ABSTRACT
In albino rats placental scars of a second pregnancy partially or completely overlapped 32 of 312 old placental scars from a previous pregnancy in 61 uterine horns.
Gross recognition of the superposed scars was often impossible. Microscopic recognition of the superposed sites was possible under the conditions of the experiment.
The major changes in the distribution pattern of second pregnancy scars as compared to those of the first pregnancy are (a) the intervals between second pregnancy scars are more variable; (b) the most posterior new scar tends to be further from the cervix than the last first pregnancy scar; (c) the most anterior new scar tends to be closer to the oviduct than the corresponding first pregnancy scar; and (d) although the oviducal scar-free segment is longer than the cervical scar-free segment in both pregnancies, the difference is less in the second pregnancy.
INTRODUCTION
The purpose of this study is to investigate possible interrelationships between successive sets of placental scars in the laboratory white rat (Rattus norvegiens). The primary problem has been to determine if placental scars formed as a result of a second pregnancy are ever superposed upon those of a first pregnancy. An attempt has also been made to learn if there are any changes in location or spacing of second sets of scars when they are compared with the distribution of scars remaining from a previous pregnancy.
A placental scar may be defined as a pigmented area of the uterus occurring at the site of a previous placenta. In the rat the placental scars appear as orange to dark brown pigmented spots along the mesometrial border of the uterus. The scars are produced by an accumulation of hemosiderin in the cells of the reticuloendothelial system. These cells are concentrated in the area between the longitudinal and circular muscle layers as well as in the deeper part of the endometrium. As the age of the scars increases the pigmented areas decrease in size and become darker in colour. It is thus relatively easy to distinguish old placental scars from very new ones. Since the rat placenta forms at the same uterine level and opposite the antimesometrial implantation site, inferences about the spacing of blastocysts at implantation can be drawn from studies of the spacing of scars.
METHODS
Virgin rats approximately three months in age and of the Wistar strain were used in this study. These animals were bred at known dates and isolated after breeding. The number of young born in each litter and the date of birth were recorded. Litters were weaned on day 21 and at this time the first laparotomy was performed on the uniparous females. The reproductive tract was exposed and examined by holding the mesometrium and uterine horn between the fingers in front of a strong light. Placental scars appeared as dark spots in sharp contrast to the pink uterine horn. The number and approximate location of scars in each horn were recorded.
The uniparous female rats were bred a second time 10 to 20 days after the first laparotomy. On day 12 of this pregnancy, a second laparotomy was performed and the number of gestation sacs present was determined. Placental scars of the first pregnancy were not clearly visible at this time. The animals were permitted to go to term and then were sacrificed on post-partum day 12.
The uteri were removed, pinned in a dissecting pan by placing a pin at each end of the horn, and flooded with 10 per cent, formalin. They were later bleached with hydrogen peroxide to increase the contrast of the placental scars with the surrounding tissue. Measurements of the total length of each uterine horn and the intervals between scars were made by calipers. The diameters of placental scars were measured by using a dissection microscope and ocular micrometer. The measurements of scar lengths and intervals between scars in each horn were converted to percentages of the total length of that horn, so that comparisons between horns could be made.
Scars of the first and second pregnancy were usually easily differentiated. Those of the first pregnancy (old scars) were approximately one half as large as the new scars, dark brown to black in colour, and restricted to the mesometriouterine junction. The new scars were large, swollen areas extending well into the mesometriuçn and of yellow or golden colour.
A total of 51 experimental animals was used, but results are presented for only 61 horns. Other horns were excluded because of failure to become pregnant, infection, post-operative adhesion, or similarity of appearance of old and new placental scars.
The technique of pinning and stretching the uterine horns may be subject to criticism because conclusions drawn from the measurements would be invalidated if the elongation were not uniform in all segments of the horn. The method was checked by applying spots of dye near both cervical and oviducal ends of the horns of two females and measuring in situ the length of the horns, the interval between spots, and the distance from each spot to the nearest end of the uterus. The measurements were repeated after the tracts were removed, stretched, pinned, and fixed. When the measurements were converted to percentages, it was found that the maximum variation in intervals was only 1·9 per cent. It is concluded that stretching was essentially uniform throughout the tract and that the method is sufficiently accurate to warrant comparisons of measurements.
RESULTS
Superposition and merging of scars
The number of placental scars observed in the preserved uteri varied from the number expected. During the first laparotomy 312 old scars were recorded, while 368 developing gestation sacs were noted during the second laparotomy. At the final examination only 280 old placental scars were found (thus 32 or 10 per cent, were missing). One or more placental scars were missing in 17 of the 61 uterine horns.
In one horn, the number of old placental scars was greater than expected. Only three old placental scars were seen during the first laparotomy, but four old placental scars were found in the preserved tract. The most probable explanation is that one placental scar had been overlooked during the first laparotomy. The extra scar was located near the cervix, which is the most difficult area to see at laparotomy.
The number of new placental scars found was 369, as compared to 368 expected. The discrepancy occurred in a horn where the additional scar was located at the cervical end and almost within the cervix. Again it is assumed that the embryo count was incorrect.
Gross examination and comparison of the number and position of visible placental scars with the diagrams showing scar distribution at the first laparotomy made it possible to predict probable locations for ‘missing’ scars. A close examination of the suspected scars often revealed that overlapping was incomplete and the double scars were larger than single new ones (Plate 1). Frequently a dark streak of pigment was also apparent at the junction of the swollen new scar and the mesometrium. Often the position of the missing scars was doubtful even though the number which should have been present was known. Identification of double scars was made more difficult because some single old or new placental scars had two areas of pigment concentration, making them appear double (Plate 1). This pattern was produced when pigment accumulated inside the circular muscle stratum in the deeper stroma of the mesometrial side and also in the area between the circular and longitudinal muscle layers. No consistent diagnostic character of the double scar was found in the gross studies. Unless the previous reproductive history of an animal were precisely known, it would be impossible to determine from gross examination of placental scars the actual number of implantation sites represented.
Histological preparation of examples of old, new, and superposed scars were made. These scars were serially sectioned and stained with haematoxylin and eosin. Microscopic study showed that superposed sites could be clearly distinguished (Plate 2, figs. A-F). Pigment accumulates as large granules in macrophages lying between the circular and longitudinal muscle layers (Plate 2, fig. D) in old scars. In new scars, these cells are larger, the cytoplasm more abundant, and the pigment granules much smaller (Plate 2, fig. E). A more detailed description of the histology of placental scars is given by Conaway (1955). Both types of pigmented cells are apparent in superposed scars (Plate 2, fig. F). The small cells laden with large granules occupy a more peripheral position and lie just under the longitudinal muscle. This distribution produces the visible dark line observed in gross examination. The large cells with small granules are more abundant and more centrally located. Superposed scars can thus be clearly recognized under the conditions of this experiment. If both scars were much regressed, however, it is doubtful that microscopic examination would differentiate between them.
The expected frequency of merging first and second pregnancy scars
It has been indicated that 32 of the 312 old scars were not visible as separate spots along the uterus. These concealed old scars were either in contact with or partially or completely overlapped by new scars.
A crude estimate of the expected number of old scars which would be so concealed can be made if it is assumed that the new embryos implant randomly.
Merging of old and new scars will occur if the centre of the new scar is within a radius of itself from either edge of an old scar. The average diameter of old scars is 1·7 per cent, of the length of the horn and the average radius of new scars is 1·4 per cent. Therefore, merging of old and new scars will occur if the new scar lies within a segment equal to 4·5 per cent, of the uterine length. Each additional old scar adds 4·5 per cent, to this total length. The number of merging scar pairs expected for each group of uteri with the same number of old scars equals N x 0·045 x T, where N=number of old scars per horn, 0·045 = the segment over which merger will occur for one old scar, and T = total number of new scars in all horns with N old scars. As an example, the number of merging scar pairs in uteri with four old scars equals 4 x 0·045 x 61 = 10·98. Therefore approximately 11 of the 61 new scars would be expected to merge with an equal number of old scars.
The above calculations were repeated for each old scar number class (Table 1). These calculations show that 87·99 merging pairs would be expected. This is considerably in excess of the 32 merging pairs observed. Table 1 also shows that the number of superposed new scars observed is much less than expected when the number of old scars in the uterine horn is low. When five or more old scars are present the observed and expected frequency of merging are in somewhat closer agreement.
Distributional pattern of scars of first and second pregnancies
Comparisons of the intervals between old and new placental scars were made to determine if there were differences in the intervals between scars of the two pregnancies. The data were analysed by calculating variances of the intervals between scars for all sets of 3·8 old and 3·8 new scars (Table 2). The variances were greater in all sets of new scar intervals than the variances for the corresponding old scar intervals, except in the six-scar class where the reverse relationship occurred. The ‘F’ test indicates the difference was significant only in the three-, four-, and eight-scar classes, and approached significance in the five-scar class. This analysis indicates that there is probably greater variation in the intervals between scars in sets of three, four, five, and eight new scars than in similar sets of old scars.
There is a tendency for placental scars of both sets to occur much closer to the cervix than to the oviducts. A greater distance free from old placental scars occurred at the oviducal end in 87 per cent, of the 61 horns and the same distributional pattern for new scars was found in 73 per cent, of these horns. The distances from the first scar to the oviduct and the last scar to the cervix were divided into scar number classes for both old and new scars. Comparison of the means of these distances showed that with one exception the greatest scar-free distance is at the oviducal end in all twelve categories. The ‘t’ test showed that six of these differences were highly significant (Table 3).
The data (Table 3) also suggest changes in the relationships of the terminal scar-free segments in second pregnancies as compared to first pregnancies. The terminal old scars tend to lie closer to the cervix but farther from the oviduct than corresponding new scars. When the means are compared, the only exception to this trend is that the mean of the first pregnancy cervical segment is greater than the corresponding second pregnancy segment in the three-scar class. The ‘t’ test shows that only the difference between the old and new oviducal segment in the four-scar class is significant. The uniform trend in the differences between the means, however, suggests that there may be a real difference in the pattern of distribution of the scars in the two pregnancies.
DISCUSSION
Some relationships of placental scars or implantation sites of later pregnancies to scars of previous pregnancies have been noted by other workers. Fortuyn (1920) reported that in the mouse there was a tendency for the gestation sacs of a second pregnancy to occur mid-way between the ‘brown cells’ of the first pregnancy. In a later study, Fortuyn (1929) reported that when embryos of the striped hamster (Cricetulus griseus) implanted in areas other than mid-way between old placental scars, resorption always occurred. De Lange (1934) states that in the gundi (Ctenodactylus gundi) either implantation does not occur or an abnormal gestation sac is produced as long as placental scars are distinct in the uterus. He indicates, however, that if only a few gestation sacs are present placentae of the second pregnancy occur between or beside old placental scars.
Bull (1949) emphasized that implantation in rodents is antimesometrial and placentation mesometrial. Thus, even when pregnancies occur in rapid sequence the two endometrial regions concerned are well separated until the embryos of the second pregnancy establish placentae. This arrangement, together with rapid and complete regeneration of the endometrium at the old placental sites, should make it possible for embryos of successive litters to occupy the same sites. No evidence that this does occur was presented by Bull. Davis & Emlen (1948) suggested superposition of placental scars might occur in the rat, since they found multiparous rats with fewer placental scars than the recorded total of young born.
The data obtained in the present study show that placental scars of the second pregnancy were partially or completely superposed upon 10 per cent, of the first pregnancy scars. It was estimated that if the new placental scars were distributed at random 28 per cent, of the old scars (88 of the 312 total) would contact or be overlapped by the new scars. This estimate is obviously crude since clearly the embryos of a set do not implant at random but tend to space along the horn. There is also a general anterior shift in the position of the second pregnancy scars which would tend to reduce the amount of overlap. However, the estimate does suggest that there is no preferential implantation at the level of old sites and it is possible that these sites are less favourable for implantation.
The major changes in the distributional pattern of second pregnancy scars as compared to old scars are: (1) the intervals between new scars are more variable; (2) the last new scar tends to be farther from the cervix than the last old scar; (3) the first new scar tends to be closer to the oviduct than the corresponding old scar; (4) although the oviducal scar-free segment is longer than the cervical segment in both pregnancies, the difference is less in the second pregnancy. Thus, the second set of scars tends to be more centrally spaced along the length of the uterine horn.
Frazer (1955) has shown that rat embryos are not distributed equally along the uterine cornua. He found that more embryos were implanted in the caudal half than in the cranial half. These findings are in agreement with those presented here. He also noted that the difference was more apparent when the embryo number was low than when it was five or more. The present analysis shows that the difference between cranial and caudal scar-free segment lengths is actually more marked when the scar number is high. The different methods of analysis used may account for this variation in findings, especially since Frazer apparently did not consider the number of pregnancies per animal as a factor.
The cause of the changes in distribution is not known. Possibly one factor may be the vascular pattern of the uterus. Various workers have suggested that implantation tends to occur in regions of greater vascularity or of local hyperemia (Mossman, 1937; Reynolds, 1949; and Young, 1952). Preliminary work suggests that in the virgin uterus the cervical portion of the uterine horn is more highly vascular than the oviducal region. Such a pattern fits the distribution of first pregnancy scars. If this vascular gradient becomes less marked as a result of pregnancy, or is destroyed, it would be expected that implantation would normally occur over more of the uterine length in the second pregnancy.
Data obtained in the present study fail to show whether or not normal embryo development occurs at the superposed sites of second pregnancies. Since placental scars are formed when a portion of the embryos of a litter undergo resorption any time after the seventh day of pregnancy (Conaway, 1955), the presence of a scar does not indicate the fate of the embryo. The number of young born was always less than the number of gestation sacs recorded at the second laparotomy in all animals where superposition was recorded. However, the number of young recorded for the animals in which superposition did not occur was also usually less than the number of gestation sacs noted. Probably embryo resorption was high in all of the experimental animals, as a result of the laparotomies. Also the number of young recorded is not an accurate measure of the young born since the counts were often made as late as twelve hours after birth.
ACKNOWLEDGEMENTS
We are indebted to Dr. H. W. Mossman for suggestions and comments. Dr. Peter Frank and Dr. Edward Novitski gave helpful suggestions concerning analysis of data.
REFERENCES
EXPLANATION OF PLATES
PLATE 1
Uteri showing placental scars. Specimens are cleared and photographed with transmitted light. Wratten 39 filter.
PLATE 2
Fig. A. Cross-section of uterus 4L through an unscarred area. The myometrium is unmodified and no pigment is seen, x 125.
Fig. B. Cross-section of uterus 12R through an area showing a superposed old and new scar. The region of the old scar (O) appears more dense than the new scar area (N). x 125.
Fig. C. Cross-section of uterus 4L through an unscarred area at the mesometrio-uterine junction. x 555.
Fig. D. Cross-section of uterus 4L through an old placental scar at the mesometrio-uterine junction. The large pigment granules are apparent, x 555.
Fig. E. Cross-section of uterus 4L through a new scar at the mesometrio-uterine junction. The hypertrophied cells are laden with fine pigment granules, x 555.
Fig. F. Cross-section of uterus 4L through superposed new scar at the mesometrio-uterine junction. The smaller cells and large pigment granules of the old scar are at the right while the large cells and small granules of the new scar are at the left of the field, x 555.
PLATE 2
Fig. A. Cross-section of uterus 4L through an unscarred area. The myometrium is unmodified and no pigment is seen, x 125.
Fig. B. Cross-section of uterus 12R through an area showing a superposed old and new scar. The region of the old scar (O) appears more dense than the new scar area (N). x 125.
Fig. C. Cross-section of uterus 4L through an unscarred area at the mesometrio-uterine junction. x 555.
Fig. D. Cross-section of uterus 4L through an old placental scar at the mesometrio-uterine junction. The large pigment granules are apparent, x 555.
Fig. E. Cross-section of uterus 4L through a new scar at the mesometrio-uterine junction. The hypertrophied cells are laden with fine pigment granules, x 555.
Fig. F. Cross-section of uterus 4L through superposed new scar at the mesometrio-uterine junction. The smaller cells and large pigment granules of the old scar are at the right while the large cells and small granules of the new scar are at the left of the field, x 555.