Autoimmune haemolytic anaemia: emergency blood transfusion

Abstract

A 50-year-old woman, with a background of autoimmune haemolytic anaemia, presented to the emergency department with lethargy and shortness of breath. Investigations revealed a haemoglobin level of 50 g/L. High dose steroids were administered and blood transfusion prescribed. However, the blood transfusion was delayed due to a positive antibody screen and concerns regarding administering blood when the patient was pyrexic. The delay resulted in a significant deterioration in the patient’s clinical state with her haemoglobin level falling to 26 g/L 24 hours later. She was urgently transfused with blood and made a full recovery. This report analyses the delays for transfusion and how these could have been minimised. First, guidelines advise that emergency blood should be considered in life-threatening circumstances. Second, fever is not always a contraindication for transfusion, particularly in an emergency setting.

Background

This report describes a case of acute haemolysis in the context of warm primary autoimmune haemolytic anaemia (AIHA) that has relapsed despite third-line treatment. While this condition only affects 1 per 100 000 per year,¹ the key learning points here relate to emergency blood transfusions for acute haemolysis and also generally in patients who are feverish.

AIHA is a rare condition, where red cell antigens are targeted and attacked by the host immune system, resulting in haemolysis. Typically patients may present with symptomatic anaemia (weakness, dyspnoea, dizziness) or less frequently jaundice and dark urine.² A positive direct antiglobulin test (DAT) supports the diagnosis, which indicates presence of either IgM or IgG and complement C3. Evidence of haemolysis and the clinical picture are key additional elements for diagnosis. The disease is categorised serologically according to the optimum temperature of the antibodies; the most common—warm, cold and mixed. They are then further categorised into primary or secondary depending on the aetiology.³ The mainstay of treatment for this is immunosuppression, primarily steroids and second-line treatment with rituximab. Third-line treatment includes splenectomy and other immunosuppressive agents such as ciclosporin. The patient in this case was relapsing while on third-line treatment, attributable to the splenunculus that had developed.

In acute haemolysis, guidance recommends investigating underlying causes such as infection or drugs; then blood transfusion if necessary, high dose steroids and to consider use of immunoglobulins. Additionally, if anaemia is life threatening it is suggested to transfuse with ABO, Rh and K matched red cells rather than usually waiting at least 4–6 hours for full compatibility testing.³ The case highlights the need to ensure that this is actioned in practice as within those hours of waiting, haemolysis can develop rapidly. Additionally, given that concurrent infection may play a role in precipitating acute haemolysis, if emergency transfusion is required—it is essential that blood is administered despite fever.

Case presentation

A 50-year-old female patient with known warm AIHA was brought in by ambulance to the emergency department with a 2-day history of lethargy and dyspnoea. The patient was diagnosed with warm AIHA in 2005, with a positive DAT/Coombs test. She was trialled on azathioprine therapy, which failed but following splenectomy in 2006 had been in remission. She relapsed in August 2019 and was started on a month’s course of rituximab and prednisolone with only a short period of remission. Since then, she had commenced a prolonged course of co-trimoxazole for pneumocystis pneumonia. Following this, she was admitted multiple times, including with a seizure secondary to hyponatraemia and anaemia (haemoglobin 60 g/L) 2 weeks prior to the discussed admission. Medications included 60 mg prednisolone, folic acid, famotidine (switched from lansoprazole due to hyponatraemia) and weekly alendronic acid. She had no other relevant medical history.

On admission, she was cushingoid in appearance and feverish 37.7°C. She was tachycardic (115 beats/min), tachypnoeic (respiratory rate 30 breaths/min), maintaining oxygen saturation on 3 L of oxygen. Aside from mildly jaundiced sclera, further examination was unremarkable.

Investigations

Admission investigations revealed; haemoglobin 50 g/L (115–165 g/L), white cell count 19.4×10⁹/L (4–11×10⁹/L), neutrophil count 16×10⁹/L (1.7–7.5×10⁹/L), C reactive protein 28 mg/L (<5 mg/L) and bilirubin 59 µmol/L (<21 µmol/L). Reticulocyte count was initially within normal range, 75×10⁹/ L (20–80×10⁹/L), then increased to a peak of 407×10⁹/L over the next few days. Lactate dehydrogenase, a marker of haemolysis, levels were sent but haemolysed and were not subsequently measured in this case. The raised inflammatory markers alongside fever were suspected to be due to infection of unknown source, which was treated with broad spectrum antibiotics. The low haemoglobin and raised bilirubin, supported by the history, prompted suspicion of acute haemolysis. Chest X-ray was normal. Recent B₁₂ 380 pg/mL (180–1000 pg/mL) and folate 20 ng/mL (>4 ng/mL) showed no deficiency as this was being chronically managed. The following day, the haemoglobin level had fallen to 26 g/L.

Subsequent investigations demonstrated growth of splenic tissue and incidental finding of pulmonary embolus on CT scan (figure 1).

Figure 1.

CT scan demonstrating regrowth of splenic tissue.

Treatment

The on-call haematologist advised broad spectrum antibiotics, prednisolone dose to be increased to 100 mg and blood transfusion. However, the blood was not administered overnight. There were two main reasons for the delay. First, the process for full serological testing for patients with a positive red cell antibody screen requires transfer to the regional blood transfusion centre, 100 miles away, which takes several hours. The blood unit was ready 13 hours after admission. Second, due to ongoing fever there was a delay in administration and the transfusion was not given overnight. Twenty-four hours after initial presentation to the emergency department, no blood had been administered to the patient, she had clinically deteriorated with temperature 38°C, blood pressure 120/60 mm Hg, tachycardic (136 beats/min), tachypnoeic (36 breaths/ min) and oxygen saturation of 94% on 4 L of oxygen. A medical emergency call was sent out for the patient; the haemoglobin level had dropped to 26 g/L. The patient was resuscitated with a red blood cell transfusion. The red cells were fully cross matched, all units given were O-positive, R2r phenotype, kell negative, M negative and deemed safe to give. Over the next few days, the patient received 8 units of packed red cells, the haemoglobin level increased to 136 g/L and she became clinically stable. She was then subsequently treated with furosemide to manage mild fluid overload and steroids were slowly tapered down.

Diagnosis of pulmonary embolus was also made following CT scan report and commenced on anticoagulant therapy.

Outcome and follow-up

The patient was discharged on slow weaning dose of 100 mg prednisolone with famotidine and alendronic acid, 5 mg apixaban two times per day and commenced a second course of rituximab. Four months later, prednisolone has been tapered down slowly to 40 mg daily and her haemoglobin level had remained stable between 120 and 130 g/L and she is now being considered for surgical removal of splenic tissue. However, due to the COVID-19 pandemic, there are anticipated delays for this.

Discussion

Acute transfusion reactions are rare, febrile haemolytic reactions are the most reported of these and are usually mild.⁴ The British Society for Haematology suggests that in the case of an isolated fever, identified as a temperature of 38°C–39°C with a rise of 1°C–2°C from baseline, transfusion can be continued under observation. If there are additional features such as vomiting, chills, respiratory symptoms—the transfusion should be stopped and appropriately managed depending on the reaction.⁵ Guidelines recommend a baseline set of observations prior to transfusion, but do not specify when transfusion would be inappropriate based on this.⁶ However, given that febrile non-haemolytic reaction is usually mild—the concept that fever should delay urgent transfusion, particularly in this case, is unsupported.

In cases of acute autoimmune haemolysis, the British Society of Haematology recommends that if anaemia is life threatening in time taken for full compatibility testing (usually at least 4–6 hours), to transfuse with ABO, Rh and K matched red cells.³ These patients by the nature of their disease have red cell antibodies resulting in a positive red cell antibody screen, consequently the presence of alloantibodies should be excluded to prevent haemolytic transfusion reaction.⁷ This requires reference laboratory input, which takes time. About one-third of patients with AIHA have alloantibodies.⁸ A retrospective study demonstrated 32 patients with AIHA receiving least incompatible blood transfusions based on cross-match testing, in emergency, 29 of these did not have any detectable alloantibodies.⁹ All transfusions did not result in acute reaction or noticeable acceleration of haemolysis following transfusion by analysing lactate dehydrogenase and haemoglobin levels. The three patients who did not receive transfusion or it was delayed due to overestimation of positive cross match died as a result. Fortunately, in our case, despite the 24-hour delay, the patient survived with little sequalae. A more recent retrospective study in China, similarly used the least incompatible blood by cross-match testing for transfusion in AIHA for 269 patients, which did not alter the risk of transfusion-related reactions or accelerate haemolysis, particularly in patients with a haemoglobin level <50 g/L.¹⁰ This emphasises the case that in acute haemolysis, the priority is delivering the blood to the patient rather than delay with full serological testing. The decision regarding when emergency blood is required is based on clinical judgement and requires specialist input.

Patient’s perspective.

I remember not feeling quite right and asking my family how I looked, they did not think I looked so bad but then I became so tired that I could not walk the stairs, my husband called an ambulance. When I was there, I recognised some of the staff from my previous admissions, which was nice. I don’t remember much from that day, but I was thinking—when is the blood going to come—and asked a few people this. I remember from before that it would make me feel better. The next thing I remember is people standing around my bed, trying to stick needles in me.

It is taking a long time to feel better again, I have been feeling weak and did not realise having the multiple admissions last year would take so much strength. I am pleased that I am on the lowest dose of prednisolone for a long time now and my haemoglobin level is stable. I am awaiting an appointment with the surgeon to discuss my splenic tissue removal.

Learning points.

• In emergency, communication between the laboratory and clinician is key to ensure the safest type of transfusion is given within an appropriate time frame.

• Pretransfusion fever is not a contraindication for blood transfusion in severe anaemia.

• Ensure prompt specialist input from haematology in cases of suspected acute haemolysis for consideration of least incompatible blood transfusions based on cross match with ABO, Rh, K matched cells, if full compatibility testing may cause fatal delay.