The surplus human embryo: its potential for growth, blastulation, hatching, and human chorionic gonadotropin production in culture

FERTILITY AND STERILITY

Vol. 51, No.6, June 1989

Printed in U.S.A.

Copyright" 1989 The American Fertility Society

The surplus human embryo: its potential for growth, blastulation, hatching, and human chorionic gonadotropin production in culture Alexander Lopata, M.B., Ph.D.* D. Lachlan Hay, Ph.D. Department of Obstetrics and Gynecology, University of Melbourne, Reproductive Biology Unit and Department of Pathology, Royal Women's Hospital, Carlton, Victoria, Australia

Early embryos that were unsuitable for transfer to patients or for cryopreservation were cultured either in a human tubal fluid (HTF) or a minimum essential medium (MEM). A significantly higher proportion of embryos developed to blastocysts in MEM (26.8%) than in HTF (14.5%). Approximately similar proportions of embryos formed blastocysts in MEM in the presence or absence of serum. The rate of embryo growth to blastocysts was similar in all media. Blastocyst hatching occurred in MEM + or - serum, but it failed to occur in HTF with serum. Released human chorionic gonadotropin (hCG) from hatched and intrazonal blastocysts was detected by day 8 after fertilization. The mean amount of hCG produced by day 14 was 19,500 miU from hatched and 1,550 miU from intrazonal blastocysts. Serum stimulated the output ofhCG. Fertil Steril51:984, 1989

Despite nearly 20 years' experience in human in vitro fertilization (IVF) and embryo culture, we still have only a rudimentary understanding of the factors that influence embryo growth. One of the causes for our lack of progress is the practice of transferring embryos to the patient's genital tract at an early stage of development. Consequently, the growth potential of over 90% of embryos that fail to produce pregnancies remains unknown. Similarly, there is limited knowledge of the ability of human blastocysts to escape from their zona pellucidas, their propensity for attachment and growth, and the relationship between these events and human chorionic gonadotropin (hCG) secretion. In this study, we compared the efficacy of two growth media in supporting human preimplantation growth and blastocyst hatching. The simpler culture medium was human tubal fluid (HTF), 1 and Received November 7, 1988; revised and accepted February 16,1989. · * Reprint requests: Alexander Lopata, M.B., Ph.D., Department of Obstetrics and Gynecology, Royal Women's Hospital, Carlton, Victoria 3053, Australia.

984

Lopata and Hay Human embryo development

the complex medium was the alpha modification of minimum essential medium (MEM). 2•3 We both monitored the pattern of embryo growth and evaluated hCG production.

MATERIALS AND METHODS Patients

Observations were made on embryos assessed as poor quality and surplus to our IVF and embryo transfer program at The Royal Women's Hospital from March 1987 to May 1988 inclusive. The hospital's Research and Ethics Committees approved the project for monitoring surplus embryos since it did not involve destructive procedures on surviving embryos. The Committees advised us to maintain surplus embryos in culture until they succumbed and then to discard the nonviable tissues. Patient selection, counseling, and management for IVF and ET have been described elsewhere. 4 All women and their husbands signed consent forms that described the therapy and associated procedures. Fertility and Sterility

IVF Procedures

Oocytes were collected after hyperstimulation with clomiphene citrate (CC) and human menopausal gonadotropin (hMG) as described previously.4 Briefly, CC (100 mg/day) was given for 5 days, starting 10 days before the cycle midpoint, and hMG (150 IU /day) was commenced 1 day later. The hMG was stopped when plasma estradiol exceeded 4 nmol/1, and the largest follicle visible by ultrasonography was > 16 mm, with at least three follicles > 10 mm in diameter. After variable intervals, patients were given 5,000 IU of hCG. Oocytes were aspirated transvaginally under ultrasound control approximately 34 hours after hCG injection. Preincubation of oocytes, preparation of spermatozoa, and in vitro insemination have been described previously. 4•5 Human tubal fluid supplemented with 10% heat-inactivated human cord serum (HCS) was used in 18-hour insemination culture, and embryo growth was maintained for an additional 24 to 28 hours in HTF with 1% HCS. Embryos then were assessed and up to four embryos were transferred into the patient's uterus, providing embryos arose from bipronucleate fertilizations. Embryo Assessment

Embryos were assessed as being suitable for transfer or for freezing on the second day after insemination (i.e., at the 2-cell to 4-cell stage). Good quality embryos contained uniform blastomeres and minimal or no cytoplasmic fragmentation. Fair quality embryos had irregular or unequal blastomeres and some cell fragmentation. Poor quality embryos had more severe cell irregularities and more numerous fragments. Embryos arising from tripronucleate oocytes were never transferred. In cases where five or more embryos were obtained, four embryos judged to be of good quality on morphologic criteria were transferred. Embryos that were not transferred were cryopreserved for transfer to the oocyte donor in subsequent cycles, providing the couple had given consent and the embryos were morphologically suitable for freezing. Those remaining that were of poor quality, or those in excess because consent for cryopreservation was not given ( <5% of residual embryos), or those derived from tripronucleate oocytes were classified as surplus.

Flow Laboratories (Irvine, Scotland). The HTF was prepared as described above and embryos were cultured either in the medium in which they originated in the IVF laboratory (original HTF group) or in freshly equilibrated HTF + 1% HCS (fresh HTF group). The MEM was supplemented with 25 mmoljl sodium bicarbonate, penicillin G 10 mg/ 100 ml and streptomycin sulphate 5 mg/100 ml, all from the Sigma Chemical Company (St. Louis, MO). The osmolarity of the medium was adjusted to 280 mOsm/kg and its pH to 7.35 when equilibrated with 5% C0 2 • The embryos were cultured in this medium supplemented with 1% HCS (MEM + 1% HCS group) or in MEM without serum (MEM alone group). Individual embryos were cultured in 1-ml aliquots of medium in 4-well Nunclon plates (Nunc, Kamstrup, Denmark). On the second day after insemination, surplus embryos, arising from bipronucleate eggs, were allotted to one of the culture groups. Ten successive surplus embryos were assigned to each medium in the chronologie sequence in which they became available, without any selection, and this allotment of ten embryos in a row to the same medium (obtained from two to five patients) continued in rotation through the four culture groups. The data within each group was combined and the proportion of embryos attaining the blastocyst stage in each medium was analyzed by means of a 2 X 2 chisquare incorporating a Yates correction. Initially, embryos arising from tripronucleate eggs were distributed between culture media groups as described above. When it became apparent that their potential for forming blastocysts was very poor, however, most of these embryos were retained in HTF + 1% HCS. In most instances, daily observations were made to assess the stage of embryonic development. When growth arrested before the blastocyst stage was attained, the embryo was discarded but the medium was usually kept for biochemical assays. Those embryos that progressed to the blastocyst stage, and particularly those that hatched from their zona pellucidas; were observed for up to 14 days after insemination, and then discarded. In some cases the medium in which blastocysts developed was replaced with freshly equilibrated medium at about 6.00 P.M. every 2 to 3 days. The aliquots of medium that supported cleavage stages and blastocyst growth were stored at -8o·c until assayed for hCG.

Embryo Culture

hCGAssay

·The growth potential of surplus embryos was evaluated in HTF and in MEM, obtained from

Samples of culture medium were assayed in a coded rand~mized batch procedure. The levels of

Vol. 51, No.6, June 1989

Lopata and Hay Human embryo development

985

Table 1

a

Development of Surplus Human Embryos, Derived From Bipronucleate Eggs, in Simple and Complex Culture Media Medium

Total embryos

Blastocysts

Percent

Chi-square

HTF+ 1% HCS Original HTF MEM+1%HCS MEMalone AUHTF AllMEM

121 113 130 105 234 235

19 15 38 25 34 63

15.7 13.3 29.2 23.8 14.5 26.8

0.116a 0.616a 10 (P = 0.0014)

Not significant.

dimer hCG in the culture fluids were measured with commercial immunoradiometric assay kits from Bioclone Australia (Sydney, New South Wales, Australia) which used specific monoclonal antibodies. 6 Assay standards were the first international reference preparation. The assay sensitivity for hCG was 0.5 miU/ml. The specificity, precision, and sensitivity of the assay has been described previously. 7 In this assay, levels> 2.5 miU/ml were considered to be significant hCG rises. This level was greater than the range (mean + standard deviation) of hCG levels measured in medium containing 10% HCS (blank controls, n = 15) of 1.5 ± 0.2 miU/ml, and in medium that contained arrested embryos (tissue control, n = 20) of 1.7 ± 0.4 miU/ ml, which failed to develop to the blastocyst stage. Because the medium used for culture was completely replaced every 2 or 3 days, the hCG levels measured at each stage represent the total amount of hCG secreted by a single blastocyst into 1 ml of medium during a 2- to 3-day period.

RESULTS Blastocyst Development and Hatching

The outcome of culture for surplus embryos derived from bipronucleate oocytes is summarized in Table 1. A lower proportion (13.3%) of embryos developed to the blastocyst stage when embryos were retained in the original HTF than in freshly equilibrated HTF (15.7%), but this difference was not statistically significant. Also, a slightly lower percentage (23.8%) develop to blastocysts in MEM without HCS, compared with the same medium supplemented with 1% HCS (29.2%), but this difference was not significant. However, when all embryos cultured in HTF (combined HTF groups) we,re compared with all embryos cultured in MEM (combined MEM groups), it was evident that a significantly higher percentage of embryos (26.8% 986

Lopata and Hay Human embryo development

versus 14.5%, P = 0.0014) attained the blastocyst stage in MEM. The capacity of blastocysts to hatch from the zona pellucida, following their development in MEM or HTF media, is evaluated in Table 2. Approximately one in every three blastocysts was capable of hatching in MEM, both in the presence and absence of an HCS supplement. By contrast, HTF containing 1% HCS failed to support blastocyst hatching. Embryos that arose from tripronucleate oocytes performed poorly in culture (Table 3). Of the 25 embryos that were incubated in the MEM groups, only 2 gave rise to blastocysts. Only a single blastocyst developed from the 68 triploid embryos cultured in the HTF groups. The Morphology and Timing of Development

Embryos that continued to grow in either MEM or HTF media attained the early blastocyst stage by the evening of day 4 and became expanded blastocysts by day 5 or 6 after insemination. In the blastocysts that developed in MEM and subsequently escaped from their zona pellucidas, herniation of the trophoblast usually commenced on day 6 and hatching was completed on day 7 after insemination (Fig. 1a). In the presence of 1% HCS in the medium, the hatched blastocysts became attached to the surface of the dish, and a monolayer outgrowth became established at the periphery of the embryo by day 8 (Fig. 1b). A more extensive outgrowth, spreading from a cellular clump, with Table 2 Hatching of Human Blastocysts in Simple and Complex Media Medium

Total blastocysts

Hatching blastocysts

MEM+1%HCS MEM HTF+ 1% HCS Original HTF

38 25 19 15

11 8 0 0

Percent 28.9 32.0 0 0

Fertility and Sterility

Table 3 Development of Surplus Human Embryos, Derived from Tripronucleate Eggs, in Simple and Complex Media Medium

Total embryos

Blastocysts

Percent

AllMEM AHHTF

25 68

2" 1

8.0

Chi-square

1.5

• One blastocyst hatched. b Not significant.

prominent pseudopod-like extensions in the monolayer (Fig. 1c), was established during the next 2 or 3 days. By about day 11, the monolayer became dispersed to give rise to stellate cells, which seeded the dish well beyond the attached cellular clump_ At this stage, folds of cuboidal epithelium were discernible in the cellular clump (Fig. 1d). By day 14 after insemination, some of the clumps contained degenerative changes and floated free in the culture medium. Occasionally the clumps broke up into fragments, some of which contained multinucleated structures with attached spherical blebs at their periphery (Fig. 1e). In the absence of HCS in the MEM, the hatched blastocysts did not attach to the surface of the culture dish. About 1 day after hatching, the trophoblast of such blastocysts appeared to contract to form a cellular clump. These clumps enlarged because of cell proliferation, but a monolayer was not established. By day 11 after insemination, most of these structures contained darkish foci, due to degenerative changes. The expanded blastocysts that failed to initiate the hatching process, in the presence or absence of HCS contracted to form a cellular clump within the stretched zona pellucida. Rapid cell proliferation within the clumps caused them to enlarge and completely fill the intrazonal space (Fig. 1f). In a few cases, the zona pellucida appeared to be cracked open by the enlarging cell mass. Degenerative foci usually appeared within these intrazonal cellular masses by day 11 after insemination. Production of hCG The total amount of hCG released by hatched blastocysts when grown in MEM + 1% HCS is shown in Figure 2a. In three of these blastocysts, the medium was replaced at regular intervals, but in some cultures the medium was sampled only once at either 10, 11, 12, or 14 days after insemination. The calculated total amount of hCG released iQto a sequence of replaced media samples, or into single samples, at various times after insemination in vitro, are shown together in Figure 2a. Vol. 51, No.6, June 1989

Following hatching, which usually occurred on day 7, the mean level of hCG released rose from 1.8 miU on day 6 to 4.0 miU by the evening of day 7, and over 9 miU by the end of the next day (Fig. 2a). Thereafter the hCG output per blastocyst rose rapidly to reach a mean level of 19,500 miU by day 14 after insemination. In three embryos that hatched in MEM without HCS, the blastocysts continued to be cultured to day 12 in the absence of serum, and the tissues then were transferred to MEM + 1% HCS for the next 2 days (Fig. 2b). As may be seen from the plot, the mean hCG output into the medium by day 12 was about 500 times less in the absence of serum compared with the mean level released by blastocyst tissues grown in MEM containing 1% HCS. However, when the blastocyst tissues grown in serum-free medium were transferred into MEM + 1% HCS on day 12, there was a rapid rise of hCG output during the next 2 days (Fig. 2b). In embryos that failed to hatch (zonal blastocysts) grown in MEM + 1% HCS, hCG levels released are shown in Figure 2c. In two blastocysts, a sequence of three samples (taken on days 8, 11, and 14) were analyzed, and in five blastocysts single samples were assayed on day 11 after insemination. The mean output of hCG per zonal blastocyst was similar to that of hatched blastocysts on days 6 and 8, but it was lower throughout the rest of the culture period compared with hCG output from hatched embryos. By day 14 after insemination, the mean level of hCG released by blastocysts that failed to hatch was about 10 times less than that derived from hatched embryonic tissues. Two embryos that developed to expanded blastocysts in MEM without HCS, and which failed to hatch, continued to be cultured in the absence of serum up to day 12 after insemination (Fig. 2d). Thereafter both blastocysts were transferred into MEM + 1% HCS for the next 2 days. As may be seen from Figure 2d, the mean hCG level did not rise above the control value up to day 12 in the absence ofHCS in the medium. However,when these embryos then were placed in MEM + 1% HCS, there was a marked increase in hCG output (from 0.4 to 1,520 miU/ml) by one of the intrazonal blastocysts and a less dramatic increase of hCG release (from 2.0 to 4.5 miU/ml) by the other. DISCUSSION Embryonic Development Our observations have shown that MEM promoted bla~tocyst development in a significantly Lopata and Hay Human embryo development

987

r

f

Figure 1 Stages of development of embryonic tissues following expansion of the human blastocyst. (a), Hatched blastocyst showing inner cell mass {ICM), trophectoderm (TR), and a zona pellucida (ZP) containing discarded cells. (b), An embryonic cell mass (CM) with an early monolayer (ML) outgrowth. (c), A more advanced monolayer outgrowth with cellular extension (CE). (d), An embryonic cell clump containing folds of cuboidal epithelium (EP). (e), A multinucleated (MN) fragment with attached blebs (BL). (f), A blastocyst that failed to hatch containing a stretched ZP and proliferating cells. All tissues were photographed in culture medium at 200X magnification using differential interference contrast optics.

988

Lopata and Hay Human embryo development

Fertility and Sterility

10000

10000

c

a 1000

1000

~

I

100

•oo

~

~

10

10

& 11 8 9 10 Days after '-'*'"lion

6

12

13

14

6

1000

100C

I

roo ~

~ ~

10

0.1'----~~~-----~-~

6

10

11

12

11

10

13

14

d

100

10

t _ • ....,. t

MEMWITHOUTHCII

9

8

Days alter nsermation

b

~

7

12

13

14

Days- Nwrrelion

MEMwmt:JIJTHCS

0.1'---------~----~

6

9

10

11

12

13

14

Days after irlserr*lation

Figure 2 Production of hCG by blastocyst tissues in vitro at various times after insemination (AI). (a), Hatched blastocysts grown in MEM + 1% HCS. (b), Hatched blastocysts grown in MEM without HCS up to day 12 AI and then transferred into MEM + 1% HCS. (c), Blastocysts that failed to hatch grown in MEM + 1% HCS. (d), Zonal blastocysts grown in MEM without HCS up to day 12 AI and then transferred into MEM + 1% HCS. (e)= individual hCG values.

higher proportion of early embryos than did HTF. In MEM, approximately the same percentage of embryos progressed to blastocysts in either the presence or absence of 1% HCS, indicating that serum was not essential for differentiation to the blastocyst stage. In a study on the influence ofHCS on early embryonic development, Menezo et al. 8 also questioned the need for serum, since human embryos cultured for 2 days in medium B2 and B3 in the presence and absence of serum supplements, respectively, had the same pregnancy potential after transfer to the uterus. Vol. 51, No.6, June 1989

It has been postulated that the Na+jK+ ratio in the culture medium influences the quality of early embryonic development. 1 A comparison of this ratio in MEM (26.5) with that in HTF (29.3) indicates that the proportions of N a+ and K+ are similar. The main difference between the two media, even when both are supplemented with 1% HCS, is the additional presence of a full complement of amino acids and water-soluble vitamins in MEM. Our observations suggest that these additional compounds promoted blastocyst development in cultured human embryos. Further studies, similar Lopata and Hay Human embryo development

989

to those reported recently,9 are required to evaluate the compounds that stimulate human embryogenesis. Our observations also have shown that HTF supplemented with a low level ofHCS failed to support blastocyst hatching, whereas MEM with or without serum supported the hatching of about 30% of the blastocysts that developed in culture. It is likely that the amino acids, or vitamins, or a combination of each, are the factors that promoted not only growth but also hatching of embryos inMEM. Daily microscopic examination of living blastocysts is not sufficient to determine whether there were differences in the rate of cell proliferation in advanced embryos growing in MEM or HTF. Differences in the rate of cell growth in the two media can only be determined if ethical approval were available to carry out histologic preparations on blastocysts, which would permit cell counts andestimations of mitotic indices, or studies on tritiated thymidine incorporation by the embryos. 10 The developmental status and approximate age of human embryos recovered from the uterus has been reported by Buster et al. 11 It was of interest to note that of the 25 embryos recovered between day 4 and 5 of potential development in vivo, only 5 were at the blastocyst stage. This represents a similar incidence of blastocyst formation to that occurring in embryos grown in MEM. In addition, it would appear that the growth rate of embryos to blastocysts in MEM and HTF is similar to the rate of embryo growth following fertilization and growth in vivo11 (the estimated mean duration being 109.1 hours, or 4.5 days). Although there is no information on the timing of blastocyst expansion and hatching in vivo, the evidence summarized by Hertig12 suggests that these events probably occur between days 6+ and 8+ after ovulation (i.e., fertilization). This data corresponds closely to the timing of blastocyst expansion and hatching that occurred in MEM. In view of these encouraging results, and the demonstration that transferred blastocysts arecapable of producing a high pregnancy rate, 11 a new approach could be attempted in clinical IVF and ET. All embryos could be grown in MEM + 1% HCS for 4 to 5 days to identify those that are capable of forming blastocysts. Two of these could be transferred to the patient and any remaining early blastocysts frozen for transfer in subsequent natural cycles. This procedure would avoid the transfer and cryopreservation of embryos that do not have a potential for growth. However, a disadvantage of 990

Lopata and Bay Human embryo development

this procedure is that it would increase the duration of embryo culture in the IVF laboratory, and that a higher proportion of patients may not attain the embryo quality criteria (blastocyst formation) required for transfer. Production of hCG

We have found that hatched blastocysts, growing in MEM + 1% HCS, released detectable levels of hCG by day 7 or 8 after fertilization. Attachment of trophectoderm cells to the surface of the culture dish was not essential for hCG production. Blastocysts that failed to hatch, and remained enclosed by their zona pellucidas, released hCG into the culture medium. The hatched blastocysts, however, produced higher levels of hCG than did those that were trapped within their zona. This suggests that the trophectoderm of hatched blastocysts either grew more rapidly, or its cells were more responsive to growth factors present in HCS, 13•14 than those of the intrazonal blastocysts. Serum was not essential for hatching, but it was found to be required for boosting hCG production by hatched embryos. Thus, hCG output remained low by day 12 after insemination but increased rapidly after adding 1% HCS to the medium (Fig. 2b). In intrazonal blastocysts, hCG remained at basal levels but increased markedly when 1% HCS was added to the culture medium on day 12 after insemination (Fig. 2d). Although the mechanism by which HCS stimulated hCG secretion by blastocyst tissues is unknown, one possibility is that growth factors present in the serum act as the stimulatory agents. 13•14 In natural conceptual cycles, the implanting blastocyst was found to produce about 10 miU of hCG per ml of plasma by day 10 after fertilization. 15 If for simplicity of calculation, it is assumed that the secreted hCG is retained within a 1,600-ml plasma compartment,16 it can be estimated that the implanting blastocyst produces at least 16,000 miU of hCG by day 10 and over 160,000 miU by day 14 after fertilization. This is at least eight times more than the mean level of hCG released by hatched blastocyst tissues growing in culture at similar times after fertilization. This suggests that there was a severe retardation of trophectoderm cell growth in vitro, and/or an inadequate production of hCG by these cells in response to growth factors present in HCS. It has been reported that in natural conceptual cycles the first detected increase in plasma hCG occurred, in the. majority of cases, between day 9 and Fertility and Sterility

10 after the luteinizing hormone surge 15 (i.e., from day 8 to 9 after ovulation). Based on our findings, this would correspond to about 1 or 2 days after blastocyst hatching, attachment and the start of hCG production in vitro. It also would correspond to the early stages of blastocyst implantation described by Hertig. 12

7.

8.

9.

Acknowledgments. We are grateful to Ms. Yvonne duPlessis, Mr. Harold Bourne, Dr. Gayle Jones, Ms. Janine Clark, Ms. Michelle Mullens, and Mr. Felix Nieto for the care with which they assessed embryos in the IVF laboratory and for providing the information and organization required for the continuity of embryo culture.

REFERENCES 1. Quinn P, Kerin JF, Warnes G M: Improved pregnancy rate in human in vitro fertilization with the use of a medium based on the composition of human tubal fluid. Fertil Steril 44:493, 1985 2. Schroeder AC, Eppig JJ: The developmental capacity of mouse oocytes that matured spontaneously in vitro is normal. Dev Biol102:493, 1984 3. Lopata A, Summers PM, Hearn JP: Births following the transfer of cultured embryos obtained by in vitro and in vivo fertilization in the marmoset monkey (Callithrix jacchus). Fertil Steril50:503, 1988 4. Johnston WIH, Lopata A, Pepperell RJ, Trounson A, Wood C: The use of in vitro fertilization in the infertile couple. In The Infertile Couple, Edited by RJ Pepperell, B Hudson, C Wood. Edinburgh, Churchill Livingstone, 1987, p 263 5. Lopata A, Gronow MJ, Johnston WIH, McBain JC, Speirs AL, Leung PCS: In vitro fertilization and embryo implantation. In Infertility: Male and Female, Edited by V Insler, B Lunenfeld. London, Churchill Livingston, 1986, p 496 6. Hay DL: Discordant and variable production of human chorionic gonadotrophin and its free alpha- and beta-sub-

Vol. 51, No.6, June 1989

10.

11.

12.

13.

14.

15.

16.

units in early pregnancy. J Clin Endocrinol Metab 61:1,195, 1985 Hay DL, Murphy JR: Evaluating human choriogonadotropin heterogeneity in nongestational malignancy with monoclonal immunoassays. Oncology 44:174, 1987 Menezo Y, Testart J, Perrone D: Serum is not necessary in human in vitro fertilization, early embryo culture, and transfer. Fertil Steril 42:750, 1984 Carney EW, Bavister BD: Stimulatory and inhibitory effects of amino acids on the development of hamster eightcell embryos in vitro. J In Vitro Fert Embryo Transfer 4: 162, 1987 De Luca E, Lopata A: Effect of an insulin/transferrin/selenite supplement on DNA synthesis in mouse embryos cultured in vitro. (Abstr. 7). Presented at the Twentieth Annual Conference, Australian Society for Reproductive Biology (ASRB), Newcastle, New South Wales, Australia, August 28 to 31, 1988. Published by ASRB, 1988 Buster JE, Bustillo M, Rodi IA, Cohen SW, Hamilton M, Simon JA, Thorneycroft IH, Marshall JR: Biologic and morphologic development of donated human ova recovered by nonsurgical uterine lavage. Am J Obstet Gynecol 153: 211,1985 Hertig AT: Implantation of the human ovum. In Progress in Infertility, Edited by SJ Behrman, RW Kistner. Boston, Little, Brown & Co, 1975, p 411 Maruo T, Matsuo H, Oishi T, Hayashi M, Nishino R, Mochizuki M: Induction of differentiated trophoblast function by epidermal growth factor: relation of immunohistochemically detected cellular epidermal growth factor receptor levels. J Clin Endocrinol Metab 64:744, 1987 Ladda RL, Bullock LP, Gianopoulos T, McCormick L: Radioreceptor assay for epidermal growth factor. Anal Biochem 93:286, 1979 Lenton EA, Neal LM, Sulaiman R: Plasma concentrations of human chorionic gonadotropin from the time of implantation until the second week of pregnancy. Fertil Steril37: 773,1982 Fishel SB, Edwards RG, Evans CJ: Human chorionic gonadotropin secreted by preimplantation embryos cultured in vitro. Science 223:816, 1984

Lopata and Hay Human embryo development

991