Abstract
Anaemia in pregnancy is common, however, only a few cases of pregnancy-associated autoimmune haemolytic anaemia have been documented. Typically, such cases involve a positive direct antiglobulin test and have the potential to cause haemolytic disease of the fetus and newborn. Rarely, no autoantibodies are detected. We report two cases of direct antiglobulin test negative haemolytic anaemia occurring in multiparous women with no cause found. Both women had a haematological response to corticosteroid therapy and delivery.
Introduction
Anaemia is a common phenomenon in pregnancy, affecting 30–40% of all pregnancies. 1 The incidence of autoimmune haemolytic anaemia (AIHA) in the general population is 1 in 100,000 however there are only limited reports in pregnancy. 2 Typically, the hallmark feature is a positive direct antiglobulin test (DAT). 5–10% of AIHA cases, however, are associated with a negative DAT therefore it is important that a diagnosis of AIHA is not excluded on this basis. 1
We present two cases of pregnant women with DAT‐negative AIHA. Both women received oral corticosteroids antenatally and demonstrated biochemical response to immunosuppression and subsequently delivery.
Case 1
A 30-year-old woman consulted at 27 weeks' gestation in her second pregnancy for progressive lethargy. Her first pregnancy had been uncomplicated. Conjunctival pallor was the only abnormality on examination.
Her haemoglobin had dropped to 90 g/L from a baseline of 126 g/L prior to pregnancy. Red cell indices and blood film review were consistent with the noted history of beta-thalassaemia trait with a mean cell volume of 65 fL (normal 70–100 fL), haematocrit 0.28 (0.30–0.48), mean cell haemoglobin 22 pg (27–32 pg) and mean cell haemoglobin concentration 318 g/L (310–360 g/L). Reticulocyte count was elevated at 4.5% (0.2–2.0), serum haptoglobins were reduced at <0.08 g/L (0.40–2.4 g/L), polyspecific DAT was negative, total bilirubin was elevated at 19 µmol/L (3–15 µmol/L) and lactate dehydrogenase was normal at 158 μ/L (120–250 μ/L). Blood group antibody screen, haematinics, viral serology, renal function, paroxysmal nocturnal haemoglobinuria flow cytometry and autoimmune screen were unremarkable. The findings were consistent with a diagnosis of pregnancy‐associated haemolytic anaemia.
After commencement of prednisolone 15 mg per day the haemoglobin improved to 104 g/L with normalisation of serum haptoglobins to 1.96 g/L. The reticulocyte count remained elevated at 3.1%. The patient remained on corticosteroids for the remainder of her pregnancy with haemoglobin being maintained at more than 100 g/L. Stress dose intravenous hydrocortisone was given intrapartum. Rapid wean of steroids post-partum was performed with no evidence of recurrent haemolysis. The baby was well with no signs of anaemia.
Case 2
A 34-year-old woman with dichorionic diamniotic twins presented at 24 weeks' gestation with progressive lethargy and dark urine. This pregnancy was the result of in-vitro fertilisation. She had had one previous pregnancy, which had been complicated by post-partum haemorrhage secondary to retained products of conception resulting in Asherman's syndrome. She also had a bicornate uterus and a history of Hashimoto's thyroiditis.
At the time of presentation her haemoglobin was 94 g/L, mean cell volume 105 fL (81–99 fL), haematocrit 0.28 (0.30–0.48), mean cell haemoglobin 35.1 pg (27–32 pg) and mean cell haemoglobin concentration 336 g/L (310–360 g/L). Blood film review confirmed the presence of occasional spherocytes. Reticulocytes were elevated at 3.9% (0.2–2.0%), polyspecific DAT was negative, serum haptoglobins were reduced 0.16 g/L (0.40–2.4 g/L), total bilirubin was normal at 6 µmol/L (3–15 µmol/L) and lactate dehydrogenase was increased 272 μ/L (120–250 μ/L). Blood group antibody, autoimmune screen and haematinics were unremarkable.
At 27 weeks' gestation the patient received two doses of intramuscular betamethasone due to concern about evolving pre-eclampsia. She was then transitioned over to oral prednisolone 40 mg once daily with prompt normalisation of serum haptoglobins. At 30 weeks' gestation the woman developed pre-eclampsia necessitating caesarean section. This was complicated by placenta accreta and massive post-partum haemorrhage. The prednisolone was weaned postpartum with no evidence of recurrence of haemolysis. Both twins had a negative DAT and normal haemoglobin at delivery.
Discussion
Anaemia in pregnancy is defined by the World Health Organisation as Hb less than 110 g/L (105 g/L in second and third trimesters). 3 Causes to consider include physiological, nutritional, defective haemoglobin or erythrocyte synthesis, haemolysis, and infection (Table 1). History, physical examination and corresponding investigations including blood film review are therefore essential. In the two cases we present, such secondary causes were excluded. Whilst pregnancy can be associated with a mild reticulocytosis, the combination of reduced serum haptoglobins and a negative DAT favoured a diagnosis of pregnancy‐associated haemolytic anaemia.4–6
Causes of anaemia in pregnancy. 3
Where DIC = disseminated intravascular coagulation, TTP = thrombotic thrombocytopenic purpura, HUS = haemolytic uraemic syndrome, aHUS = atypical haemolytic uraemic syndrome, TB = Tuberculosis, HIV = human immunodeficiency virus, CMV = cytomegalovirus
Autoimmune haemolytic anaemia (AIHA) involves red cell autoantibodies directed against one's own red cell antigens leading to their premature destruction by the spleen or reticuloendothelial system. The antibodies are typically classified into warm (typically IgG isotype), cold (typically IgM isotype) or mixed (IgG and IgM isotype) according to the optimum temperature for haemolysis. 2 As IgG crosses the placenta, warm AIHA has the potential to cause severe anaemia in the fetus or neonate. 7
AIHA can be primary or secondary to malignancy, infections, immunodeficiency and medications. 8 Despite extensive investigation the cause of the aberrant antibody production in both cases remained unknown; however, normalisation of serum haptoglobins in response to corticosteroids and delivery suggests a placentally-derived immune or hormonal aetiology. 9
Typically, the hallmark feature in AIHA is a positive DAT demonstrating the presence of antibodies bound to the red cell surface. Potential causes for a negative DAT include technical errors, use of antisera lacking the corresponding antibody subclass (most polyspecific reagents including those used in these two cases, have little activity against IgA and IgM) and the presence of an antibody that is readily dissociable or easily eluted, insufficient antibodies or homologous destruction of coated red cells. 10 Flow cytometry has been proven to be a more sensitive but was not performed in either case. 11
Corticosteroids are usually the first‐line treatment. Haptoglobins were an early guide to treatment efficacy in our cases. Immunosuppressive agents including azathioprine, and cyclophosphamide have been documented with success. The introduction of anti-CD20 monoclonal antibodies (e.g. Rituximab) have been reported to have a favourable response however the potential for interfering with neonatal immunity should be considered. 7 As it typically presents around the third trimester and often remits postpartum one must also consider delivery as a treatment option for severe, refractory cases. 9
In conclusion we present two cases of pregnancy-associated immune-mediated haemolytic anaemia. Unlike typical autoimmune haemolytic anaemia our cases had a negative DAT with no other cause found. Both women were commenced on prednisolone and demonstrated a favourable response. Neither woman has gone on to have subsequent pregnancies but if so, it would be advisable to monitor them regularly given the potential risk of recurrence.9,12
Footnotes
Acknowledgements
None.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.
Ethical approval
St John of God Hospital and King Edward Memorial Hospital do not require ethics approval for reporting of individual cases or case series.
Informed consent
Written informed consent was obtained from all subjects before the study.
Guarantor
Dr Katherine Creeper
Contributorship
Dr Katherine Creeper was responsible for reviewing the literature and writing the manuscript. Dr Dorothy Graham was responsible for obtaining patient consent, reviewing and editing of the manuscript.
