunitaire, mais aussi les différents éléments de l’implantation comme l’adhésion et les phénomènes vasculaires. Ainsi, certaines cytokines sécrétées localement comme l’interleukine 18 (IL18) participent au contrôle de l’implantation.

The materno-foetal relationship is not simply maternal tolerance of a foreign tissue, but a series of intricate mutual cytokine interactions governing selective immune regulation and also control of the adhesion and vascularisation processes during this dialogue. There is strong evidence that locally secreted cytokines, such as interleukine 18 (IL18) control the implantation process and can cause implantation failure in case of absence or overactivation. Uterine flushing fluids may be analysed to determine the level of several cytokines. At the time of egg retrieval, the flushing procedure does not adversely affect pregnancy rates. We report a strong positive correlation between the presence of IL18 in the uterine flushing and bad implantation rates. The presence of IL18 in the lumina is the traduction of an overactivation of endometrial IL18 that should be diagnosed and treated. Moreover, endometrial biopsy could define which type of cytokinic dysregulation is implicated in repeated implantation failure and define which type of treatment need to be introduced.

Objective: To investigate the endometrial immunohistochemical staining of interleukin (IL)-12 and IL-18 and to quantify the CD56 bright natural killer (NK) cells in relation to Doppler vascular disorders.Design: Controlled clinical study.Setting: Research unit of a university hospital.Patient(s): Thirty-five women with repeated implantation failure after ET in IVF and 12 fertile control patients.Intervention(s): Ultrasound evaluation and endometrial biopsy on day 20.Main Outcome Measure(s): The balance between IL-12 and IL-18, the number of NK cells, and the vascular status among fertile and implantation failure patients.Result(s): The control patients displayed normal vascular parameters, a weak anti-IL-12 staining, a consistent moderate stromal anti-IL-18 staining, and fewer than 15 NK cells/field. This pattern was observed among only 17% (6/35) of the implantation failure group. The remaining patients fit into one of two patterns: [1] 37% (13/35) had more than 40 NK cells/field with a strong anti-IL-12 and/or anti-IL-18 staining, and [2] the remaining 46% (16/35) had a marked local depletion of IL-18 and IL-12. Respectively, 85% and 31% of two groups displayed abnormal vascular parameters.Conclusion(s): Distinctions between the different local dysregulations of the cytokine network may provide clues for further exploration and treatment.

We briefly review the history of concepts (some of which are still valid) which have lead to the present situation where pregnancy is viewed as being a Th2 phenomenon. We recall some of the early evidence which has been taken as supporting the general validity of this concept in murine and human pregnancy. We then recall some of the recent data dealing with “newer” cytokines and the role of uterine natural killer (NK) cells at the feto-maternal interface which fit neither with a steady-state concept nor with inflammatory cytokines, being solely “bad guys” as the paradigm would predict, nor with the concept of reduction of NK “activity” being required for successful pregnancy. As an example of the newer complexity, we briefly recall some of our recent micro-array studies in mice, and describe briefly our most recent data in human pointing out the importance of the tripod IL-12/IL-18/NK in successful or failed pregnancy in human, perhaps under IL-15 control. We conclude by a repeated warning against the so-called rationales of lymphocyte alloimmunization for therapy of recurrent spontaneous abortion and improvement of implantation rates.

Patients with a history of repeated and unexplained implantation failure after IVF-embryo transfer and a high uterine natural killer cell count during estrogen-progestin treatment may benefit from controlled natural IVF-ET.

Pregnancy is controlled primarily, though not exclusively, by a delicate equilibrium between locally acting growth factors and cytokines, some under steroid control. The hypothesis considered here is that stress is able to influence the equilibrium between cytokines and thus lead to abortions or implantation failure. We thus detailed the studies on that topic in order to explore the psycho-neuro-immunological mechanisms concerned. The duration of stress, the patient's strategy for coping with this and the social context might be able to produce some opposite immunological effects. Thus, the link between stress and the immunological events induced is complex, and much care is needed for such patients.

We restate briefly why we consider that the Th1/Th2 paradigm, as useful as it has been, is now no longer adequate and is obsolete. We take as an example the role of IL-18, abortifacient at high doses but cardinal for the control of natural killer (NK) cell effects on spiral artery remodelling in mice, and likely also in humans. We then describe briefly our recent studies on cytokine defects and implantation failure in humans, a key feature being the link between uterine cytokine dysregulation and abnormal uterine vascular scores. We draw lessons for preeclampsia, and describe features of a model for its immune aetiology.

Objective: To assess intrauterine levels of leukemia inhibitory factor (LIF) by uterine flushing at the time of egg retrieval and to confirm that the procedure has no detrimental effect on pregnancy rates.Design: Prospective study.Setting: Assisted reproductive unit of a university hospital.Patient(s): Uterine flushing was performed in 148 IVF patients. The first 100 patients were compared with a matched control group.Intervention(s): Uterine flushing at the time of egg retrieval.Main Outcome Measure(s): IVF-ET results, pregnancy rates, and intrauterine LIF levels.Result(s): Pregnancy rates were not different in the group of patients with (27%) or without uterine flushing (28%). Leukemia inhibitory factor was detected in 60 patients (46%). Pregnancy rates did not differ between patients' detectable LIF and those in whom LIF was undetectable. Mean levels of LIF were 30.1 +/- 49.3 pg/mL and 28.6 +/- 51.2 pg/mL in pregnant and nonpregnant patients respectively.Conclusion(s): The flushing procedure at the time of egg retrieval did not adversely affect pregnancy rates. Leukemia inhibitory factor was detected in 46% of patients at the time of egg retrieval, but no correlation were observed with better pregnancy rates in patients with detectable LIF. Mean LIF levels did not differ in pregnant and nonpregnant women. Access to endoluminal secretions of the endometrium during IVF-ET may represent a new research in human implantation. (C) 2003 by American Society for Reproductive Medicine.

Objective: To assess intra-uterine LIF levels by uterine flushing at the time of egg retrieval. To confirm that the procedure had no detrimental effect on pregnancy rates. Design: Prospective study. A matched control group without flushing was used for the first 100 patients. Materials/Methods: Uterine flushings were performed in a total of 148 infertile patients undergoing IVF-ET. The first 100 patients in the study group were compared with a controlled group of 100 patients without flushing matched for age, number of attempts, day-3 FSH and type of procedure (classical IVF or ICSI). Uterine flushing were performed at the time of egg retrieval. The main outcome measures were: IVF-ET results, pregnancy rates, intra-uterine LIF levels. Results: The pregnancy rates were not different in the group of patients with (27%) or without uterine flushing (28%). In 18 cases (12%), the flushing could not be used for technical reasons (small volume, debris). For 60 patients (46%), we were able to detect LIF in the liquid collected after exposure to the uterine cavity. In the group where LIF was detected, the range of detection was 20 to 260 pg/ml (mean = 63.7 SD= 54.8 pg/ml). The pregnancy rates in the LIF detectable and undetectable group were 28.3% and 27.1% respectively (NS). The mean levels of LIF were mean=30.1 SD= 49.3 and mean=28.6 SD=51.2 in the pregnant and not pregnant patients respectively and these results are not statistically different. Conclusions: The flushing procedure at the time of egg retrieval did not adversely affect the pregnancy rates. We report for the first time that LIF could be detected in some patients (46%) at the time of egg retrieval. No correlation were observed towards better pregnancy rates in patients with detectable LIF in the uterine flushing. The mean level of LIF was not different in the pregnant and not pregnant women. The measurements of other cytokines levels is under progress. Access to the endo-luminal secretion of the endometrium during IVF-ET, open a new research area in human implantation. Supported by: None.

Objective: Ultrasound assessment of the endometrium has become a standard procedure during the diagnostic work-up and treatment of infertility. The clinical differences in endometrial thickness remain controversial and clinicians have not been able to establish an ideal thickness for conception. However, the concept that some minimum thickness is required to establish a clinical pregnancy is widely accepted and thin endometrium, unresponsive to intensive estrogen therapy, is associated with a poor outcome in ART. A combination of pentoxifylline (PTX) and tocopherol (Vit. E) has been reported to be an effective treatment of musculocutaneous radiation-induced fibrosis, both in an experimental model and in humans. We therefore hypothesized that any thin uterine endometrium, even when the causative agent is unknown, can be treated like iatrogenically induced fibroDesign: We tested this treatment in 18 oocyte recipients whose endometrium remained thin in the late proliferative phase, despite vaginal treatment with micronized estradiol started the first day of the cycle observed. All patients received the combination of PTX and Vit. E for six months. We then evaluated the effect on endometrial thickness and the pregnancy and delivery rates.Materials/Methods: The inclusion criteria were designed to enroll oocytes recipients (n = 18) who failed to develop a preovulatory endometrial thickness of at least 6 mm after receiving vaginal micronized estradiol. The eighteen patients received a combination of PTX (800 mg/d) and Vit. E (1000 IU/d) for six months. The main outcome measurements were the change in endometrial thickness and the pregnancy and delivery rates after treatment.Results: Endometrial thickness (ETh) increased significantly (p <0.001): the mean Eth was 4.9 +/− 0.6 mm before and 6.2 +/−1.4 mm after treatment, with 72% (13/18) good responders. Five patients did not respond to the treatment. Three patients, among two after previous radiotherapy, were spontaneously pregnant, and two were pregnant after embryo transfer. Three patients did not have embryo transfer. The pregnancy rate was thus 33% and the delivery rate 27%Conclusions: Treatment by PTX and Vit. E together appears to improve the pregnancy rate in patients with a thin endometrium by increasing the endometrial thickness and improving ovarian function especially after total body irradiationSupported by: Fondation fertilité stérilité et l'INSERM. Objective: Ultrasound assessment of the endometrium has become a standard procedure during the diagnostic work-up and treatment of infertility. The clinical differences in endometrial thickness remain controversial and clinicians have not been able to establish an ideal thickness for conception. However, the concept that some minimum thickness is required to establish a clinical pregnancy is widely accepted and thin endometrium, unresponsive to intensive estrogen therapy, is associated with a poor outcome in ART. A combination of pentoxifylline (PTX) and tocopherol (Vit. E) has been reported to be an effective treatment of musculocutaneous radiation-induced fibrosis, both in an experimental model and in humans. We therefore hypothesized that any thin uterine endometrium, even when the causative agent is unknown, can be treated like iatrogenically induced fibro Design: We tested this treatment in 18 oocyte recipients whose endometrium remained thin in the late proliferative phase, despite vaginal treatment with micronized estradiol started the first day of the cycle observed. All patients received the combination of PTX and Vit. E for six months. We then evaluated the effect on endometrial thickness and the pregnancy and delivery rates. Materials/Methods: The inclusion criteria were designed to enroll oocytes recipients (n = 18) who failed to develop a preovulatory endometrial thickness of at least 6 mm after receiving vaginal micronized estradiol. The eighteen patients received a combination of PTX (800 mg/d) and Vit. E (1000 IU/d) for six months. The main outcome measurements were the change in endometrial thickness and the pregnancy and delivery rates after treatment. Results: Endometrial thickness (ETh) increased significantly (p <0.001): the mean Eth was 4.9 +/− 0.6 mm before and 6.2 +/−1.4 mm after treatment, with 72% (13/18) good responders. Five patients did not respond to the treatment. Three patients, among two after previous radiotherapy, were spontaneously pregnant, and two were pregnant after embryo transfer. Three patients did not have embryo transfer. The pregnancy rate was thus 33% and the delivery rate 27% Conclusions: Treatment by PTX and Vit. E together appears to improve the pregnancy rate in patients with a thin endometrium by increasing the endometrial thickness and improving ovarian function especially after total body irradiation Supported by: Fondation fertilité stérilité et l'INSERM.

Objective: Implantation remains an unsolved problem in reproductive medicine and is the major factor restricting the success of assisted conception in humans. There is strong evidence that locally secreted cytokines control the implantation process and can cause implantation failure. We have, however, no reliable immunological tools for the routine exploration of uterine receptivity. Leukemia Inhibitory Factor (LIF) has been documented as one of the main cytokines for implantation and increased TNF?expression has been associated with implantation failure. In the present study, we analysed LIF and TNF production in uterine flushing and by endometrial biopsies. Design: Infertile patients (n=33) underwent uterine flushing on day 26, in the late luteal phase, of two consecutive cycles, to confirm the reproducibility and hence the reliability of the data. Then after a reference uterine flushing and biopsy with measurements of the LIF and TNF concentrations and determination of the LIF index production (LPI) for 30 patients, we followed the outcome of their first consecutive cycle of either IVF or intrauterine insemination. Ten of the thirty patients became pregnant. Materials/Methods: All patients received estrogen-progestin replacement treatment to standardize the data of the endometrial evaluation. We adapted the standard method of uterine flushing previously described by inserting into the uterus an embryo transfer catheter to instill twice 1 ml of saline water which is immediately aspirated. The concentrations of LIF and TNF were measured in the flushing liquid by ELISA and bioassay, respectively. Explant cultures from the endometrial biopsy tissue made it possible to calculate the LIF production index (LPI). Results: The median concentration of LIF was 0 pg/ml (range: 0–177) and of TNF, 0 U/ml (range: 0–6.17) among those who became pregnant and 203 pg/ml (range: 0–1620) and 2.14 U/ml (range: 0–16.1 U/ml), respectively, among those who did not. The LIF concentration was significantly lower in the pregnant group (p = 0.0013). No difference was observed in the LPI between the pregnant (5.4) and the non pregnant (4.6) patients. Conclusions: A low concentration of LIF in the uterine flushing fluid at day 26 was predictive of subsequent implantation. Use of this procedure should increase the number of IVF attempts yielding successful pregnancies and also lead to corrective therapies. Supported by: Fondation fertilité stérilité.

Objective: Progress in improving early pregnancy rates depends on a better understanding of the mechanisms underlying the uterine receptivity. We therefore set up an exploration of the uterine receptivity combining the quantification/localisation of some cytokines which might control the vascular maternal network and a Doppler /ultrasound analysis. Design: We performed among 35 women with an unexplained failure of repeated in vitro fertilization attempts (UF-IVF) and in 10 control fertile women a luteal uterine standardized exploration (J20) that comprises an ultrasound evaluation, an uterine flushing for evaluation of LIF concentration by ELISA and an immunohistochemical analysis of the endometrium using anti-IL-12, anti-IL-18 (two cytokines co-inducer of interferon gamma) and anti-CD56 (a marker of Natural Killer cells) antibodies. We previously demonstrated that detection of LIF in flushing was associated with a poor pregnancy outcome. Materials/Methods: The uterine vascular scoring includes the following parameters: uterine artery pulsatility index (PI) (score: 4 pts if <3), protodiastolic notch (score: 2 pts if absent), end diastolic blood flow (score: 2 pts if present), endometrial vascularization (score: 2 pts if present). The intensity of anti-IL-12,IL-18 staining was assessed semi-quantitatively. For the NK staining (anti-CD56), image capture of all the slides allow to select three representative 10X fields in which we quantified the mean number of NK cells stained using a semi-automated method Results: The control patients exhibited a remarkably constant common immunohistochemical pattern: lack or poor anti IL-12 staining, a constant stromal anti IL-18 staining, and less than 20 NK cells/field. All the patients of this control group had a normal vascular scoring and an undetectable LIF in uterine flushing. For patients in UF-IVF, only 17% (6/35) exhibited the same immunohistochemical pattern as observed in controls. In the other ones, we could define four distinct patterns different both from the control one and within themselves: a) excess of both anti-IL-12 and anti-CD56 staining for 10 patients (28,6%) b) an isolated excess of anti-CD56 staining in 3 cases (8,6%). The two other groups had in common an absence of anti-IL18 or IL12 staining. However, profile c) was characterized by a normal anti CD56 staining in twelve cases (34,3%) and profile d) four cases (11,4%)showed an excess of such staining. These four groups had a low vascular scoring in 60%,100%, 25% and 0% respectively and a detectable LIF in 90%, 66%, 0% and 100% of the cases respectively. The concentration of LIF and the number of NK were significantly correlated (r = 0.36, p <0.02) and the vascular scoring was significantly negatively correlated to the LIF concentration in flushing (r = −0;46, p = 0.001). Conclusions: Most of the cases of so far UF-IVF are probably related to ignored local dysregulation of the local cytokine network. Ultrasound assessment and assessment of the LIF concentration by uterine flushing which are both non-invasive procedures seem to correlate with some (but not all) abnormal immunological patterns. These results suggest that we have to develop several specific molecular approaches for uterine therapy. Supported by: INSERM, fondation fertilité stérilité.

BACKGROUND:To evaluate the effect of an antifibrotic treatment by a combination of pentoxifylline (PTX) and tocopherol (vitamin E) in patients with a thin endometrium who were enrolled in an oocyte donation programme.METHODS:Eighteen oocyte recipients who failed to develop a pre-ovulatory endometrial thickness of at least 6 mm after receiving vaginal micronized estradiol were enrolled in the study. The patients received a combination of PTX (800 mg/day) and vitamin E (1000 IU/day) for 6 months. The main outcome measurements were the change in endometrial thickness and the pregnancy and delivery rates after treatment.RESULTS:Endometrial thickness increased significantly (P <0.001), with a mean of (+/-SD) 4.9 +/-0.6 mm before and 6.2 +/- 1.4 mm after treatment, with 72% (13/18) of patients being good responders. Five patients either did not respond to the treatment or responded only slightly. Three patients, of which two had received previous radiotherapy, became spontaneously pregnant, and two became pregnant after embryo transfer. Three patients did not have embryo transfer. A total of four babies were delivered. The pregnancy rate was thus 33% and the delivery rate 27%.CONCLUSION:Treatment by combination of PTX and vitamin E appears to improve the pregnancy rate in patients with a thin endometrium by increasing the endometrial thickness and improving ovarian function. This was especially noticeable in patients who had previously received total body irradiation.