Rh-prophylaxis in early abortion

Abstract

In most countries anti-D immunoglobulin G (anti-D IgG) is given to rhesus (Rh)-negative women, although evidence is lacking for the need of this intervention after abortion in early pregnancy. This is especially true for medical abortion, which has been used increasingly in recent years. Mifepristone was approved for medical abortion in France in 1988 under the brand name of Mifegyne. The UK and Sweden followed in 1991 and 1992, respectively, and most other European countries in 2000. So far more than one million women have used this method in Europe alone, and the use is still increasing. The need for the application of immunoglobulin in medical abortion has never been evaluated. The necessity seems questionable when the intervention is performed in early pregnancy and in many cases even before a fetal heart rate can be demonstrated by ultrasound. The Swedish Board of Health and Welfare gave a recommendation against the use of anti-D IgG in early spontaneous or medical abortions in 1997. However, there is still no evidence-based analysis for or against prophylaxis. This leads to regional variations in the treatment of early abortion. This paper reviews publications on this subject with the particular aim of providing a background for recommendations or further research initiatives. A review of this kind could not be found in the literature, and has probably not been done recently. This article is based on: a review of the literature in MedLine on the Rh blood group system, fetomaternal hemorrhage (FMH) and Rh (D)-immunization in relation to spontaneous or induced surgical and medical abortion; further articles found as reference in the above-mentioned publications and recommendations from specialists working in the field; recommendations and references from health authorities, abortion providers and manufacturers of Rh immunoglobulins; personal communication with specialists in gynecology and obstetrics, haematology and embryology. The origin of the hematopoietic stem cells (HSCs) that can be found in the hematopoietic organs of the embryo have long been the subject of controversy. The first recognizable blood cells appear within the wall of the yolk sac at about day 17 postconception. The numbers of erythroid burst-forming units (and granulomacrophage progenitors) within the yolk sac decline between the fourth and fifth week after conception, or 6–7 weeks after the last menstrual period (LMP), as they increase reciprocally in the liver. This suggests that HSCs of the yolk sac migrate to the liver, where they establish nests of proliferating cells that may eventually seed other hematopoietic organs of the embryo, including spleen, lymph nodes, thymus and bone marrow. There is a concomitant switching from embryonic to fetal hemoglobin (Hb F) isoforms (1). The role of the yolk sac as the primary source of blood cells is short lived. Its hematopoietic activity decreases rapidly until 10 weeks postconception (12 weeks LMP), when it finally ceases to produce blood cells. At around the fifth week postconception (7 weeks LMP) the hematopoietic activity of the liver quickly becomes the primary source of red blood cells, a position it holds until the 30th week of gestation. The majority of hemoglobin synthesized during the hepatic phase is Hb F. The fetal spleen also commences production of red blood cells at about 10 weeks LMP and continues throughout the second trimester of pregnancy. Bone cavities begin to form red blood cells at about 20 weeks LMP and rapidly become the sole source of hematopoiesis in humans. Additionally, Hb F is gradually replaced after delivery by adult hemoglobin (Hb A), the latter being produced only in the bone marrow. By the end of the first year after birth, Hb A is the only type of hemoglobin that can be found, although some individuals continue to produce a low level of Hb F (1–3). The Rh blood group system consists of a number of inherited antigens. It is the commonest cause of hemolytic disease of the fetus and the newborn. The presence of anti-D antibodies in an Rh (D)-negative woman causes hemolytic disease in an Rh-positive fetus/newborn. The Rh blood group system is also of major importance for hemolytic transfusion reactions. The Fisher–Race nomenclature assumes the presence of three genetic loci for the antigens, each with possible alleles. Today more than 45 antigens in the Rh system are known. The most important are D, c, C, e and E. There are two Rh genes, RhD and RhCE, located on the short arm of chromosome 1. The genes that control their synthesis are not clearly known. There are three codominant alleles; D, d, C, c, E and e. The entire gene complex is inherited as a unit (haplotype). The resulting antigens are named D, C, c, E and e and are inherited in two sets of three; one set from each parent. The most common sets of antigens are Cde (42%), cde (38%) and cDE (14%). Many variations of the D-antigen expression form a heterogeneous group called weak D (previously Du). All weak D forms are considered to be D-positive because of their ability to produce an antibody response. Rh-antigens are proteins (nonsugar containing, like other antigens) that are associated with large membrane complexes on the red blood cell surface. The number of antigenic sites for a specific red cell depends on the genotype (3–5). About 40–50% of all Rh-positive individuals are homozygous for D, having inherited D from both parents. The others (50–60%) are heterozygous for D, having inherited a D-containing set from one parent and a non-D set from the other parent. A child from homozygous Rh-positive parents can only be Rh-positive. A child from a heterozygous Rh-positive couple on the other side can be Rh-positive or Rh-negative. Only an Rh-positive fetus or a blood transfusion of Rh-positive blood is able to cause Rh-immunization in an Rh-negative woman. The antibodies produced by a woman affect only an Rh-positive fetus (3). The presence or absence of the D-antigen denotes an Rh-positive or an Rh-negative individual and for most clinical purposes, people are classified as Rh-positive or Rh-negative by testing their red cells with anti-D antiserum. Because no antiserum specific for d has been found, d in the blood group classification signifies the absence of a discernible allelic product (the d-antigen does not exist) (3,6). On the first antenatal visit, every pregnant woman has her blood group and Rh determined, in case there is no previous record of her blood group. In addition, all Rh-negative women are tested for the presence of anti-D Igs. This is easily done by the indirect (Coombs' test) antiglobulin test for the detection of IgG antibodies. A positive antiglobulin test indicates the presence of antibodies capable of reacting with Rh-positive red cells, and capable of hemolysing them (3). In case of an induced abortion antibodies are not tested. Rh-negativity is a Caucasian trait and its prevalence is around 15%, ranging from 12% in Finns to 35% in Basques. Other populations were probably entirely Rh-positive at one time, but by intermingling genes with Caucasian genes, Rh-negativity appeared. As a result, 8% of American blacks are now Rh-negative, as well as 5% of West Africans and 1% of American Indians. Among Indo-Eurasians 4% are Rh-negative. Less than 1% of indigenous Chinese and Japanese are Rh-negative. Caucasian Rh-negative women have a 60% risk of being pregnant with an Rh-positive fetus. Consequently Rh-incompatibility between an Rh-negative mother and an Rh-positive fetus can be found in 10% of all pregnancies in Europe (5,7,8). The D-antigen is highly immunogenic. The other Rh-antigens are less immunogenic (less risk for sensitization) than D but are of significance. Rh-negative people develop circulating antibodies of anti-D Ig following exposure to Rh-positive red cells (3). Small amounts of Rh-positive blood may be sufficient to produce Rh-immunization. In one study 50% of cases were immunized with 10 mL of Rh-positive blood (3). In other experiments 80% of cases were immunized after one injection of 0.5 mL Rh-positive red cells and 30% were immunized with repeated injections of 0.1 mL of Rh-positive red cells (3). A secondary immune response may occur after exposure to much smaller volumes (0.03 mL) of Rh-positive red cells. About 30% (range 25–50%) of Rh-negative people are never sensitized (nonresponders), even when exposed to large volumes of Rh-positive blood (reviewed in 3). When the fetus is AB0 incompatible to the mother, fetal red cells are less frequently detected in the woman's circulation, and then only in small numbers. Caucasians have a high prevalence of Rh-negativity (15%) and the blood group A (40–50%). It has been estimated that group A incompatibility between mother and infant gives 90% protection against Rh-immunization. Incompatibility of the blood group B gives 55% protection. Consequently AB0 incompatibility confers very significant protection against primary Rh-immune response, but it confers no protection against the secondary immune response (3,7). Studies have also shown that an Rh-positive fetus/infant that initiates an Rh-immunization in the mother is more frequently male than female. The sex ratio of male to female was found to be 1.44–1.74 to 1 (6,7). The primary immune response. The primary immune response develops slowly. In experimental Rh-immunization of male volunteers, the first antibodies develop as early as 4 weeks after injection but it usually takes 8 to 9 weeks before a response appears. The primary response is usually weak and often IgM in nature. IgM does not cross the placenta. IgG appears between 6 weeks to 6 months after antigen exposure. IgG anti-D crosses the placenta and leads to hemolysis of Rh-positive fetal red cells (8). The secondary immune response. Once the primary response has developed, a new exposure to Rh-positive red cells leads to a rapid increase in antibodies, which are mostly IgG in nature. All further exposures thereafter may produce even higher levels of antibodies. If the intervals between antigen exposures are long, both antibody titer and avidity of the antibody for the Rh-antigen will be markedly increased, and the severity of Rh erythroblastosis greater. The antibody titer decreases over time without a new booster. Because of the slow immune response only later pregnancies are at risk for hemolytic disease of the newborn and the fetus (3,8). The transfer of IgG is an active process and takes place only from the mother to the fetus and only via the placenta. IgG is the only immunoglobulin able to cross the placental barrier, via Fc receptors on the plasma membrane of the placenta (7). The heterogeneity of the IgG subclasses produced and the differences in transplacental transfer may explain the differences in the severity of fetal symptoms. The reaction of anti-D antibody with D-antigens leads to the destruction of Rh-positive red cells. This is not a complement-fixing reaction, and most IgG-coated red blood cells are hemolysed extravascularly in the splenic reticuloendothelial system. Antibody-dependent cellular cytotoxicity mechanisms may also contribute (mononuclear phagocytes, lymphoid cells) (6). Immunization to the D-antigen can be prevented by administration of anti-D IgG either before or shortly after exposure to Rh-positive red cells. From an immunological point of view, there are three potential mechanisms of action for the anti-D Ig, antigen blocking, clearance and antigen deviation and central inhibition by generation of antigen-specific suppressor cells near the time of exposure to Rh-positive red cells (6). Rh-immunoglobulin was first released for general use in 1968 and its introduction has very successfully reduced immunization to D-antigen (9,10). Rh-immunoglobulin is very effective if given in adequate amounts at the right time before sensitization has occurred (9). However, it should be noted that anti-D IgG does not protect against development of other antibodies, which may also cause hemolytic disease of the fetus and the newborn. Anti-D IgG is extracted by cold alcohol fractionation from plasma donated by male or female persons with high levels of anti-D IgG antibodies. The donated plasma is pooled and fractionated by commercial manufacturers. Anti-D IgG is prepared in different dosages and is administered as an intramuscular injection (in the deltoid muscle or in the gluteal region). The half-life of this exogenous anti-D IgG is 24 days, but titers decrease over time (11). Studies have shown that the postpartum administration of a single dose of anti-D IgG to Rh-negative women within 72 h of delivery reduces the risk for Rh-immunization by about 85–90%, provided the woman has not already been immunized. The longer the delay between anti-D IgG administration and delivery, the less likely it will be effective (6). The rate of Rh-immunization at 6 months after delivery with subsequent application of anti-D IgG is less than 1% compared to an estimated 16% (range 5–16% in the literature) of Rh-immunization in untreated women after the first Rh-positive and AB0 compatible pregnancy (3,9,10,12). The general recommendation and practice for delivery at term is to administer 250–300 µg anti-D IgG to the mother, provided she is Rh-negative and the fetus is found to be Rh-positive or in case the Rh-status of the fetus is unknown. This routine has been under discussion for many years. An amount of 300 µg of anti-D IgG is considered sufficient to protect against the transfusion of 30 mL fetal blood or 15 mL packed blood cells. However, 99% of women have an FMH of less than 4 mL at delivery (10,13). A potential problem is the development of nonspecific anti-globulin in mothers who receive anti-D IgG. The incidence of this is estimated to be between 5% and 25%. It is unknown if this is of any clinical significance. However, polyclonal IgG will pass from the mother to the fetus (9). There is also increasing concern about the future supplies of anti-D IgG and there is already a shortage in some parts of the world. Subsequently, some physicians in Australia proposed reducing the indications during the first trimester, unless transplacental hemorrhage (TPH) could be documented. The USA might reduce the first-trimester indications and recommend using lower doses of anti-D IgG. In developing countries anti-D IgG is generally not available. However, the number of Rh-negative women is smaller in most parts of the world outside Europe and North America (11), although those few women are at higher risk of beings pregnant with a Rh-positive fetus. The risk of transmission of viral infections [human immunodeficiency virus (HIV), hepatitis B Virus (HBV) and hepatitis C virus (HCV)] via the administration of anti-D IgG is regarded as minimal to absent. No side-effects are known today. Concern about the theoretical risk of contracting Creutzfeldt–Jacob disease from blood products has led to the advice that no plasma products from the UK should be used for the preparation of blood products including anti-D IgG (11). Anti-D IgG has to be stored cold (2–8 °C), which must be considered as a practical problem in some parts of the world. The cost for one dose of anti-D IgG is 560 Skr (61 €). Termination of pregnancy involves trauma to the choriodecidual space, favoring TPH. It has been estimated that more than half of the circulating fetal blood volume is in the placenta. If its selective permeability is lost, this amount of blood might enter the maternal circulation, with the potential risk for sensitization (14). The volume of fetal blood at 8 weeks LMP is estimated to be 0.33 mL. This figure is based on a report that fetal blood represents 30% of placental weight, or approximately 4.2 mL, at 12 weeks. Assuming an exponential growth, the blood volume at 6 weeks would be less than 0.25 mL (15). However, the total blood volume in a fetus younger than 12 weeks has not been verified. No study or information could be found on this subject. It is known that Rh-immunization in an Rh-negative woman is related to previous pregnancies with an Rh-positive baby or transfusions with Rh-positive blood. It has been presumed that early and possibly unrecognized abortion can cause an Rh-immunization, as fetal red blood cells or Rh-antigens from degraded red cells might enter the maternal circulation through defects in the placental barrier during abortion. In this regard it is important to establish the earliest stage of fetal development at which blood group antigens are developed. Bergström et al. found Rh-antigens on the red blood cells of a 10-mm fetus, obtained approximately 38 days after conception or 52 days after LMP (16). The embryo was found in a 47-year-old woman, with systemic lupus erythematosus (SLE), who was undergoing hysterectomy because of a cervical myoma. The embryo had a yolk sac and was obtained within intact membranes. Microscopic examination of the suspension showed almost exclusively nucleated megaloblasts of fetal type. The Rh-antigen was found to be Rh-positive (identified by Löw's method). This is the only report of such an early fetus and the morphology of the blood cells may vary during the earliest stages of fetal life. Despite this, showing Rh-antigens in a fetus as young as approximately 38 days after conception suggests that early abortion could theoretically induce Rh-immunization in the D-negative woman. Several methods can be used for tracing fetal erythrocytes in maternal blood. One of the most commonly used is the acid-elution method, or the Kleihauer–Bethke technique. This method can detect one fetal red cell in 20 000 adult red cells but it has several sources of error and is considered as too insensitive. The method considered most reliable is flow cytometry. For this analysis the maternal blood sample is first treated with anti-D and then with fluorescein-labeled anti-IgG. Red cell progenitors are found universally in very small numbers in maternal blood during early gestation, and this very sensitive method has a risk for overestimating the FMH. When the acid-elution method was used for analysis of the same samples, estimates have been about twice as high. The different techniques used make it difficult to compare the results of published studies. During normal pregnancy. TPH can occur as early as at 4 weeks following fertilization (6 weeks LMP). This is the time when fetal and maternal circulation in the placenta has been formed and when the vascularization of the villi and the pumping action of the fetal heart begins (7,17,18). TPH as low as 0.004 mL has been detected at this stage of pregnancy (17). Using the acid-elution method Hb F cells are found in small numbers in 1–2% of normal adults. In about 25% of pregnant women, the level of maternal Hb F rises above the upper limit of normal (0.9%) starting at about 8 weeks and may reach 7% (7). This can be considered as a physiological phenomenon and complicates the interpretation of fetal cells counts made by the acid-elution method. In one study a majority of pregnant women were found to have an increase in Hb F; 3% during the first trimester, 45% in the second trimester and 65% at delivery (7). Quantification showed that in 80% of these pregnancies TPH was very small (<0.1 mL), and that most of the pregnant women were Rh-sensitized as a result of small or undetectable TPH. Large TPH was uncommon during a normal, uncomplicated pregnancy. In less than 1% of the cases more than 5 mL and in less than 0.23% of the cases more than 20 mL of TPH could be found. In another study, Jörgensen came to similar results (18). Hb F was analyzed using the acid-elution method in 381 women in the second and third trimesters. The frequency rose from 12% at 6 months to 41% at 9 months. The occurrence and amount of Hb F red cells were found to be independent of parity and the weight of the placenta; however, there was a significant increase in Hb F red cells among women with preeclampsia (18). Anti-D IgGs are formed by few Rh-negative women during their first uncomplicated pregnancy. The average rate reported for this is 0.9%, range 0.3–1.9% (3,6,7,13). The risk of Rh-immunization for an Rh-negative women is considered to be 1–2% during a pregnancy with an Rh-positive fetus, in case they are AB0 compatible. The risk is reported to be 14–17% during delivery (3). At delivery after a full-term pregnancy, there is a 99% risk for TPH less than 4 mL and 0.3% risk for TPH more than 15 mL. It is estimated that 25–35 µg anti-D IgG would successfully protect against 1 mL of fetal blood cells or approximately 2 mL of fetal blood (19). In spontaneous abortion. Evidence that spontaneous abortion occurring in the first trimester can cause primary Rh-immunization could not be found in the published literature. There is some evidence that significant TPH occurs only after curettage but does not occur after either threatened or incomplete spontaneous abortions. Litwak et al. were the first to suggest that all women undergoing spontaneous abortions, complete or incomplete, should receive Rh-prophylaxis (20). Jørgensen found Hb F red cells in 35% of women (n = 55) with spontaneous abortion in the first and second trimester (7–21 weeks LMP) (18). Most of the women (93%) had a curettage for suspected incomplete abortion. These results were significantly higher compared to 13% Hb F in women with a normal pregnancy of the same gestational age. The occurrence of Hb F red cells was found to be independent of the gestational age of the pregnancy in which the abortion took place. In all cases the amount of red cells corresponded to a TPH of less than 0.1 mL. It can therefore be expected that a TPH of less than 0.1 mL will occur during spontaneous abortion (with or without curettage) during the first trimester. No immunization was reported among the women undergoing spontaneous abortion. The author suggests that there might be no correlation between the amount of TPH and the risk for Rh-immunization after abortion. However, the recommendation for giving anti-D IgG prophylaxis to all Rh-negative women undergoing abortion is given, no matter whether Hb F is found or not. Von Stein et al. investigated whether women with threatened spontaneous abortion had a greater incidence of TPH compared to a control group matched for gestational age (21). Pregnant women (n = 89) with a mean gestation of 9 weeks (all less than 20 weeks) presenting with threatened abortion (defined as vaginal bleeding without cervical dilatation or passage of the fetus) were studied. Pregnant and nonpregnant age-matched control groups were included. Ultrasound and/or human chorionic gonadotropin (HCG) measurements were used to document the pregnancy and the acid-elution method was used for tracing fetal red cells. The results of the study were not statistically significant, although women with threatened abortion had a higher incidence of TPH (11%). In comparison, women with a normal pregnancy had a 4% incidence of TPH during the first and early second trimesters. One nonpregnant control was found to have 4% Hb F, probably due to a hereditary persistence of Hb F. Even if this study did not give a clear result, the authors recommend Rh-prophylaxis for all women with threatened abortion. In 1997, the Swedish Board of Health and Welfare published a recommendation to stop giving anti-D IgG to women after early spontaneous or medical abortion (22). The immunization rate in women with early (<12 weeks) spontaneous abortion (n = 620) was studied during the following years by Messeter, who could find no evidence for an increase in Rh-immunization (23). However, there is a risk that only a few immunizations per year would not significantly influence the data. The recommendation to stop giving prophylaxis was later changed to giving no general recommendation. Thus the absence of a national guideline resulted in different approaches among hospitals in Sweden. In surgical abortion. The fragmentation of placental and fetal tissue occurring during surgical abortion and curettage probably releases fetal red blood cells into the maternal sinuses. The uterine relaxation under general anesthesia may facilitate the passage of fetal red blood cells into the maternal circulation. Recommendations to give anti-D to all Rh-negative women undergoing any kind of abortion often refer to a study by Leong et al. (15). This is a randomized study to determine whether or not TPH occurs in pregnant women (n = 75) undergoing induced abortions less than 8 weeks LMP (15). The duration of the pregnancies was calculated from LMP and physical examination (no ultrasound). Karman aspiration catheters were used and syringes were used as a vacuum source. Blood samples were taken before and after the procedure to test for fetal cells and were analyzed with the acid-elution method. Fetal cells were found in two (2.6%) women before the abortion and in 12 (15.5%) after the procedure. Fetal red blood cells could be found in a significant number of women with a pregnancy older than 6 weeks LMP. No fetal red blood cells could be found in women at less than 6 weeks LMP. The authors recommended that all Rh-negative women at 6 weeks LMP or more should be given anti-D IgG following induced surgical abortion (15). The problems with this study is the imprecise method used to determine TPH and the duration of the pregnancy as no ultrasound was used. This is a problem found in several studies. Jørgensen compared blood samples before and after surgical termination of pregnancy (and other abortion methods excluding medical abortion) in the first and second trimesters (n = 186) (18). An increase in the number of Hb F cells was found in 17% of the women after the abortion. In 5% the extent of the TPH was more than 0.1 mL and exceeded 1 mL in 2% of the cases. No details are given about how length of pregnancy was estimated. Freda et al. estimate the risk of Rh-immunization to be negligible during the first month and 2% at 2 months, but substantial (>9%) at 3 months and beyond (24). The risk of immunization was reported to increase directly with the gestational age at abortion. Rh-negative primigravidae (n = 1785) were followed during 9 years. Of these, 4.2% had positive antibody titers in the early phase of their next term pregnancy. It was concluded that the TPH after abortion at LMP less than 8 weeks is negligible and all Rh-negative nonimmunized women who have an abortion at 2 months or beyond should receive anti-D IgG whenever and wherever possible. The method for estimating gestational age as well as methods of abortion were not described. In a frequently cited study, Simonovoits et al. studied Rh-negative secundi gravidae women (n = 386) whose first pregnancy ended in induced abortion and who were not given anti-D IgG at the time of abortion (25). The hypothesis was that a secondary immune response never occurs at the beginning of the second pregnancy in nonimmunized women. In all cases more than 6 months elapsed between the abortion and the second pregnancy. The estimated rate of immunization turned out to be about 2.6%. It was also pointed out that 2.6% risk is related to Rh-negative women irrespective of the Rh group of the fetus. However, in a population in which the rate of Rh-positive is 85%, and with random mating, it can be assumed that 40% of the fetuses of Rh-negative mothers will be Rh-negative because of the percentage of Rh-negative or heterozygous fathers. Consequently only 60% of the mothers are at risk of Rh-immunization. Calculated on this basis, the correct risk of immunization in nonprotected Rh-negative women after abortion amounts not to 2.6% but to 4.3% (25). Abortion method, duration of first pregnancy and dating of the pregnancy were, however, not well described in the study, and the rate of immunization by the end of the first pregnancy was not well known. Because of the smaller blood volume of the fetus during the first trimester it could be assumed that less anti-D IgG should be needed during this period (in full-term pregnancies 300 µg anti-D IgG is generally considered to protect the women for being Rh-immunized). Stewart et al. studied Rh-negative pregnant women (n = 755) who presented for surgical abortion at 4–12 weeks LMP (26). The women were given 50-µg doses of anti-D IgG and tested 6 months later for Rh-sensitization. A control group was given 300 µg anti-D IgG. There was no group of untreated controls. None of the groups showed evidence of Rh-sensitization. The conclusion was that 50 µg will prevent Rh-immunization in women up to 12 weeks of gestation after surgical abortion. The treatment failure rate was assumed to be about 0.25%. Nevertheless, Rh-immunoglobulins are available only in 250- or 300-µg ready-to-use syringes in most countries. This makes it impossible to follow the recommendation of a reduced dosage in early abortion. In medical abortion (mifepristone–prost- aglandin). In theory and as discussed by Urquart and Templeton, a difference in TPH could be expected between surgical and medical abortion and possibly be attributed to damage of blood vessels at curettage allowing the passage of fetal blood into the maternal circulation, whereas medical abortion stimulates uterine contractions and constrictions of blood vessels reducing the opportunity for fetal cells to enter the maternal circulation (27). These authors performed a small study where they estimated serum alpha-fetoprotein (AFP) in women (n = 20) before giving a single dose of mifepristone (27). Concentration of AFP was measured again 48 h later, before prostaglandin insertion, and at the end of the following 4-h observation period. Data were compared to a control group (n = 20) with similar gestation age before and after surgical abortion. In the medical abortion group maternal serum AFP did not increase in any of the 20 patients during the hours between mifepristone treatm