Follicular flushing and in vitro fertilization outcomes in the poorest responders: a randomized controlled trial

Transcription

1 Human Reproduction, Vol.28, No.11 pp , 2013 Advanced Access publication on September 5, 2013 doi: /humrep/det350 ORIGINAL ARTICLE Infertility Follicular flushing and in vitro fertilization outcomes in the poorest responders: a randomized controlled trial Evelyn Mok-Lin*, Anate Aelion Brauer, Glenn Schattman, Nikica Zaninovic, Zev Rosenwaks, and Steven Spandorfer The Ronald O. Perelman and Claudia Cohen Center for Reproductive Medicine, Weill Cornell Medical College, 1305 York Ave, 7th Floor, New York, NY 10021, USA *Correspondence address. Tel: ; evelyn.moklin@gmail.com Submitted on May 21, 2013; resubmitted on July 30, 2013; accepted on August 6, 2013 study question: Does follicular flushing during oocyte retrieval improve the number of oocytes retrieved in the poorest responders? summaryanswer: Follicular flushing in the poorest responders does not increase the number of oocytes retrieved and may result in lower implantation and clinical pregnancy rates. what is known already: Although previous studies have shown no beneficial effect of follicular flushing in normal responders, no study has demonstrated a detrimental effect and many IVF centers continue to perform this technique in poor responders. Data on follicular flushing in this patient group are limited, with no randomized trial to date assessing its utility in the poorest responders. study design, size, duration: This randomized controlled trial compared the effects of follicular flushing and direct aspiration on IVF outcomes in the poorest responders, defined as having four or fewer follicles 12 mm on the day of hcg administration. Fifty patients were randomized during the 12-month enrollment period. participants/materials, setting, methods: The patients were treated at an academic fertility center at Weill Cornell Medical College, New York. main results and the role of chance: Fifty women were randomized to follicular flushing (n ¼ 25) or direct aspiration (n ¼ 25). One patient in the direct aspiration group was canceled prior to oocyte retrieval for premature ovulation and was included in the intent-to-treat analysis. There was no difference in the number of oocytes retrieved with a median (IQR) of 4 (2 6) in the aspiration group versus 3 (2 5) in the flushing group (95% CI: 20.78, 1.98; P ¼ 0.41). Patients who underwent follicular flushing had significantly fewer embryos transferred {1.7 [standard deviation (SD) 0.6] versus 2.5 (SD 1.2), P ¼ 0.03}, a lower implantation rate (5.3 versus 34.2%, P ¼ 0.006) and a lower clinical pregnancy rate (4 versus 36%, P ¼ 0.01). The difference in pregnancy rates remained significant after adjusting for embryos transferred. limitations, reasons for caution: Findings, including results for secondary outcome measures, may not be generalizable to natural IVF cycles as these were excluded from the study. wider implications of the findings: This is the first randomized trial to evaluate the utility of follicular flushing in the poorest responders, and the first to demonstrate a potentially detrimental effect of flushing on IVF outcomes. study funding/competing interest(s): None. trial registration number: NCT Key words: flushing / IVF / poor responders & The Author Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please journals.permissions@oup.com

2 Follicular flushing in the poorest responders 2991 Introduction Performed for nearly three decades, ultrasound-guided transvaginal follicular aspiration is the standard of care for oocyte retrieval (Wikland et al., 1985; Gembruch et al., 1988; Wiseman et al., 1989; Ludwig et al., 2006). Early in the transition from laparoscopy to the transvaginal approach, double-lumen needles were developed that allow for aspiration of the follicle through one lumen, reexpansion by fluid injection through a second lumen and subsequent reaspiration without removing the needle from the follicle. Potential benefits of follicular flushing include the ability to overcome oocyte retention within the follicle or the retrieval collection system. In the past, flushing s routine performance was supported by prospective, non-randomized studies that demonstrated an increase in the number of oocytes retrieved compared with those who underwent direct aspiration with a single-lumen needle (el Hussein et al., 1992; Waterstone and Parsons, 1992; Bagtharia and Haloob, 2005). Recent reviews, drawing on data from randomized trials with normally responding patients, have recommended against follicular flushing in these patients as it prolongs operative time without an increase in oocyte yield (Scott et al., 1989; Kingsland et al., 1991; Tan et al., 1992; Wongtra-Ngan et al., 2010; Haydardedeoglu et al., 2011; Levy et al.,2012). Most institutions no longer perform follicular flushing except in the treatment of poor responders, for whom maximization of oocyte recovery may be critical to a successful outcome. The only randomized trial that assessed the utility of flushing in poor-responding patients included 15 women per group with 4 8 follicles 12 mm and at least 2 follicles 16 mm on the day of hcg administration (Levens et al., 2009). This study thus excluded patients with fewer than four follicles, the poorest responders who may potentially benefit most from flushing. To address this lack of data on the procedure s utility in this challenging patient group, we performed a randomized trial of follicular flushing compared with direct aspiration in the poorest responders. Materials and Methods Women undergoing IVF at the Ronald O. Perelman and Claudia Cohen Center for Reproductive Medicine of Weill Cornell Medical College were identified. The poorest-responding patients, as defined by four or fewer follicles 12 mm on the day of hcg administration, were eligible for this study. Patients without a planned, fresh embryo transfer, patients undergoing natural IVF and patients whose cycles were canceled prior to hcg administration were ineligible. Those who were offered enrollment or were randomized in a previous cycle were ineligible. Informed consent was obtained after the study procedures were explained to patients and all questions answered. The study was approved by Weill Cornell Medical College s Institutional Review Board and registered at clinicaltrials.gov (NCT ). Specific IVF protocols and starting gonadotrophin dosages were selected based on age, weight, measures of ovarian reserve and previous response to controlled ovarian hyperstimulation (COH). The majority of poor responders at our center utilized a gonadotrophin-releasing hormone (GnRH) antagonist protocol with luteal estrogen priming. These patients were placed on a 0.1 mg estradiol patch every other day beginning 8 10 days after a LH surge, followed by COH on Day 2 of menses. Other protocols included the use of a GnRH antagonist without priming, 5 days of oral clomiphene citrate or letrozole, and a daily subcutaneous micro-dose (40 mg) of leuprolide with COH. In all patients, COH was performed in a standard fashion using recombinant FSH and human menopausal gonadotrophins. Patients were monitored with serial serum estradiol concentrations and transvaginal ultrasounds. hcg IU was administered intramuscularly when one to two follicles 17 mm were present, and transvaginal oocyte retrieval was performed 35 h later. Eligible patients were offered enrollment and randomized on the day of hcg administration to either the direct aspiration group or the flushing group. Treatment allocation was performed using a computer-generated randomization sequence constructed by a statistician in randomly generated block sizes of two to ensure equal distribution between groups. Allocation was performed by the research team and concealed using sequentially numbered, opaque envelopes that were opened after obtaining informed consent for study participation. All patients were blinded to the randomization for the duration of the study. By necessity, the providers performing the procedure were notified of the treatment allocation on the day of the retrieval. In the direct aspiration group, a 16-gauge single-lumen oocyte retrieval needle (Echotip w ovum aspiration needle; Cook Medical; Bloomington, IN, USA) was used to aspirate the follicles with transvaginal ultrasound guidance. In the flushing group, each aspirated follicle was flushed up to four times using a manually pressed syringe with 5 ml of culture media warmed to 378C and re-aspirated using a 16-gauge double-lumen needle (Echotip w Double Lumen Aspiration Needle; Cook Medical). All retrievals were performed by one of two physicians in a single center to minimize surgical variability. Both physicians were experienced in flushing and direct aspiration techniques. The oocytes and embryos were cultured in the same type of media as that used for follicular flushing. Determination of need for ICSI and selection of embryos for transfer were performed by embryologists blinded to the allocation scheme. Grading of the cleavage embryos was performed according to the Atlas of Human Blastocysts (Veeck and Zaninovic, 2003). All embryos were transferred on Day 3 and transferring physicians were blinded to the assigned intervention. The primary end-point was the number of oocytes retrieved. A review of poor-responding patients in our center who met the inclusion criteria in the 6 months prior to initiation of this study demonstrated a standard deviation of 1.4 oocytes. An a priori sample size calculation determined that 25 subjects in each group would be required to detect a one-oocyte difference between groups with 80% power (b ¼ 0.2, a ¼ 0.05). Secondary outcomes included anesthesia time, procedure time, number of mature oocytes, number of embryos transferred, implantation rate, clinical pregnancy rate and live birth rate. The total anesthesia time was defined as the time from initiation of propofol until transport to the recovery area. The total procedure time was recorded from the time of transvaginal insertion of the retrieval needle into the first ovary to the removal of the needle from the second ovary. Baseline characteristics were collected and cycle outcomes compared between the patients who received direct aspiration versus follicular flushing in an intent-to-treat fashion. Anti-Müllerian hormone (AMH) levels below the limit of detection (LOD) of 0.16 ng/ml were accounted for as one-half the LOD in the analysis. Statistical analyses included the Mann Whitney U-test for continuous variables and two-sided Fisher s exact test for categorical variables. Relative risks (RR) were calculated for clinical pregnancy and live birth rates asthe ratio of the probability of the outcome occurring in the direct aspiration group versus the flushing group. A logistic regression model including the number of embryos transferred (continuous independent variable) and treatment allocation (dichotomous independent variable where direct aspiration was coded 1 and flushing 0 ) was utilized in the secondary analysis of clinical pregnancy and live birth rates. Odds ratios (OR) were reported in these models as the ratio of the odds of the outcome occurring in the direct aspiration group versus the flushing group. P, 0.05 was deemed statistically significant. Stata Statistical Software version 11 (StataCorp, College Station, TX, USA) was used to perform all data analyses. Results During the 12-month enrollment period, 78 patients met the inclusion criteria and were offered enrollment; 50 patients consented and were

3 2992 Mok-Lin et al. randomized (Fig. 1). Each patient contributed a single cycle of treatment. There were no differences in demographic or cycle characteristics between the groups (Table I). Patients in the direct aspiration and flushing groups were on average 39.5 [standard deviation (SD) 3.2] and 38.2 (SD 4.3) years old, respectively, ranging from 31 to 44 years old. Mean AMH levels were 0.45 (SD 0.39) and 0.31 (SD 0.42) ng/ml, including 12 (24%) patients with undetectable AMH levels. Patients received an average of 4551 (SD 2290) IU of gonadotrophins and reached a mean peak estradiol level of 721 (SD 246) pg/ml. There was no difference in the type of IVF protocol utilized. The two groups had similar mean number of follicles 12 mm [3.3 (SD 0.8) versus 3.0 (SD 0.9), P ¼ 0.19] and total follicles 1 mm [5.4 (SD 2.5) versus 4.7 (SD 2.0), P ¼ 0.49] on the day of hcg administration. Twenty-two percent of the patients had one or two follicles on the day of hcg. There were no statistically significant differences in the ages or total motile sperm count of patients male partners. A total of 208 oocytes were retrieved 112 from the direct aspiration group and 96 from the flushing group. One patient in the direct aspiration group was canceled immediately prior to the planned retrieval due to premature ovulation and was included in the intent-to-treat analysis. There were no cancelations in the flushing group and no drop-outs in either group. All patients who underwent retrieval had oocytes retrieved. There was no difference in the primary end-point: patients who underwent direct aspiration had a median (IQR) of 4 (2 6) oocytes retrieved compared with 3 (2 5) oocytes among patients who underwent follicular flushing (95% CI: 20.78, 1.98; P ¼ 0.41). The number of mature oocytes, maturity rate and what we termed mature oocyte recovery rate calculated as the number of mature oocytes/follicles 12 mm on the day of hcg were also similar between the groups (Table II). These outcomes were similar whether evaluated per cycle started (intent-to-treat) or per retrieval performed. The retrieval procedure time was higher among those who underwent follicular flushing, with an estimated increase of 2.3 min [7.0 (SD 2.5) versus 4.7 (SD 1.7), P ¼ 0.001]. The anesthesia time was increased by an average of 2.6 min in the flushing group compared with patients who underwent aspiration alone [16.9 (SD 3.2) versus 14.3 (SD 2.5), P ¼ 0.003]. There were no operative complications or adverse events in either group. Five patients in each group did not undergo an embryo transfer. All five in the flushing group were due to failure of fertilization. In the direct aspiration group, one patient was canceled prior to retrieval as noted above, three had no fertilizations and one had no embryo development. There were no significant differences in the proportion of ICSI or fertilization rate between the two groups. Among those who did have an embryo transfer, all received embryos of similar quality on Day 3. However, direct aspiration patients had more cleavage embryos available for transfer [2.8 (SD 1.2) versus 1.9 (SD 0.7), P ¼ 0.01] and more Table I Demographic and cycle characteristics among poor responders who received direct aspiration versus follicular flushing at oocyte retrieval. Direct aspiration Flushing (n 5 25) (n 5 25)... Age (years) 39.5 (3.2) 38.2 (4.3) BMI 24.6 (3.6) 23.9 (3.0) AMH (ng/ml) 0.45 (0.39) 0.31 (0.42) Previous cycles 1 (0 4) 2 (1 3) Gravity 1 (0 2) 1 (0 2) Parity 0 (0 1) 0 (0 0) Diagnosis Diminished ovarian reserve 23 (92%) 22 (88%) Endometriosis 1 (4%) 1 (4%) Male factor 1 (4%) 2 (8%) Starting gonadotrophin dose (IU) 441 (169) 409 (153) Total gonadotrophins (IU) 4788 (2303) 4313 (2276) Days of stimulation 11.2 (2.1) 10.8 (2.7) Peak E2 (pg/ml) 752 (260) 690 (231) Follicles 12mm 3.3 (0.8) 3.0 (0.9) Follicles 1mm 5.4 (2.5) 4.7 (2.0) Male age 41.3 (7.3) 42.6 (6.4) Total motile sperm (10 6 ) 3.6 (3.8) 3.0 (4.1) Figure 1 Diagram illustrating the flow of patients through the study. Randomization did not produce any unexpected differences in baseline characteristics between groups. Continuous data presented as mean [standard deviation (SD)] for normally distributed data or median (IQR) for non-normally distributed data; categorical data presented as number (%).

4 Follicular flushing in the poorest responders 2993 Table II Cycle outcomes among poor responders who received direct aspiration versus follicular flushing at oocyte retrieval. Direct aspiration (n 5 25) Flushing (n 5 25) P-value... Cancelation 1 (4%) Oocytes 4 (2 6) 3 (2 5) 0.41 Mature oocytes 3.3 (1.9) 2.5 (1.3) 0.18 Maturity rate 85.7 (60 100) 80.0 (60 100) 0.65 Mature oocyte recovery rate a 95.8 (44.0) 86.1 (37.5) 0.46 Fractured zona pellucida ICSI 17 (68%) 21 (84%) 0.32 Fertilization rate 81.7 ( ) 66.7 ( ) 0.18 Total cleavage embryos 2.8 (1.2) 1.9 (0.7) 0.01 Embryos transferred 2.5 (1.2) 1.7 (0.6) 0.03 Embryo grade b 2.2 (0.6) 2.0 (0.4) 0.27 Embryos cryopreserved 0.16 (0.55) Procedure time (min) 4.7 (1.7) 7.0 (2.5) Anesthesia time (min) 14.3 (2.5) 16.9 (3.2) Continuous data presented as mean (SD) for normally distributed data or median (IQR) for non-normally distributed data; categorical data presented as number (%). a Mature oocyte recovery rate calculated as (MII oocytes/follicles 12 mm) 100%. b Cleavage embryos graded from 1 to 5, with 1 representing the best quality and 5 the worst. Figure 2 Implantation, pregnancyand live birth rates among poor responders who received direct aspiration (n ¼ 25) versus follicular flushing (n ¼ 25) at oocyte retrieval. embryos transferred [2.5 (SD 1.2) versus 1.7 (SD 0.6), P ¼ 0.03]. Two patients in the direct aspiration group each had two embryos cryopreserved compared with no frozen embryos in the flushing group (P ¼ 0.15). Patients in the aspiration group had a higher mean implantation rate of 34.2% (SD 41.8) compared with 5.3% (SD 22.9) in the flushing group (P ¼ 0.006), as well as a 9-fold increase in clinical pregnancy rates per cycle started (36 versus 4%; RR 9, 95% CI: 1.23, 65.85; P ¼ 0.01) and per embryo transfer (45 versus 5%; RR 9, 95% CI: 1.25, 64.58; P ¼ 0.008; Fig. 2). This increase in clinical pregnancy rates in the aspiration group remained significant after adjusting for the number of embryos transferred (OR 22.5, 95% CI: 2.21, ; P ¼ 0.009). Additionally, there was a non-significant increase in higher live birth rates in the direct aspiration group (20% versus 4% per cycle; RR 5, 95% CI: 0.63, 39.79; P ¼ 0.19; 25% versus 5% per embryo transfer; RR 5, 95% CI: 0.64, 39.05; P ¼ 0.18), which persisted after controlling for embryos transferred (OR 9.3, 95% CI: 0.88, 98.63, P ¼ 0.06).

5 2994 Mok-Lin et al. Discussion This is the first randomized controlled trial to evaluate the utility of follicular flushing in the poorest responders with one to four follicles at the time of hcg administration. While it is commonly believed that flushing could potentially maximize the number of oocytes retrieved in this patient group, we found no difference in the oocyte number among those who received flushing compared with direct aspiration. Our study is also the first to demonstrate a negative effect of follicular flushing on IVF outcomes. While prior randomized controlled trials (RCTs) and meta-analyses found no benefit of flushing on the number of oocytes retrieved, they also did not demonstrate any adverse effects aside from prolonged procedure time (Wongtra-Ngan et al., 2010; Levy et al., 2012). Conversely, patients in our study who underwent follicular flushing had fewer embryos transferred, lower implantation rates and lower clinical pregnancy rates that remained significant even after adjusting for the embryos transferred. These were unexpected findings in light of previous data that demonstrated the relative safety of the flushing technique. Of note, while the pregnancy rates were low in the flushing group, patients who were directly aspirated with a singlelumen needle had implantation rates of 34% and clinical pregnancy rates of 36 45%. These rates are quite encouraging given that these patients were very poor responders, many of whom likely would have been canceled. Increased intra-follicular pressure at retrieval is one potential etiology accounting for the significantly lower implantation and pregnancy rates in the flushing group. While the pressure at which we aspirated the follicles was strictly maintained at 80 mmhg, the pressure generated within the follicle from fluid injection is unknown. Although none of the oocytes retrieved with flushing had evidence of a fractured zona pellucida, it is possible that the additional pressure may have impacted oocyte quality via other mechanistic effects, which may explain why flushed patients had fewer embryos transferred despite a similar number of oocytes retrieved. Oocyte quality may have also been affected by the increased procedure and anesthesia times, although there was a,3 min difference in anesthesia time between the groups. Another possible etiology is the presence of flushing fluid in the pelvis that may have altered the uterine milieu at the time of implantation. This is less likely, as the fluid used for flushing was the same type of media that the oocytes and embryos were cultured in, the flushed follicle was thoroughly re-aspirated, and 3 days passed before the embryo transfer was performed. However, that does not preclude the potential for retention of a residual amount. Additionally, serial flushing could have impacted granulosa cell number and function, although secondary analysis of luteal estradiol levels did not demonstrate a significant difference between the groups. While none of these variables were unique to this study compared with prior studies on follicular flushing, it is likely that one or more of these factors played a small role in the difference in outcomes and their cumulative effect was amplified in the oocytes and embryos of these older, poor-responding women with less buffering capability. Strengths of the study include our patient selection criteria. This is the first randomized trial to assess the utility of follicular flushing in the poorest responders with four or fewer follicles 12 mm. In addition to having the worst response with the lowest mean number of oocytes retrieved, our study population was also older with a mean age of 38.2 years in the flushing group and 39.5 years in the direct aspiration group. This is compared with 36.2 and 37.1 years in the only other available RCT that specifically targeted poor responders (Levens et al., 2009), and 30.8 and 30.6 years in the largest randomized trial to date assessing flushing in normal responders (Haydardedeoglu et al., 2011). In addition, we performed an a priori sample size calculation to detect a one oocyte difference and our study was adequately powered for the primary outcome. The majority of previous RCTs did not report an a priori calculation or were unable to enroll the necessary number of patients (Scott et al., 1989; Kingsland et al., 1991; Tan et al., 1992; Haydardedeoglu et al., 2011). Our study was limited by a relatively small sample size of 50 poorresponding patients from a tertiary referral center, and findings may not be generalizable to all poor responders in other centers or to patients undergoing natural or semi-augmented IVF cycles. We excluded natural IVF cycles as these patients had an intentionally low number of follicles at retrieval and were not necessarily poor responders. While some observational data favor flushing for natural- and minimal-stimulation IVF cycles (Lozano et al., 2006; Mendez Lozano et al., 2008), there is no randomized trial supporting flushing s use in this context to date. An additional limitation and potential source of bias is the necessary unblinding of the retrieval physicians. While these providers were different from the enrollment physicians and were unaware of the allocation until immediately prior to the procedure, it is possible that their knowledge of the treatment may have contributed to the findings. However, our study hypothesis was such that if a bias did exist, we would have expected it to favor flushing. The results of this RCT in the poorest responders found no benefit to follicular flushing on the number of oocytes retrieved and a detrimental effect on implantation and pregnancy rates. Based on these results and those of prior studies, one should exercise caution in the use of follicular flushing in poor responders. Additional studies are needed to determine optimal techniques for maximizing successful outcomes in this difficult patient population. Acknowledgements The authors thank the attending physicians, fellows, embryologists, nurses and research staff at The Ronald O. Perelman and Claudia Cohen Center for Reproductive Medicine. Authors roles E.M.-L. was involved in the substantial contribution to conception and design, acquisition of data, and analysis and interpretation of data, in addition to drafting the article. A.A.B., G.S. and N.Z. were involved in acquisition of data and revising the content. Z.R. was involved in interpretation of data and revising the article critically for content and final manuscript approval. S.S. was involved in conception and design, acquisition and interpretation of data and in revising the article critically for content and final manuscript approval. Funding No external funding was either sought or obtained for this study. Conflict of interest The authors have nothing to disclose.

6 Follicular flushing in the poorest responders 2995 References Bagtharia S, Haloob AR. Is there a benefit from routine follicular flushing for oocyte retrieval? J Obstet Gynaecol 2005;25: el Hussein E, Balen AH, Tan SL. A prospective study comparing the outcome of oocytes retrieved in the aspirate with those retrieved in the flush during transvaginal ultrasound directed oocyte recovery for in-vitro fertilization. Br J Obstet Gynaecol 1992;99: Gembruch U, Diedrich K, Welker B, Wahode J, van der Ven H, Al-Hasani S, Krebs D. Transvaginal sonographically guided oocyte retrieval for in-vitro fertilization. Hum Reprod 1988;3: Haydardedeoglu B, Cok T, Kilicdag EB, Parlakgumus AH, Simsek E, Bagis T. In vitro fertilization-intracytoplasmic sperm injection outcomes in singleversus double-lumen oocyte retrieval needles in normally responding patients: a randomized trial. Fertil Steril 2011;95: Kingsland CR, Taylor CT, Aziz N, Bickerton N. Is follicular flushing necessary for oocyte retrieval? A randomized trial. Hum Reprod 1991;6: Levens ED, Whitcomb BW, Payson MD, Larsen FW. Ovarian follicular flushing among low-responding patients undergoing assisted reproductive technology. Fertil Steril 2009;91: Levy G, Hill MJ, Ramirez CI, Correa L, Ryan ME, DeCherney AH, Levens ED, Whitcomb BW. The use of follicle flushing during oocyte retrieval in assisted reproductive technologies: a systematic review and meta-analysis. Hum Reprod 2012;27: Lozano DH, Fanchin R, Chevalier N, Feyereisen E, Hesters L, Frydman N, Frydman R. Optimising the semi natural cycle IVF: the importance of follicular flushing. J Indian Med Assoc 2006;104: Ludwig AK, Glawatz M, Griesinger G, Diedrich K, Ludwig M. Perioperative and post-operative complications of transvaginal ultrasound-guided oocyte retrieval: prospective study of.1000 oocyte retrievals. Hum Reprod 2006;21: Mendez Lozano DH, Brum Scheffer J, Frydman N, Fay S, Fanchin R, Frydman R. Optimal reproductive competence of oocytes retrieved through follicular flushing in minimal stimulation IVF. Reprod Biomed Online 2008;16: Scott RT, Hofmann GE, Muasher SJ, Acosta AA, Kreiner DK, Rosenwaks Z. A prospective randomized comparison of single- and double-lumen needles for transvaginal follicular aspiration. J In Vitro Fert Embryo Transf 1989;6: Tan SL, Waterstone J, Wren M, Parsons J. A prospective randomized study comparing aspiration only with aspiration and flushing for transvaginal ultrasound-directed oocyte recovery. Fertil Steril 1992;58: Veeck LL, Zaninovic N. An Atlas of Human Blastocysts, 1st edn. FL, USA: CRC Press, Waterstone JJ, Parsons JH. A prospective study to investigate the value of flushing follicles during transvaginal ultrasound-directed follicle aspiration. Fertil Steril 1992;57: Wikland M, Enk L, Hamberger L. Transvesical and transvaginal approaches for the aspiration of follicles by use of ultrasound. Ann N Y Acad Sci 1985; 442: Wiseman DA, Short WB, Pattinson HA, Taylor PJ, Nicholson SF, Elliott PD, Fleetham JA, Mortimer ST. Oocyte retrieval in an in vitro fertilizationembryo transfer program: comparison of four methods. Radiology 1989; 173: Wongtra-Ngan S, Vutyavanich T, Brown J. Follicular flushing during oocyte retrieval in assisted reproductive techniques. Cochrane Database Syst Rev 2010:CD