SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. BACKGROUND
  4. FETOMATERNAL HAEMORRHAGE DURING PREGNANCY
  5. ANTI-D ALLOIMMUNISATION DURING PREGNANCY
  6. ANTENATAL ADMINISTRATION OF ANTI-D IMMUNOGLOBULIN
  7. SEARCH STRATEGY
  8. RESULTS OF CLINICAL STUDIES ()
  9. OPTIONS
  10. References
  11. Appendices

Background Standard post-delivery administration of anti-D, together with further anti-D for events known to result in fetomaternal haemorrhage (FMH) during pregnancy, has reduced the incidence of alloimmunisation in RhD women to 0.83—1.5% in the UK. Residual alloimmunisation occurs mainly for two reasons: i) failure to administer sufficient anti-D at the correct time after known at-risk events, either during pregnancy or at delivery; and ii) alloimmunisation during pregnancy as a result of ‘silent’ FMH. The RhD antigen is well developed by 6 weeks’ gestation and the fetoplacental blood volume increases during pregnancy. Studies show that 3% of pregnant women have FMH in the first trimester, 12% in the second, and 45% in the third. Analysis of alloimmunisation in primigravidae clearly shows that on average, 90% are detectable after 28 weeks’ gestation. Additional anti-D prophylaxis during the course of pregnancy, starting at 28 weeks, can reduce alloimmunisation to a minimum by protecting against occult FMH. The identification of intrapartum alloimmunisation as being the ‘true’ cause of alloimmunisation is best assessed by studying first pregnancies, and rigorous analysis of antenatal anti-D efficacy should preferably include observation in second pregnancies to avoid underestimation of alloimmunisation. Ideally there should be no exclusions, i.e. treating both arms of the study on an ‘intention to treat’ basis, otherwise there will be an overestimate of efficacy under routine practice conditions. Initial safety concerns about effects of antenatal anti-D on the fetus have not been confirmed in practice.

Options Although there is only one randomised controlled clinical trial (with small numbers) demonstrating a further reduction in alloimmunisation following antenatal administration of anti-D1 meriting a grade A recommendation (see appendix II), the total body of evidence of efficiency is compelling. Whilst two doses of 300μg are effective, this is no more so than the single dose in practice, and as it requires considerably more anti-D immunoglobulin, it is probably not cost effective. If a single dose is to be given, it is too late at 34 weeks, and 28 weeks is to be recommended. If divided doses are to be given at 28 and 34 weeks, 50μg is insufficient, and 100 μg is recommended. Two dose regimes can be recommended, as follows:

  • i) 
    single dose of 300μg at 28 weeks the results of the single 300pg dose in first pregnancies is limited to the Canadian study2 with observed reduction from 1.6% (45/2768) in concurrent nonrandomised controls to 0.18% (2/1086) in the treatment group.
  • ii) 
    two doses of 100μg at 28 and 34 weeks—the two controlled studies give similar results in first pregnancies3,4. A reduction in alloimmunisation is seen from 1.11 % (4/360) in controls to <0.28% (0/362) in the treatment group in the French study3, and from 0.95% (19/2000) in controls to 0.32% (4/1238) in the treatment group in the English study4.

Whilst anti-D is in limited supply, it is more cost effective to restrict antenatal prophylaxis to first pregnancies. It is also probable that a single dose of 250μg (as used in Europe) will be as effective in practice as the 300μg dose, given the limitations of the anti-D quantification assay, and the vial overfill introduced by manufacturers, but this has not been formally proven in clinical trials. The number of RhD deaths is now very low, even with standard postpartum prophylaxis, but there is a systematic underreporting in the UK, due to early fetal deaths being recorded as ‘abortion’ rather than as haemolytic disease of the newborn (HDN). There have been no systematic studies on the reduction in mortality observed with antepartum anti-D. Nevertheless, it is self-evident that if immunisation is largely prevented, then so will fetal morbidity and mortality.


BACKGROUND

  1. Top of page
  2. Abstract
  3. BACKGROUND
  4. FETOMATERNAL HAEMORRHAGE DURING PREGNANCY
  5. ANTI-D ALLOIMMUNISATION DURING PREGNANCY
  6. ANTENATAL ADMINISTRATION OF ANTI-D IMMUNOGLOBULIN
  7. SEARCH STRATEGY
  8. RESULTS OF CLINICAL STUDIES ()
  9. OPTIONS
  10. References
  11. Appendices

During pregnancy, and especially at the time of childbirth, small amounts of fetal red blood cells (RBC) enter the maternal circulation. When the woman is RhD-negative, and the fetus RhD-positive, the woman becomes alloimmunised to the RhD antigen, and produces IgG anti-D which crosses the placenta to cause Rh haemolytic disease. Passive immunisation with anti-D immunoglobulin can prevent the RhD-negative woman from being actively alloimmunised, if given early after exposure to RhD-positive fetal RBC (within 72 h).

The introduction of routine post-partum prophylaxis by administration of 500 IU (100 μg) anti-D immunoglobulin to RhD-negative women delivering a RhD-positive baby has reduced the alloimmunisation of at-risk women in the UK by 90%, from 17% (in unprotected ABO compatible pregnancies) to 0.83–5 5–7. Nevertheless, even when combined with the administration of additional anti-D on the basis of an estimate of fetomaternal haemorrhage(FMH) by the Kleihauer-Betke test, or given after events known to be associated with FMH, the rate of alloimmunisation has remained constant in the UK since the introduction of routine postpartum prophylaxis in the early 1970s.

Residual alloimmunisation occurs mainly for two reasons: i) failure to administer sufficient anti-D at the correct time after known at-risk events, either during pregnancy or at delivery; and ii) alloimmunisation during pregnancy as a result of occult FMH; even when correctly administered, the current UK prophylaxis programme cannot prevent the latter.

FETOMATERNAL HAEMORRHAGE DURING PREGNANCY

  1. Top of page
  2. Abstract
  3. BACKGROUND
  4. FETOMATERNAL HAEMORRHAGE DURING PREGNANCY
  5. ANTI-D ALLOIMMUNISATION DURING PREGNANCY
  6. ANTENATAL ADMINISTRATION OF ANTI-D IMMUNOGLOBULIN
  7. SEARCH STRATEGY
  8. RESULTS OF CLINICAL STUDIES ()
  9. OPTIONS
  10. References
  11. Appendices

Although there is a number of recognised events that are associated with FMH during pregnancy, such as spontaneous or therapeutic abortion, antepartum haemorrhage, amniocentesis, abdominal trauma, ectopic pregnancy, etc., a number of studies have shown that ‘occult’ FMH occurs in the absence of a precipitating event.

The RhD antigen is well developed by 6 weeks’ gestation, although the risk of FMH in the first trimester is limited by the amount of fetal RBCs. The fetoplacental blood volume increases from 25 ml at 19 weeks to 150 ml at 31 weeks, and as might be expected, spontaneous FMHs occur with increasing frequency and volume during the course of pregnancy, especially during the second and third trimesters. There are some difficulties in accurate assessment of very small transplacental haemorrhage (TPH) due to persistent haemoglobin F (HbF) or increase in F-cells above the normal limit (0.9%) between 8 and 32 weeks’ gestation in 25% of normal pregnancies. The incidence of very small spontaneous FMH in the first trimester is therefore uncertain. Using a standardised sensitive test capable of detecting 0.01 ml FMH, and testing prospectively every 2 weeks during pregnancy, Bowman, Pollock and Penston8 detected 3% (1/33) FMH in the first trimester, 12.1% (4/33) in the second, and 45.4% (15/33) in the third. In this series, only one woman (3%) had FMH > 5 ml.

Routine Kleihauer testing during antenatal anti-D trials gives an indication of spontaneous FMH at 28 and 34 weeks’ gestation when the women were attending for anti-D prophylaxis (Table 1). The Prevalence of FMH increases at the end of pregnancy. Large FMH in the last trimester, >5 ml fetal blood, was observed in in o.5% women1, and >10 ml in 0.27% of women9. These analyses from only two samples taken during each pregnancy almost certainly underestimate the true incidence throughout pregnancy, especially in the third trimester, when prophylactic anti-D would be expected to result in clearance of some RhD positive fetal RBC from the maternal circulation.

Table 1.  Prevalence of spontaneous FMH (ml whole blood).
 FMH*28–30 weeks30–39 weeks
  1. * equivalent to approximately 50% less as fetal RBC.

Bowman9> 1 ml23/5680 (0.4%)88/1895 (1.84%)
 >2.5 ml11/5680 (0.2%)45/1895 (0.94%)
 > 10 ml3/5680(0.05%)13/1895 (0.27%)
Huchet1<0.5ml55/957 (5.7%)67/957 (7%)
 >0.5ml17/957 (1.8%)18/957 (1.9%)
 5 ml1/957 (0.1%)5/957 (0.5%)

ANTI-D ALLOIMMUNISATION DURING PREGNANCY

  1. Top of page
  2. Abstract
  3. BACKGROUND
  4. FETOMATERNAL HAEMORRHAGE DURING PREGNANCY
  5. ANTI-D ALLOIMMUNISATION DURING PREGNANCY
  6. ANTENATAL ADMINISTRATION OF ANTI-D IMMUNOGLOBULIN
  7. SEARCH STRATEGY
  8. RESULTS OF CLINICAL STUDIES ()
  9. OPTIONS
  10. References
  11. Appendices

As little as 0.1 ml RBC may sensitise, i.e. initiate an immune response, in a good responder10. There is, therefore, ample opportunity for anti-D to appear during pregnancy before delivery, even in the absence of an observed precipitating event. The overall incidence during uneventful first pregnancies reported worldwide’ I is approximately 0.87% (263/30, 155). As might be expected from the natural history of spontaneous FMH, anti-D appears much more frequently towards the end of pregnancy, as a result of repeated challenge of larger volumes of RhD-positive RBC (Table 2).

Table 2.  Time of anti-D detection during pregnancy.
 Up to 28 weeks29–36 weeks37 weeks-term
Eklund1115/107 (14%)41/107 (38%)5/107 (48%)
Bowman95/62 (8%)10/62 (16%)47/62 (76%)
Tovey51/19 (5%)7/19 (37%)11/19 (58%)

Analysis of the time when anti-D is first detected in primigravidae clearly shows that on average, less than 10% of intrapartum alloimmunisations are detectable before 28 weeks’ gestation. The major risk occurs after 28 weeks, with most alloimmunisations detected in the last trimester, when the fetus is fully developed, and the fetoplacental volume is greatest. Even these data underestimate the true rate of alloimmunisation as some women do not have detectable anti-D in their blood until early in a second pregnancy.

As a result of such findings, several workers, but particularly Bowman and colleagues in Canada, have advocated the use of additional anti-D during the course of pregnancy, starting at 28 weeks, in an attempt to reduce alloimmunisation to an absolute minimum.

ANTENATAL ADMINISTRATION OF ANTI-D IMMUNOGLOBULIN

  1. Top of page
  2. Abstract
  3. BACKGROUND
  4. FETOMATERNAL HAEMORRHAGE DURING PREGNANCY
  5. ANTI-D ALLOIMMUNISATION DURING PREGNANCY
  6. ANTENATAL ADMINISTRATION OF ANTI-D IMMUNOGLOBULIN
  7. SEARCH STRATEGY
  8. RESULTS OF CLINICAL STUDIES ()
  9. OPTIONS
  10. References
  11. Appendices

Dose and timing of anti-D administration

Based on observations of the incidence of FMH and anti-D alloimmunisation during pregnancy, an adequate dose of antepartum anti-D at 28 weeks could, in theory, prevent the 90% of intrapartum sensitisation that occurs as a result of occult FMH after 28 weeks’ gestation. A single dose of 300μg anti-D, as used in North America, would, in theory, result in approximately 20μg remaining in the mother at the time of delivery 12 weeks later, assuming a circulating half- life of 21 days for anti-D IgG. This would be sufficient to neutralise up to 1 ml fetal RBC, but the 1 % of women with FMH >1 ml at 30-39 weeks9 would not be protected. In principle, it is more effective to administer anti-D in divided doses, and 100μg at 28 and 34 weeks would result in a slightly higher residual anti-D at term.

Bowman9 observed that the same dose (300 μg), whether given by the intramuscular (i.m.) or intravenous (i.v.) route, resulted in a similar plasma level from 10-14 days onwards. At 10 weeks’ post injection, 0.7-2.6 ng/ml anti-D was detected in the circulation (equivalent to 6-21 μg total anti-D in the mother; 1 ng/ml in the circulation = 8.4μg in intra- and extravascular compartments), but at 12 weeks’ post injection, some women had no detectable anti-D (< 0.5 ng/ml) and the average was approximately 1 ng/ml (8.4μg in total). This is less than calculated, probably due to interaction with, and clearance of, fetal RBC from the circulation. Toveys5 observed that 14% of women given 100 μg at 28 and 34 weeks had anti-D detectable at term by enzyme-only tests, all < 0.1 IU/ml (<20ng/ml), the limit of the assay being 0.0 1 IU/ml (2 ng/ml), but exact figures were not given.

In practice, the observed failure rates of either pro- gramme are similar (see later); the divided dose schedule requires less anti-D, but incurs greater costs of administration and the respective cost effectiveness is addressed in a companion paper in this supplement (Cairns 1998).

Safety of antenatal anti-D

Two main concerns exist - the risk of ‘augmentation’, or enhanced anti-D immunisation, and the effect of passive anti-D on the fetus. In theory, the existence of low levels of passive anti-D in the maternal circulation on exposure to RhD positive RBC could result in enhancement of a primary immune response, as has been observed in experimental models1,3 - the opposite of what is intended. In practice, this has not been observed during trials5, and the experimental data in immunised volunteers are conflicting14,15.

Regarding transplacental transfer of IgG affecting the fetus, in theory a 300μg dose to the mother may result in the transfer of up to 10% to the fetus, i.e. a maximum of 30μg, based on transfer in immunized women16. The anti-D might be expected to cause fetal anaemia, but observations during trials do not confirm this. Bowman12 observed 28% of ABO compatible babies had a weak positive direct antiglobulin test (DAT) after administration of 300μg at 28 and 34 weeks. None had anaemia or hyperbilirubinaemia sufficient to warrant phototherapy (<3.4 mg/dl =58 μMol/L). The UK study5 with 100mg at 28 and 34 weeks identified 21/1238 DAT positive babies, mainly ABOincompatible, and no adverse effects. In a Swedish study, 250μg at 34 weeks17 resulted in l0/529 RhD-positive babies with a positive DAT, and no significant differences were seen in cord haemoglobin and bilirubin between the RhD- ositive and RhD-negative babies. In a Danish study18, 9% of 609 women given 300 μg anti-D at 28 weeks had anti-D in the circulation at delivery, but no instances of positive DAT, anaemia or neonatal jaundice were seen in the 346 RhD-positive babies delivered. The conclusion is that passive anti-D crosses the placenta and the highest dosage used can interact with fetal cells to produce a weak DAT, but this is insufficient to cause observable haemolysis or anaemia in the fetus or neonate.

Some concerns have been expressed about possible adverse effects of anti-D immunoglobulin, such as a trend towards increased fetal loss when given before 20 weeks19 but examination of perinatal mortality and morbidity in infants whose mothers received anti-D after second-trimester amniocentesis did not confirm these findings20, and extensive review of outcomes in the Yorkshire trial did not confirm an adverse effect of routine antepartum anti-D in the absence of amniocentesis4. On the other hand, the RhD-negative babies of mothers given antenatal prophylaxis are exposed to anti-D immunoglobulin during pregnancy without benefit, and there has been no systematic study of outcomes in these babies. However, the fetus is exposed to so little anti-D IgG compared to the amount of IgG transferred naturally from mother to baby, that it is extremely unlikely to produce any unrecognised adverse effect.

There are also potential risks due to transfer of viral infection by the anti-D, which is a blood product. Transmission of hepatitis C to RhD-negative women by a contaminated batch of anti-D has been noted in Ireland2. This product was manufactured for intra- venous use by a chromatography process, and became contaminated as a result of faulty procedures prior to the introduction of hepatitis C virus screening of blood donors. Anti-D is manufactured in the UK by a different process (cold-ethanol precipitation) which is known to be safer with respect to virus transmission, and there have been no recorded incidents of viral transmission in the UK (or North America) since the introduction of prophylaxis in the early 1970s. Nevertheless, a potential risk remains of transmission of known and unknown viruses for which screening tests are not available, and manufacturers are developing virucidal processes to reduce any risk even further.

SEARCH STRATEGY

  1. Top of page
  2. Abstract
  3. BACKGROUND
  4. FETOMATERNAL HAEMORRHAGE DURING PREGNANCY
  5. ANTI-D ALLOIMMUNISATION DURING PREGNANCY
  6. ANTENATAL ADMINISTRATION OF ANTI-D IMMUNOGLOBULIN
  7. SEARCH STRATEGY
  8. RESULTS OF CLINICAL STUDIES ()
  9. OPTIONS
  10. References
  11. Appendices

To identify clinical studies and trials involving antenatal anti-D prophylaxis, the following electronic databases were searched: Medline 1966-1 997, Embase 1987-97, the Cochrane Library. In addition, the reference sections of general review and original articles pertaining to clinical trials were searched manually, and cross referenced. Copies of the full articles pertaining to the clinical studies reported here were obtained and examined.

RESULTS OF CLINICAL STUDIES (Table 3)

  1. Top of page
  2. Abstract
  3. BACKGROUND
  4. FETOMATERNAL HAEMORRHAGE DURING PREGNANCY
  5. ANTI-D ALLOIMMUNISATION DURING PREGNANCY
  6. ANTENATAL ADMINISTRATION OF ANTI-D IMMUNOGLOBULIN
  7. SEARCH STRATEGY
  8. RESULTS OF CLINICAL STUDIES ()
  9. OPTIONS
  10. References
  11. Appendices
Table 3.  Summary of clinical trials of antenatal anti-D (see appendices for details).
ReferenceDoset† (μg anti-D)ParityTrial group*Control groupEvidence levels§
  1. † At 28 weeks for a single dose: at 28 and 34 weeks for two doses; * no exclusion of cases with anti-D detected before 28 weeks; 8 evidence-based levels of RCPEd (see appendix 11).

Appendix (A)122 × 300Prirnigravidae1/1205 at delivery (0.1%)4512768 at delivery (1.6%)IIa
   1/905 at 6min  
Appendix (B I)21 × 300Prirnigravidae2/1086 at delivery (0.18%)45/2768 at delivery (1.6%)IIa
   2/5 12 at 6 rnin  
  Multigravidae6/7 I9 at delivery (0.84%)17/765 at delivery (2.2%) 
Appendix (Bla)9I x300Unselected2519295 at delivery (0.27%)62/3533 at delivery (1.8%)III
Appendix (B2)IXI × 300Unselected0/291 at 10 min (432%)6/322 at l0min (I 236%)IIa
Appendix (C)17I × 250 (at 34 weeks)Unselected5/529 at 8 min (0.95%)101645 at 8 rnin (1 .6%)Ila
Appendix (DI)52 × 100Primigravidae2/1238at6min(0.16%)I8/2000 at delivery (0.9%)Ila
Appcndix (Dla)42 × l00Primigravidae4/1238 (0.32%)19/2000 (0.95%)Ila
Appendix (D2)32 × 100Primigravidae0/362 at 2-12 min (<0.28 %)4/360 at 2-12 min (1.11%)Ib
  Primiparae1/472 at 2-12min (0.21%)7/468 at 2-1 2 min (1.5%) 
  Unselected1/599at 2-12min (0.17%)7/590al2-12min (1.19%) 
Appendix (E)62 × 50Primigravidae4/5 13 at 6min (0.78%)9695 at 6 min (I 5%)Ih

To determine RhD protection rate it is necessary to determine the percentage of women who would have been alloimmunised without protection, the number given prophylaxis, and the number alloimmunised despite prophylaxis. The identification of intrapartum alloimmunisation as being the ‘true’ cause of immunisation is best assessed by the incidence at delivery in ABO-compatible RhD-negative primigravidae with no history of transfusion, or precipitating event requiring prophylactic anti-D. Not all studies have identified separate data for primigravidae, and none has been restricted to ABO compatible pregnancies, as in the original studies of postpartum prophylaxis.

Studies of effectiveness of anti-D prophylaxis have the advantage of a defined end-point, i.e. appearance of anti-D in the plasma of the women. However, the sensitivity of the assays will influence the detection rate, with assays such as enzyme-enhanced agglutination or Auto Analyzer quantification being more sensitive than antiglobulin tests; also, more recent AHG techniques tend to be more sensitive due to better quality reagents. Where historical controls have been used, they invariably antedate control groups by several years and detection of alloimmunisation may err on the side of underestimation, if anything, therefore underestimating the degree of protection by additional anti-D in the treatment group.

It is also necessary to distinguish between active alloimmunisation and prophylactic anti-D at the time of delivery. For rigorous analysis, it is conventional to repeat the antibody screen after 6 months or so, by which time passive anti-D would have been catabolised, with alloimmune anti-D persisting.

The ‘true’ rate of alloimmunisation is less than identified by the appearance of anti-D, as a proportion of women are ‘sensitised’ and will give a secondary immune response during a second challenge, such as FMH during a later pregnancy, but do not have detectable anti-D during or after the current pregnancy. This figure can never be known with certainty, but appearance of anti-D before 28 weeks in a subsequent pregnancy is a strong indication of alloimmunisation in the earlier pregnancy. Rigorous analysis of antenatal anti-D efficacy should preferably include second pregnancies in primigravidae.

A number of authors justify exclusions of alloimmunisations prior to 28 weeks in the treatment group, or where there have been events likely to precipitate FMH. Whilst justified scientifically, in proving efficiency, this approach will overestimate efficiency under routine practice conditions rather than in the context of a clinical trial. In determining the likely achievable reduction in practice, ideally there should be no exclusions in either group, i.e. treating both arms of the study on an ‘intention to treat’ basis. This is particularly important in cost- benefit analysis so as not to overestimate the benefits of antenatal anti-D in routine use rather than clinical trial conditions.

OPTIONS

  1. Top of page
  2. Abstract
  3. BACKGROUND
  4. FETOMATERNAL HAEMORRHAGE DURING PREGNANCY
  5. ANTI-D ALLOIMMUNISATION DURING PREGNANCY
  6. ANTENATAL ADMINISTRATION OF ANTI-D IMMUNOGLOBULIN
  7. SEARCH STRATEGY
  8. RESULTS OF CLINICAL STUDIES ()
  9. OPTIONS
  10. References
  11. Appendices

Although there is only one randomised controlled clinical trial (with small numbers) demonstrating a further reduction in alloimmunisation following ante- natal administration of anti-D1 meriting a grade A recommendation (see appendix II), the total body of evidence of efficiency is compelling. Whilst two doses of 300 μg are effective, this is no more so than the single dose in practice, and as it requires considerably more anti-D immunoglobulin, it is probably not cost effective. If a single dose is to be given, it is too late at 34 week17, and 28 weeks is to be recommended. If divided doses are to be given at 28 and 34 weeks, 50 μg is insufficient6.

Two dose regimes can be recommended - a single dose of 300 μg at 28 weeks, or two doses of 100 μg at 28 and 34 weeks, respectively:

  • i) 
    single dose of 300 μg at 28 weeks - the results of the single 300μg dose in first pregnancies is limited to the Canadian study2 with observed reduction from 1.6% (45/2768) in concurrent nonrandomised control to 0.18% (2/1086) in the treatment group.
  • ii) 
    two doses of 100 pg at 28 and 34 weeks - the two controlled studies give similar results in first pregnancies3,4. A reduction in alloimmunisation is seen from 1- 1 1 % (4/360) in controls to <0.28% (0/362) in the treatment group in the French study3’, and from 0.95% (19/2000) in controls to 0.32% (4/1238) in the treatment group in the English study4.

Whilst anti-D is in limited supply, it is more cost effective to restrict antenatal prophylaxis to first pregnancies (and women with no living children), who are not already irnmunised5,7. It is also probable that a single dose of 250μg (as used in Europe) will be as effective in practice as the 3001.18 dose, given the limitations of the anti-D quantification assay, and the vial overfill introduced by manufacturers, but this has not been formally proven in clinical trials.

Although reduction in alloimmunisation can be demonstrated, and attributed to antenatal administration of anti-D with some certainty, it is more difficult and complex to confirm that antenatal anti-D will have a significant effect on RhD-associated mortality. The number of RhD deaths is very low, even with standard postpartum prophylaxis, largely due to improvements in fetomaternal medicine and intensive neonatal care. However, it has been argued that there is a systematic underestimate in the UK, due to fetal deaths before 28 weeks (now reduced to 24 weeks) being recorded as ‘abortion’, and not HDN in official statistics21. A recent survey of maternity records in Scotland has also revealed that there are approximately five times as many RhD losses to alloimmunised women as appear in official statistics7. There is also a considerable cost to the NHS of caring for immunised women and their babies, which would be avoided even if there is ultimately a favourable outcome to the pregnancy. There have been no systematic studies on the reduction in mortality observed with antepartum anti-D. Although Bowman9,12 has calculated the losses prevented in the light of historical control data, this overestimates the contribution of antepartum prophylaxis, since a systematic reduction has also been observed in the UK without antepartum anti-D, for the reasons given above22. Nevertheless, it is self evident that if immunisation is largely prevented, then so will fetal morbidity and mortality and the magnitude of such a reduction is best measured in terms of contemporary, rather than historical data when assessing the cost-effectiveness of antepartum anti-D7.

References

  1. Top of page
  2. Abstract
  3. BACKGROUND
  4. FETOMATERNAL HAEMORRHAGE DURING PREGNANCY
  5. ANTI-D ALLOIMMUNISATION DURING PREGNANCY
  6. ANTENATAL ADMINISTRATION OF ANTI-D IMMUNOGLOBULIN
  7. SEARCH STRATEGY
  8. RESULTS OF CLINICAL STUDIES ()
  9. OPTIONS
  10. References
  11. Appendices

Appendices

  1. Top of page
  2. Abstract
  3. BACKGROUND
  4. FETOMATERNAL HAEMORRHAGE DURING PREGNANCY
  5. ANTI-D ALLOIMMUNISATION DURING PREGNANCY
  6. ANTENATAL ADMINISTRATION OF ANTI-D IMMUNOGLOBULIN
  7. SEARCH STRATEGY
  8. RESULTS OF CLINICAL STUDIES ()
  9. OPTIONS
  10. References
  11. Appendices

APPENDIX I Details of clinical trials of antenatal Anti-D

A) 2 × 300 μg at 28 and 34 weeks12

The first studies involved two doses of the standard Canadian/North American dose of 300 μg at 28 and 34 weeks. There were sound theoretical reasons for using two doses, because the circulating half life of IgC anti-D is 2 I days and the better protection against larger bleeds in the third trimester with the second dose. This schedule was clearly shown to be effective in a relatively small number of primigravidae. The authors quote a 94% reduction to <1/1000 at risk pregnancies (1 ×6% to 0.1%). If one excludes logistic failures, no cases of alloimmunisalion were seen in the treatment group.

Trial details: open study: nonrandomised concurrent controls; two urban hospitals in antenatal group; other three rural hospitals in control group; both groups given standard postpartum prophylaxis of 300μg at delivery.

Treatment group (1968-1976): (300μg at 28 and 34 weeks) 0/1204 primigravidac alloimmunised at delivery; 0/905 developed immune anti-D at 6months post delivery (one immunised before 28 weeks); operational failure rate in primigravidae = 1/1204 = 0.08%.

Control group (1967-1974): 45/2768 primigravidae alloimmunised at delivery; operational failure rate in primigravidae = 1.6%.

Conclusion: alloimmunisation can be reduccd from I .6%i to 0.1% in primigravidae (946 reduction) (x2= 16.14: P<0.001).

B1) 1 × 300 μg at 28 weeks2

Primarily because of anti-D shomges, a single dose at 28 weeks was evaluated early in the Canadian Rh programme as an alternative. In a small number of primigravidae, a single dose of 300μg at 28 weeks reduced alloimmunisation from 1.6% (45/2768) to 0.18% (2/1088). If logistic failures an: excluded, none of the treatment group was alloimmunised. For multigravidae, the observed failures were 0.846, but excluding women estimated to be sensitised in previous pregnancies, the ‘true'failure rate was estimated 3% 0.42%, compared to 2.2% in the control multigravid group.

Trial details: open study with historical controls: bolh groups given standard postpartum prophylaxis of 300 pg at delivery. Treatment group examined at 6 months for anti-D.

Treatment group (1975-1977): (300μg at 28 weeks).

Group I 0/1084 prirnigravidae alloimmunised at time of delivery (expected 14/1084): 0/512 developed immune anti-D after 6months: (two alloimmunised before 28 weeks at 12.5 and 6.5 weeks); operational failure rate in primigravidae is 2/1088 =0-18%.

Group 2 6/719 rnultigravidae alloiinmunised at time of delivery; 3/716 estimated to have a ‘secondary response’ from an earlier pregnancy: 3l7 16 estimatcd ‘true failures’ (expected 127/16); ‘true failures’ in multigravidae 3/716 = 0.42%. operational failure rate in multigravidae is 6/716 = 0.84%.

Control group (1967-1974): 45/2768 primigravidae alloimmunised at delivery; opcralionnl failure rate in prirnigravidae = 1.6%. Of the 45 primigravidae it was noted that 3/45 (6.67%) were alloimmunised before 28 weeks (3/2768 = 0.108%) and 42/45 (93.33%) were alloimmunised after 28 weeks (42/2768 = 1.52%). 17/765 multigravidae alloimmunised at delivery; operational failure rate in multigravidae = 2.2%.

Conclusion: excluding other causes, antenatal alloimmunisation would reduce from 1 6 8 to 0.18 in primigravidae, and from 2.2% to 0.42% in multigravidac (x2= 12.29: P<0.001). Including all patients in treatment and control arms, for primigravidae. alloimmunisation is reduced from 1.6% to 0.18% (88.75% reduction) for multigravidae, from 2.2% to 0.84% (X2= 3 81: P =0.051)

B1a) 1 × 300 μg at 28 weeks9

This paper is an update on the earlier trial with additional numbers, over 8.5 years to February 1986. All cases reported in the earlier trials are included. There is no distinction made between primigravidae and parous women. although it is estimated that 78%: were primigravidae. The dose also changes from an estimated 24Opg for the first 5 years, with a subsequent change to 300μg as standard. It is also not clear at what point the standard postpartum dose changed to l20μg i.v. Nevertheless, the study has value as the largest number of women evaluated under routine conditions, as opposed to clinical trial conditions. It was estimated that the treatment failures (i.e. excluding alloimmunisation before 28 weeks and other causes during pregnancy) were 0.18% (I7/9295). The observed alloimmunisation rate in the treatment groups fell from an estimated (historical) 1.8% (62/3533) to 0.32% (30/9295) in routine practice. Excluding other causes, and including all women regardless or parity, the authors conclude that antenatal alloimmunisation can be reduced from 1.4 % to 0.278 (25/9295) but this method of calculation overestimates the control failures, and underestimates the treatment failures, and gives a more favourable outcome to antenatal prophylaxis. Nevertheless, the reduction in alloimmunisation as a result of the programme is significant, and evidence of the efficacy of prophylactic antenatal anti-D in routine practice.

Trial details: open study with historical controls in one province in Canada. Both groups given standard postpartum anti-D it is not clear if the dose reduced to 120μg i.v. from 300μg, and at what point.

Treatment group (1977–1986): (1977–83 = 240μg = 1200IU at 28 weeks). (1983–86 = 300μg = 1500IU at 28 weeks). 25/9295 alloimmunised (not all primigravidae). failure rate = 0.27%. 8/25 (32%) alloimmunised < 28 weeks: 17/25 (68%) treatment failures. treatment failure rate (17/9295) = 0.18% NB: actual total failures over the time was 30/9295 = 0.32%.

Control group (1967–74 data — same as in earlier studies): 62/3533 total (i.e. not all primigravidae) [note: this also includes anti-D detected up to 3 days after delivery, which gives a higher figure than reported by othcr groups]; operational failure rate = 1.8%. Of the 45 primigravidae in this group it was noted that 3/45 (6.67%) were alloimmunised before 28 weeks (3/2768 = 0.1088) and 42/45 (93.33%:) were alloimmunised alter 28 weeks (42/2768 = 1.52%).

Conclusions: excluding other causes. and including all women regardless of parity. antenatal alloimmunisation can be reduced from 1.8% to 0.274 = 854, reduction in alloimmunisation (x2= 81.72: P < 0.001). Under routine conditions, with no exclusions, anti-D alloimmunisation can be reduced from 1.8% to 0.32% (78% primigravidae in the group) = 82% reduction in alloimmunisation (x2= 78.1: P < 0.001).

B2) 1 × 300 μg at 28 weeks18

In this small study using the same dose as the Canadian trials, alloimniunisation was reduced to less than 0.32% in the treatment group (0/291) compared to historical controls with 146% alloimmunisation.(6/322). The study includes women of all parities, and the data are comparable in practice to the larger Canadian study. confirming the efficacy of a single dose of anti-D i n reducing alloimmunisation. It was noted that 9% of women had anti-D at delivery in the treatment group. but no positive DAT. jaundice or anaemia was observed in the babies. Kleihauer estimation of FMH at delivery was also carried out in both groups. The number of observed FMH was the same in both groups — 63 in treatment group versus 72 in control group. but the number of l a r p FMH (> I ml) was significantly different 5/63 in treatment group versus 28/72 i n the control group (x2: P < 0.001).

Trial details: open study at one hospital in Denmark. with historical controls. Both groups given standard postpartum prophylaxis of 200μg Both groups stated to be comparable as regards first pregnancies, risk events, instrumentation and Caesarian Section rate. Postpartum 10-month follow-up sample obtained in both groups.

Treatment group (1980–85): (300μg at 28 weeks) 609 women; 346 with RhD-positive baby: 0/291 alloimmunised at 10months = <0.32 % (assuming 1/29?).

Control group (historical controls 1972–77): 354 with RhD-positive baby: 6/32? alloimmunised at 10months = 1.86% (or during subsequent pregnancy). Treatment versus controls t-test; P < 0.45.

Conclusion: alloimmunisation can be reduced from 1.86% to <0.32%(83% reduction) (x2= 3.72: P= 0.054).

C) 1 × 250 pg at 32 weeks17

In this study, the European dose of 250μg (rather than 300μg) was given later in pregnancy. at 32–34 weeks. These doses are not signilicantly different, given the limitations of the anti-D assays. Although a reduction in alloimmunisation was observed. from 1.6%) (10/645) to 0.38% (2/529) in the treatment group regardless of parity, including those women immunised during pregnancy before 32 weeks raised the failure rate to 0.95%. In practice. therefore, it is demonstrated that a single dose at 32–34 weeks is too late to provide sufficient protection against intrapartum sensitisation. as might be expected from the natural history of detectable FMH during pregnancy.

Trial details: open study at one hospital in Sweden with historical controls. Both groups given standard postpartum prophylaxis of 250μg anti-D with 8-month follow-up sample for anti-D.

Treatment group (1973–77): (250μg at 32–34 weeks) 830 women 529 with RhD-positive baby: 2/529 alloimmunised = 0.38%. 1/236 primigravidae = 0.42%. 1/293 multigravidae = 0.34%. 0/59 alloimmunised in second pregnancy. NB: three primigravidae excluded because anti-D before 32 weeks; if included, operational failures = 5/529 = 0.95%.

Control group (historical control dates not given): 10/645 alloiminunised = 1.6%. 5/286 primigravidae = 1.75%. 5/359 multigravidae = 1.39% (1.6 versus 0.38 =P < 0.05). The authors note that if weak anti-D is considered to be due to prophylaxis, then 0/529 in trial group, and 9/645 could be considered alloimmunised =P < 0.005.

Conclusions: excluding logistic failures, alloimmunisation can be reduced from 1.6% to 0.38% (x2=2.87: P= 0.09). When including immunisation before prophylactic anti-D at 32–34 weeks. the reduction is 0.95% (x2: P=0.511). In primigravidae. the reduction is from 1.75% to 0.42%. with small numbers in each group (x2= 1.0: P= 0.317). Giving anti-D at 32–34 weeks is therefore probably too late to be most effective.

D1) 2 × 100 μg at 28 and 24 weekss

This was the first study using a divided dose of anti-D calculated t o give the same (or more) anti-D in the maternal circulation at term 300 μg dose at 28 weeks. and specifically addressed first pregnancies. The impetus for such an approach in the UK was primarily hased o n economic use of anti-D, with emphasis in first pregnancie greater benefit with limited amounts of anti-D available. First pregnancies (2069) were treated, in comparison with 2000 historical control primigravidae. This initial report showed a reduction from 0.9% in controls to 0.16% in the treatment group, a similar figure to that demonstratcd by Bowman et al with a single 300μg dose (0.18%).

Trial details: multicentre open study in first pregnancies, with historical controls. Both groups given standard postpartum prophylaxis of 100 μg i.m. Six-monthly testing for anti-D was undertaken in some women.

Treatment group (1980–81): (100μg at 28 and 34 weeks) 2/1238 primigravidae alloimmunised at delivery and 6 months = 0.16% (1059 tested at 6 months).

Control group (historical 1978–79): 18/2000 primigravidae alloimmunised = 0.9%.

Conclusion: Including all Failures, alloimmunisation can be reduced in primigravidae from 0.9% to 0.16% (82% reduction), (x2=5.64: P= 0.018).

D1a) 2 × 100 μg at 28 and 34 weeks4

In a follow-up report to the original study, the treatment cohort had been followed longer. and a more detailed breakdown of the control data was given. If only failures after the first pregnancy are considered, the control alloimmunisation rate is 0.95% (19/2000); in the treatment group, it is only 0.32% (4/1238), giving a 66% reduction in alloimmunisation. It is not possible t o identify the comparative immunisation rate in both groups in subsequent pregnancies, as the original antenatal prophylaxis group were only given standard postpartum anti-D. However, there was a suggestion of protection persisting into subsequent (up to four) pregnancies in the original trial group, with a total of six failures in 2041 RhD-positive pregnancies (0.29%). compared to 32/2985 (1.07%) in RhD-positive pregnancies in the original historical cohort.

Treatment group (1980–81): 4/1238 primigravidae alloimmunised first pregnancy = 0.32%; 1/604 in second (not given antenatal prophylaxis)= 0.17%. 6/2041 all pregnancies including third and fourth = 0.29%.

Control group (historical 1978–79): 19/2OO0 primigravidae alloimmunised during first pregnancy = 0.95% (9/751 women alloimmunised in second = 1.2%; 32/2985 all pregnancies = 1.07%).

Conclusion: in first pregnancies, alloimmunisation can be reduced from 0.95 to 0.32% (66% reduction), (x2=3.42: P= 0.064).

D2) 2 × 100 μg at 28 and 34 weeks3

A second trial also using 100μg in two doses of anti-D was reported from France, this being the only randomised trial reported with a protective dose of antenatal anti-D administration. Considering first pregnancy only. the reduction in alloimmunisation was from 1.5% (7/468) to 0.21% (1/472). which is a similar reduction to that observed by Thornton et al. (1989) with first pregnancies receiving 2 × l00μg anti-D. A signficiant reduction in primigravidae was also observed. from 1.11% (4/360) to <0.28% (0/360). Kleihauer tests to identify FMH were also carried out in both groups. There was no difference observed with the frequency of larger FMH (>1/10000 maternal RBC) at 32–34 weeks — 31 versus 32 in the control group. On the other hand, fewer FMH in total were seen in the treatment group — 4.2% (39/927) versus 7% (67/957) in the control group, indicating that the 28-week anti-D was effective in eliminating small FMH. Likewise at delivery of a RhD-positive baby, 12.2% (73/599) in the treatment group had a positive Kleihauer test compared to 20.2%; (119/590) in the control group.

Trial details: open controlled study involving several centres in Paris. allocated to treatment or control arms by even/uneven date of birth. Both groups received standard postpartum anti-D of 100μg i.v. (the standard French practice). Serum was examined for anti-D at birth and at 2–1 2 months postpartum.

Treatment group: (100μg anti-D i.m. at 28 and 34 weeks) 927 women — 599 with RhD-positive baby — 461 primigravidae/l38 multipara (472 followed for 2–12months); 0/599 alloimmunised at delivery, 1/599 followed 2–12months (at 5 months) = 0.17%, 1/472 primiparae followed 2–1 2 months (at 5 months) = 0.21, 0/362 primigravidae followed 2–12 months = < 0.28%.

Control group: 955 women — 590 with RhD-positive baby — 454 primigravidae/l36 multipara (468 followed for 2–12 months); 6/590 alloinimunised at delivery = 1.02%, 7/468 primiparae followed 2–12 months = 1.92, 4/360 primigravidae followed 2–12 months = 1.11%.

Conclusion: considering primigravidae only, the reduction in alloimmunisation is from 1.1l% to <0.28% (x2=2.28: P= 0.131). Considering primipara only, the reduction in alloimmunisation is 1.58 to 0.21% (86% reduction) (x2=3.22: P= 0.073). The reduction in alloimmunisation observed. considering all groups, is from 1.19% (7/590) to 0.178 (1/599) (x2 3.22 : P= 0.073).

E) 2 × 50 μg at 28 and 34 weeks6

A prospective study to evaluate the efficacy of a lower dose of 50μg at 28 and 34 weeks was undertaken in the UK, but was stopped due to a combination of low recruitment, and evidence that protection against alloimmunisation was incomplete. Although reduction was substantial, 1.5% (9/595) in controls to 0.78% (4/513) in the treatment group, this was less than that observed in the UK trial of 2 × 100μg (Thornton et al. 1989) and it was concluded that the lower dose could not be recommended.

Trial details: open controlled multicentre study in primigravidae with randomisation to treatment or control groups by sealed envelope. Both groups received standard postpartum prophylaxis of 100μg i.m. Both groups examined for alloimmunisation at 6months. Women with preexisting anti-D at 28 weeks excluded.

Treatment group: (50μg i.m. at 28 and 32 weeks) 4/5 13 alloimmunised at delivery = 0.78%; one extra alloimmunised, but antenatal event not given anti-D therefore excluded.

Control group: 7/595 alloimmunised at delivery = 1.2%; 9/595 alloimnunised at 6months follow-up = 1.5%.

Conclusion: although a reduction in alloimmunisation was observed from 1.5% to 0.788, this was less than observed with 2 × 100μg, and 2 × 50μg is insufficient to give full protection (x2=0.72: P=0.395).

APPENDIX II Table of evidence-based levels

LevelType of evidence la
IaEvidence obtained from meta-analyis of randomised controlled trials.
IbEvidence obtained from at least one randomised controlled trial.
IIaEvidence obtained from at least one well-designed controlled study without randomisation.
IIbEvidence obtained from at least one other type of well-designed quasi-experimental study.
IIIEvidence obtained from well-designed non-experimental descriptive studies, such as comparative studies, correlation studies and case studies.
IVEvidence obtained from expert committee reports or opinions and/for clinical experiences of respected authorities.
GradeRecommendation
A(Evidence Levels la, Ib)Required — at least one randomised controlled trial as part of the body of literature of overall good quality and consistency addressing specific recommendation.
B (Evidence Levels Ila, Ilb. III)Required — availability of well conducted clinical studies but no randomised clinical trials on the topic of recommendation.
C (Evidence Level IV)Required — evidence obtained from expert committee reports or opinions and/or clinical experiences of respected authorities. Indicates absence of directly applicable clinical studies of good quality.