The Experimental Induction of Prenatal Mortality and the Subsequent Elimination of the Dead Embryos in Rabbits

Extensive prenatal mortality occurs in wild rabbits after implantation. All the embryos die and are reabsorbed in many of the litters between the tenth and fifteenth days post-coitum (Brambell, 1942, 1944). Whereas the proportion of dead embryos in surviving litters could be calculated directly, the proportion of litters represented by those with all the embryos reabsorbing, which have consequently ceased to develop, could not be estimated without knowing the duration of the reabsorptive processes. It was essential to know also whether the dead embryos were always reabsorbed or whether they might be aborted and, if so, the factors which determine whether reabsorption or abortion occurs. Search of the literature failed to provide adequate information on these points and the present work was undertaken in consequence. The object of the work was confined to establishing the time relations of reabsorption and abortion in the rabbit. A review of the extensive literature on the endocrine aspect of the maintenance of pregnancy and any attempt to interpret the results from an endocrinological viewpoint have been avoided intentionally. The work has provided the material for histological investigation of the reabsorptive processes and of the condition of the uterus after abortion compared with that after parturition. These results will be published separately.

Mammalian embryos in the uterus have been killed experimentally by a variety of means. Giacomini (1893) and Hammond (1917) removed the embryos in rabbits surgically but allowed the placentae to remain in the uteri. Newton (1935) destroyed the embryos in the uteri of mice by manipulation and van Wagenen & Newton (1943) removed the embryos surgically in monkeys. Ovariectomy has been found by many authors to result in the termination of pregnancy in most, but not in all, species. Injection into the mother of various substances such as urea (Feis, 1894), colchicine (Kerr, 1947), sex-hormones including natural and artificial oestrogens (Parkes, Dodds & Noble, 1938), etc. have been employed. Huggett & Pritchard (1945), working on rats, employed four methods: direct surgical interference, ovariectomy, oestrone injections and gonadotrophic hormone injections.

It was necessary to know, first, the times from the death of an embryo rabbit until (a) fragmentation was so advanced that it was no longer possible to determine the age at death, and (b) the reabsorptive site in the uterus was no longer recognizable as such macroscopically; secondly, how these periods varied (a) with the age of the embryos at death, and (b) when all the embryos in the litter, or only some of them, were killed ; thirdly, if abortion, rather than reabsorption, might occur. This involved killing, at will, either all the embryos simultaneously or else only selected ones; hence two contrasting methods were used, that of injection of synthetic oestrogen and that of direct surgical interference with the embryos.

The material employed consisted of three wild rabbits and forty-nine tame rabbits ; the majority of the latter belonged to the Agricultural Research Council’s Compton strain of Dutch rabbit. The tame rabbits were permitted to copulate twice in rapid succession and only those which did so were employed. The stage of development of the embryos was dated from the time of the first copulation. The wild rabbits were freshly caught and were pregnant at the time of capture. They were part of a large sample of wild rabbits upon which exploratory laparotomies were performed to determine the stage of pregnancy.

A uniform dose of 5 mg. of stilboestrol in olive oil was injected subcutaneously in all the animals in which the embryos were killed by means of an oestrogen.

Laparotomy was performed under full ether anaesthesia and with appropriate aseptic precautions in those animals in which an attempt was made to destroy the embryos by surgical means. The method employed was to perforate the selected uterine swellings through the antimesometrial wall of the uterus, using a fine suture needle. The needle was passed into the swelling and several stabs at the embryo were made. It was possible as a rule to confirm that the needle passed through the embryo since the antimesometrial wall of the swelling at the stages employed is semi-transparent. Care was taken to ensure, so far as possible, that the needle passed through the head of the embryo. A very small number of animals died under anaesthesia, but otherwise no animals showed adverse after-effects.

The embryos were judged to be dead in all experiments when they had become flaccid and white. The heart was never beating in such embryos. These criteria were certainly rigorous so far as the embryo is concerned, since they entailed some measure of autolysis, but it does not follow that the embryonic placental tissues had died simultaneously. The stage of development attained by embryos already dead at autopsy was taken to represent the age at which they had died.

Since material was required for the histological investigation of the course of reabsorption in each experimental series, some of the swellings were unopened at autopsy and were preserved intact.

Three series of experiments were performed in which stilboestrol was injected at 11,$1534$and 19–20 days post-coitum respectively. No embryos survived the injection by more than 24 hr., judged by the stage of development attained by the embryos as determined by comparison with the normal table of Minot & Taylor (1905).

The first series consisted of fifteen rabbits, of which one proved to be non-pregnant and was rejected. Particulars of the animals, which were killed and examined at intervals ranging from 1 to 10 days after injection at 11 days, are given in Table 1. Four days after injection, or 3–$312$ days after death, appears to be the limit at which the age can be determined. The reabsorption sites were barely distinguishable from old placental sites at 10 days after injection, and this may be taken as the limit at which they could be recognized, were it not known that reabsorption, rather than abortion or parturition, had occurred. The progressive decrease in diameter of the swellings after the death of the embryos is remarkably uniform. The data are represented graphically, together with the data of the diameters of the uteri between swellings, in Fig. 1. The diameters of the swellings fall about a line which would intersect that of the diameters of the uteri on the 10th day.

The second series consisted of thirteen animals injected at $1534$ days and one at 17 days post-coitum. One was excluded as the embryos had evidently begun to retrogress several days before the injection. Particulars of the remainder, which were killed and examined at intervals of from 1 to 9 days after injection, are given in Table 2. Five days after injection, or 4–$412$ days after death, was the maximum time at which the stage of development of the embryos could be determined. The 9th day was the limit at which the reabsorption sites could be recognized as such, macroscopically, since they were scarcely distinguishable from old placental sites at that time, but they remain visible as slight swellings on the uterus for some days longer. The rate of decrease in diameter of the swellings can be seen from Fig. 2, the principal difference from those killed at 11 days being in the larger size of the swellings at the time of death. Otherwise the chief difference from the previous series was the tendency for the placenta to become separated from the uterine wall. The separation had begun the day after injection but was not complete until the 5th day. The embryonic part of the placenta was not present on the 8th day and on the 9th day only the unhealed sites marked the former positions of the placentae.

The third series consisted of only three rabbits injected at 19–20 days post-coitum and is summarized in Table 3. All aborted, but, in one some of the embryos were retained and had regressed in situ at the time of autopsy, 9 days after injection. The other two aborted all the embryos on the 5th day after injection, one aborting all, and the other some, of the placentae as well.

Three series of experiments were performed, the embryos being killed in two series at 16 days, and in one series at 19–20 days post-coitum.

The first series consists of six animals in which all the embryos were perforated at 16 days and particulars of which are given in Table 4. In another intended for this series, one embryo survived though the membranes had been perforated, and it was therefore transferred to the second series. The embryos were intact, though flabby and distorted, in the first two animals killed at 5 and 7 days after operation respectively. Hence the dissolution of the embryos in these was much slower than in those of the corresponding series injected with stilboestrol at $1534$ days post-coitum, in which no intact embryos were found more than $312$ days after death. The second of these animals (E/107) was plucking fur and nest-making when killed and the uterus was observed at autopsy to be contracting strongly. It is probable that this animal would have aborted soon if it had. lived. The next two animals, killed at $734$ and at 9 days after operation respectively, had aborted but the abortions were not recovered for examination. The two remaining animals, killed at $934$ and at 11 days after operation, had no trace of embryos or placentae and it is probable that they aborted some days previously.

The second series, in which some of the embryos were killed at 16 days postcoitum and some survived, includes eight tame rabbits. Two wild ones, in which partial hysterectomy was performed, and some of the embryos in the remaining uterus were killed, can be included, as the estimated age of the embryos at operation was 15 and 17 days respectively. Particulars are given in Table 5. One (E/84) aborted the surviving embryos 3 days after operation and was killed as soon as this was discovered. Six were successful experiments in that some of the embryos survived and some died, although some of the embryos which died had not been perforated and some which were damaged survived. The dead embryos in all of these were recognizable, and their age at death could be determined. The dead embryos in most cases were intact, though flaccid and distorted, even when the animal was not killed until near full term, upwards of 12 days after operation.

The third series consists of four tame rabbits in which all the embryos were perforated at 19–20 days post-coitum and one wild one in which one embryo only was perforated. The results are summarized in Table 6. Two of the embryos in the perforated swellings survived in one tame animal (E/114). Two of the tame animals aborted all the embryos and placentae. The aborted embryos were recovered and had died at the time of operation in one of these (E/115), which aborted 3 days after operation. The embryos were not recovered and the precise time of abortion was unknown in the other (E/120), killed 8 days after operation. The condition of the uterus of the wild rabbit (W/22), when killed at 8 days after operation, was inconsistent with reabsorption and it must have aborted soon after the operation as the placental sites were healed. The remaining tame rabbit (E/121) was reabsorbing all the embryos, which were still intact, and at the 20-day-stage, when it was killed 8 days after operation. This is the only animal in which all the embryos were killed at 20 days, either by stilboestrol or surgically, and which did not abort.

The results show that 5 mg. of stilboestrol administered subcutaneously kill all the embryos with uniformity, whether at $1112$⁠, 16 or 20 days post-coitum and within 24 hr. of administration. Laparotomy and puncture of the swellings is by no means so effective. It was anticipated that the mere puncture of the membranes, with the consequent loss of some of the embryonic fluid contained by them, would be sufficient to kill the embryos. The fact that many survived loss of fluid, and some probably direct damage as well, was surprising.

The speed at which reabsorption proceeds, when it occurs, varies widely according to the age of the embryos at death, the method of killing and whether some or all in the litter die. The remains of the embryos permit determination of the age at death for at most 3–$312$ days after death at $1112$ days, as compared with 4–$412$ days after death at 16 days post-coitum, when killed by stilboestrol. The reabsorption sites remain recognizable as such macroscopically for 9–10 days after death from stilboestrol administration at either 11 or $1534$ days post-coitum; although the sites remain visible after that time, especially in the $1534$-day-series, it is not then possible to distinguish them from old post-partum placental sites. The embryos remain intact and their age at death can be determined for at least 7 days after death at 16 days, and 8 days after death at 20 days when all have been killed surgically. Hence autolysis of the embryos must proceed much more quickly after stilboestrol administration than after surgical interference. The age of the embryos at death can be determined, and they often remain intact for at least 13 days after death at 16 days post-coitum, that is until full term, when some of the other embryos in the litter survive. Obviously in this case they are very much more persistent than when all are killed with stilboestrol. They are also very much more persistent than when all are killed surgically but the evidence is inconclusive as to whether this is due to slower autolysis or to reabsorption not being curtailed by subsequent abortion, as it often is when all the embryos are dead.

The method of removal of dead embryos and membranes appears to vary according to the stage of development at which death occurs and to whether or not some of the embryos survive. It may vary also according to the manner in which the embryos are killed. Reabsorption occurred in every case where the embryos were killed by stilboestrol at $1112$ or 16 days. The results, when all the embryos were killed surgically at 16 days, were consistent with an initial period of reabsorption lasting 7 days being followed by abortion of the remnants. Conversely when some only of the embryos were killed surgically at 16 days reabsorption was the rule, one animal only aborting, and that incompletely. Although the number of animals in which the embryos were killed at 20 days was small it is significant that all three of those injected with stilboestrol aborted, as did two certainly, and probably three, of those in which the embryos were killed surgically. One in which all the embryos were killed and one in which some survived, reabsorbed the dead embryos and did not abort.

The difference in the methods of removal between the series in which all the embryos were killed at 16 days by stilboestrol and by perforation respectively is remarkable. It was observed that, in the stilboestrol series, although the embryos and placentae always remained in situ, the placentae tended to become detached through the development of clefts in the connective tissue of the decidua basalis close to the muscularis. Although this tendency was apparent soon after the death of the embryos complete separation of the placentae was not observed until 5 days after injection. Probably abortion cannot occur until this zone of weakness has differentiated. The appearances observed suggested that although the placental tissues died rapidly after stilboestrol injection they did not do so after perforation, and often survived the death of the embryos, so reaching a stage of development in advance of that at which the embryos died. This is not improbable since many authors, including Newton (1935), van Wagenen & Newton (1943) and Huggett & Pritchard (1945), have shown that the placenta can survive the death of the embryo and may even persist until normal full term in several species. Thus the difference between the two 16-day-series described herein, in that abortion did not occur in one and was frequent in the other, may depend on a difference in the stage of development attained by the placentae, resulting from the difference in the methods of killing the embryos.

It is difficult to determine the time at which abortion occurs because rabbits eat the aborted embryos and membranes immediately. It was known accurately in four of the eleven cases in which abortion was believed to have occurred. These four cases vary from $1912$ to 25 days post-coitum. The probable time of abortion in the others could be estimated from the known time of death of the embryos and the condition of the placental sites at autopsy. All the estimates are consistent with it having occurred between 20 and 25 days post-coitum.

We wish to thank Miss Patricia Allen for much assistance in the conduct of the experiments. We are indebted to Dr A. S. Parkes, F.R.S., for suggesting the use of stilboestrol as a convenient means of killing the embryos in the intact pregnant rabbit. We are indebted to Sir Michael Duff, Bart., for the supplies of live wild rabbits and we wish to acknowledge the interest and care his Head Keeper, Mr Charles Parker, took in obtaining the animals in satisfactory condition. We wish to acknowledge the technical assistance of Mr R. A. Lansdown and Mr W. Holland. This work is part of a scheme of research on prenatal mortality supported by the Agricultural Research Council, to whom we are greatly indebted.