Early embryo mortality in natural human reproduction: What the data say

Abstract

How many human embryos die between fertilisation and birth under natural conditions? It is widely accepted that natural human embryo mortality is high, particularly during the first weeks after fertilisation, with total prenatal losses of 70% and higher frequently claimed. However, the first external sign of pregnancy occurs two weeks after fertilisation with a missed menstrual period, and establishing the fate of embryos before this is challenging. Calculations are additionally hampered by a lack of data on the efficiency of fertilisation under natural conditions. Four distinct sources are used to justify quantitative claims regarding embryo loss: (i) a hypothesis published by Roberts & Lowe in TheLancet is widely cited but has no practical quantitative value; (ii) life table analyses give consistent assessments of clinical pregnancy loss, but cannot illuminate losses at earlier stages of development; (iii) studies that measure human chorionic gonadotrophin (hCG) reveal losses in the second week of development and beyond, but not before; and (iv) the classic studies of Hertig and Rock offer the only direct insight into the fate of human embryos from fertilisation under natural conditions. Re-examination of Hertig's data demonstrates that his estimates for fertilisation rate and early embryo loss are highly imprecise and casts doubt on the validity of his numerical analysis. A recent re-analysis of hCG study data concluded that approximately 40-60% of embryos may be lost between fertilisation and birth, although this will vary substantially between individual women. In conclusion, natural human embryo mortality is lower than often claimed and widely accepted. Estimates for total prenatal mortality of 70% or higher are exaggerated and not supported by the available data.

Keywords: Hertig; early pregnancy loss; embryo mortality; human chorionic gonadotrophin; occult pregnancy; pre-implantation embryo loss.

Figure 1.. Schematic representation of timelines and key events in ( A) non-fecund and ( B) fecund menstrual cycles.

Menstrual cycle lengths vary considerably and most fall within a range of 20 to 40 days . A typical menstrual cycle is usually represented as lasting for 28 days, as shown here. Differences in cycle length are mostly due to variations in the duration of the follicular phase, the time from the onset of menstruation to ovulation. The time from ovulation to the onset of the next cycle, the luteal phase, is more consistently 14 days. Therefore, in the typical 28 day cycle, ovulation occurs midway at around 14 days. The fertile period (shown in light blue) is the time during which coitus may result in a pregnancy. The probability of pregnancy is highest when coitus occurs in the two days leading up to ovulation . In a normal fecund cycle, fertilisation occurs within hours of ovulation in the fallopian tube, after which point an embryo is present and development begins. Embryonic development may fail at any stage from fertilisation through to birth. 6–7 days after fertilisation, the embryo begins to implant in the uterine wall at which stage human chorionic gonadotrophin (hCG) produced by the embryo becomes detectable in urine or serum samples. The onset and duration of pregnancy may be defined in various ways: gestational pregnancy (typically used in clinical practice) is timed from the first day of the last menstrual period; developmental pregnancy begins with fertilisation; in an IVF treatment cycle, although an embryo is present in the uterus immediately following embryo transfer, pregnancy is not considered to be established until there is evidence of implantation, usually provided by elevated hCG levels. The earliest point at which a woman could observe that she is pregnant is approximately 14 days after ovulation/fertilisation with the first missed period. The stage at which pregnancies are clinically confirmed depends on study design and clinical practice, and may be at gestational day 28 (i.e., first missed menstrual period, Zinaman (1996) , French & Bierman (1962) ), gestational day 42 (Wang (2003) ), or following a positive pregnancy test (Wilcox (1988) ) or satisfactory ultrasound scan.

Figure 2.. Distribution of embryo loss estimates from fertilisation to birth derived using a modified version of the model of Roberts & Lowe .

Embryo loss values were calculated using alternative speculative values (see text and Table 1) obtained by randomly sampling from normal distributions with means equal to the Roberts & Lowe values and a coefficient of variation of 20%. 100,000 simulated embryo loss values were obtained. Frequencies within a bin size of 0.25% are shown. The 2.5 ^th and 97.5 ^th percentiles are indicated. The simulation was performed using NONMEM 7.3.0® (Icon PLC, Dublin, Eire). Simulated values are in Dataset 1.

Figure 3.. Graphical representation of the fate of 1,000 pregnancies in progress at 4 weeks’ gestation (2 weeks’ post-fertilisation).

The figure is generated using values in Table 4.3 of Leridon (1977) and are derived from six different studies (see text). The Kauai Pregnancy Study data published by French & Bierman (1962) are shown in thick black. Data from Shapiro (1970) were analysed either with all pregnancies included (ALL) or with those pregnancies excluded that aborted within one week of study entry (EXCL.). The greater loss observed with ALL may be due to a correlation between study entry and abortion risk. Based on these data, the risk of losing a pregnancy ongoing at 4 weeks’ gestation ranges from 12.5% to 23.7% (excluding Shapiro (1970) ALL). Values are in Dataset 2.

Figure 4.. Summary of findings from thirteen studies that used hCG detection to diagnose early pregnancy.

Data are arranged by publication date and the first author of the study is shown. Three datasets are shown: (i) the percentage of at risk menstrual cycles that were hCG positive; (ii) the percentage of hCG positive cycles that did not manifest as clinical pregnancies = early pregnancy loss; and (iii) the percentage of clinical pregnancies lost prior to 12 or 28 weeks or live birth (definitions vary between studies). A clinical pregnancy may be manifest by a missed period although criteria vary between studies. Videla-Rivero et al. , Sasaki et al. , Cole and Mumford et al. do not report sufficient data to calculate all three values. Values are in Dataset 3.