Forty-five infertile women undergoing in vitro fertilization treatment were managed by uterine cavity injection of day 2 embryo culture supernatant before day 3 embryo transfer. No statistically significant differences were found in implantation and pregnancy rates when compared with control group.

Study design and sample size This was a randomized clinical trial involving ninety patients who were undergoing conventional IVF or intracytoplasmic sperm injection (ICSI) treatment at our Reproductive Medical Center. The subjects were recruited and studied for a period of 6 months between January and June 2008. The study was approved by the Ethics Committee of Shen-Zhen Maternity and Child Healthcare Hospital. Subjects all signed an informed consent form.

According to the guideline made by the Ethics Committee of Shen-Zhen Maternity and Child Healthcare Hospital, we had better accomplish the study within six months. We calculated that, in our center, a sample size of 100 or more subjects would need to be recruited over about a 6 months period. We ended up with 90 patients in this study.

Subject Ninety infertile women, aged 23-39 years, with an infertile duration of 3-8 years, were enrolled from our infertility treatment clinic between January and June 2008. The selection criteria for participants were only those women who were first going to undergo IVF or ICSI treatment. Women who were treated with natural cycle IVF or had no appropriate embryo for ET were excluded. The infertility factors were identified as the following: tubal (n=54), unexplained (n=3), male n=22), endometriosis (n=9), others (n=2). The selection and enrollment for each subject was decided by three same physicians. All subjects were informed that this procedure was a new technique, and that its effectiveness has not been confirmed, although an optimal outcome in patients with blastocyst transfer was obtained by using this technique for the same purpose (6). During the selection procedure, four eligible subjects opted not to participate when counseled about the uncertain of effectiveness, and other two subjects declined to randomize. Finally, 90 eligible patients were enrolled.

Randomization First, ninety random numbers generated by means of computer were randomly divided into study group and control group. Then, all subjects in an enrolled date order were arranged according to the sequence of random numbers. So, ninety subjects were randomly divided into the study group and control group. The random allocation sequence was concealed in a closed and dark-colored envelope until the day of oocyte pick-up. Randomization occurred after patients agreed to participate in the study (Figure 1). All women of two groups underwent a conventional IVF or ICSI procedure, but in study group, the patients were added with a uterine cavity injection of day 2 embryo culture supernatant. The IVF or ICSI procedure was performed by same physicians and laboratory technicians. The treating sequence of subjects was as same as the enrolling sequence. The study was not blinded, because the patients as well as the clinicians were aware of the study content.

IVF or ISCI procedure and injection technique of embryo culture supernatant Controlled ovarian hyperstimulation was performed using a variety of protocols, as previously published elsewhere (8). A down-regulation of the pituitary was realized during triptorelin (Triptorelin Acetate Injection, Ferring, Germany) administration S.C., 0.1 mg/d×7, started from day 21 of the previous cycles. Ovarian stimulation started from day 3 of stimulated cycle by using recombinant follicle stimulating hormone (rFSH, Gonal-F, Serono, Switzerland), 225-300IU daily S.C.. The follicle development was monitored by transvaginal ultrasound scan (TVS). 250ug of HCG (Ovitrelle; Serono, Madrid) was administered by I.M. when one to two leading follicle(s) reached a mean diameter of 18 mm. In general, transvaginal oocyte retrieval was performed after 36 h of HCG injection.

The media used for embryo culture was IVF-60. Briefly, after oocyte retrieval (day 0), cumulus-enclosed oocytes were incubated in IVF-60. Semen was prepared with a 45-90 discontinuous gradient. The resulting pellet was washed in IVF-60. For normal IVF, 100,000 to 150,000 sperm/100ul were used for each oocyte. The maturation stage of oocyte was assessed for IVF 15-18 hour after insemination (day 1). Cumulus and corona cells were removed using drawn pipettes. When ICSI was performed, oocytes were denuded with HYASE. Each mature oocyte was placed in a 6-ul droplet of GAMETE-20. The partner's sperm was placed in a 6-ul droplet of ICSI-100. All droplets were overlaid with OVOIL. ICSI was performed using a inverted microscope with RI micromanipulators (INTEGRA Micromanipulation System, Research Instruments, UK). The CO2 incubators (Forma, USA) with 6% CO2 gas were used for embryo incubation. An embryo scoring system based on cleavage rate and morphologic features was used for the evaluation of embryo quality, as previously described elsewhere (9). Briefly, grade 1 is the best score and grade 5 is the worst. For example, a grade 1 embryo has blastomeres of equal size with no cytoplasmic fragmentation. A grade 2 embryo has symmetric blastomeres, but a small degree of cytoplasmic fragmentation involving ≤10% of the embryo's surface. Grades 3 through 5 demonstrate progressive embryo cytoplasmic fragmentation and variable cell symmetry.

On the day 2, the embryos were translated from the old dish to a new dish with fresh media and cultured continuously. In study group, the embryo culture supernatant at the old dish were aspirated by using embryo transfer catheter (COOK-5000, Australia) and injected into the patient's uterine cavity. ET was performed on the third day of embryo culture. Two to three embryos were selected to transfer for each woman. The number of embryos selected for transfer was based on morphologic criteria and some additional factors, such as duration and cause of infertility and age of the patient. Pregnancy was determined by carry out a urine pregnancy test after 14 days of ET. The embryo implantation, the number of gestational sacs, and embryo heart pulsation were determined by TVS in the seventh week of pregnancy. Except the day 2 embryo culture supernatant injection, all of the above-mentioned clinical and laboratory procedures were identical between the study group and control group.

Statistical analysis Baseline characteristics of the patients were expressed as mean ±standard deviation or frequencies and proportion (%). The pregnancy and implantation rates were calculated based on the cases of pregnancy and number of gestational sacs. A Chi-Square Test was used to compare the implantation and conception proportions between two groups. The Statistics Package for Social Science (SPSS 13.0; SPSS Inc.) was used for statistical analyses. Statistical significance was set at P<0.05.

Inclusion Criteria:

• The participants were only those women who were first going to undergo IVF or ICSI treatment.

Exclusion Criteria:

• The women who were treated with natural cycle IVF or had no appropriate embryo for ET were excluded.