Fatal case of delayed-onset haemolytic anaemia after oral artemether–lumefantrine
- Lotta Gustafsson ,
- Sunil James ,
- Yimeng Zhang and
- Karunakaran Pradeep Thozhuthumparambil
- Department of Acute medicine, Sandwell and West Birmingham Hospitals NHS Trust, Birmingham, UK
- Correspondence to Dr Sunil James; sunil.james1@nhs.net
Abstract
Artemisinin derivatives are used globally in the management of falciparum malaria. Postartemisinin delayed haemolysis (PADH) is a recognised adverse event contributing to severe anaemia. To the best of our knowledge, we report the first recorded fatal case of PADH. A 60-year-old woman presented with two episodes of collapse at home and feeling generally unwell. She had recently been treated for uncomplicated falciparum malaria 1 month prior with artemether 80 mg/lumefantrine 480 mg in Congo. Her results on admission revealed an anaemia (haemoglobin 43 g/L), raised lactate dehydrogenase and positive direct antiglobulin test that suggested an intravascular haemolytic process. She made a capacitous decision to refuse blood products in line with her personal beliefs. Despite best supportive treatment, she did not survive. This case highlights the importance of postartemisinin follow-up and should encourage discussion and careful consideration of its use in the context of lack of access to/patient refusal of blood products.
Background
In 2019, there were an estimated 229 million cases of malaria worldwide, with falciparum malaria being the most deadly.1 WHO currently recommends oral artemether–lumefantrine combination therapy and intravenous artesunate for the management of uncomplicated and complicated falciparum malaria, respectively.2 This recommendation is now globally adopted and artemisinin derivatives’ rapid and early clearance of parasites has been repeatedly shown to be clinically superior to alternative antimalarial therapies such as quinine in clinical trials.3 4
A recognised side effect of artemisinin is delayed haemolysis occurring greater than 1-week post-treatment (postartemisinin delayed haemolysis, PADH).5 Those with hyperparasitaemia (including non-immune travellers who commonly experience high parasite counts) are associated with more severe episodes of PADH.2 6–9 Multiple case reports describe haemolysis secondary to both intravenous artesunate and oral combination therapies such as artemether–lumefantrine.10–16 It is estimated that up to 22% of patients treated with intravenous artesunate can experience PADH.7 Following the case reports of PADH in returning travellers, in 2013, the Medicines Malaria Venture and WHO published statements acknowledging the need both for physician awareness of PADH and for continued monitoring of patients’ haemoglobin (Hb) levels, especially those with severe malaria treated with intravenous artesunate.5 17 The success of blood transfusion in the management of the most severe PADH cases has meant that no fatal outcome had been reported to date.7 However, in the clinical context of either patient refusal of blood products, or in regions where safe blood products are not readily available, PADH can become a serious patient safety issue.7
We report the fatal case of a 60-year-old woman who suffered from haemolysis 2 weeks following treatment of falciparum malaria in the Congo with oral artesunate/lumefantrine.
Case presentation
We report the case of a 60-year-old postmenopausal woman (medical history: achalasia, migraine and hypertension) who presented to hospital after suffering two collapses at home. One month prior, she reported being treated for malaria during a visit to Congo, following presentation to hospital with recurrent fevers. She completed a course of oral artemether 80 mg/lumefantrine 480 mg and her symptoms initially resolved. Soon after completing treatment and returning to the UK, she started to feel ‘not right’ and lethargic. She recalled a period of passing dark urine 1–2 weeks prior to this presentation. She denied any lower urinary tract symptoms, vaginal and gastrointestinal bleeding.
Physical examination was generally normal bar an ejection systolic murmur heard loudest in the aortic area, in keeping with symptomatic anaemia. Her digital rectal examination revealed no acute bleeding. On admission, besides being tachycardic at 118 bpm, she was otherwise haemodynamically stable.
Investigation
On presentation, she was noted to have severe normocytic anaemia (Hb 43 g/L, last blood tests 1 year prior demonstrated Hb 130 g/L) and a stage 1 acute kidney injury (AKI). Her admission and subsequent blood results are shown on figures 1 and 2. Her iron level was 41.1 µmol/L, transferrin 1.86 g/L, transferrin saturation 88%, ferritin 4252 µg/L, vitamin B12 733 µg/L ad folate 7.8 µg/L. Further blood tests revealed a positive direct antiglobulin/Coomb’s test (DAT), reticulocyte count 34.4%, bilirubin 73 umol/L, lactate dehydrogenase (LDH) 925 U/L and haptoglobin <0.08 g/L. The patient’s malarial and glucose-6-phosphate dehydrogenase deficiency tests were negative. An ultrasound of the abdomen revealed grossly normal liver and spleen. A CT head confirmed no acute intracranial pathology or bleed following her collapses. Blood cultures did not demonstrate any bacterial growth; however rising inflammatory markers and an isolated temperature recording of 39.5C supported an ongoing inflammatory process.
Trends in haemoglobin, mean cell volume and reticulocyte count throughout admission.
Trends in white cell count and C reactive protein throughout admission.
Differential diagnosis
The laboratory test results and history suggestive of presumed haemoglobinuria in this case are in keeping with the diagnosis of haemolytic anaemia, with a >50% drop in Hb compared with a normal Hb the previous year. Alongside this, the time since artemisinin treatment (approximately 20 days) and lack of evidence of any other cause of acute or long-standing haemolysis is suggestive of a diagnosis of PADH.
Treatment
Given her life-threatening anaemia, an extensive discussion was held with the patient about transfusion. Owing to her personal beliefs, the patient made a capacitous decision to refuse transfusion with packed red cells and other blood products, which was respected throughout her stay in hospital. Meanwhile, haematology opinion was sought and supportive care with erythropoietin, folic acid and vitamin B12 supplementation was initiated. Intravenous fluids were strictly restricted in view of high risk of haemodilution and subsequent cardiac risks. Intravenous antibiotics to cover for infection of unknown source were initiated. Unfortunately, the patient continued to deteriorate and methylprednisolone was trialled.
Outcome and follow-up
Despite the above treatments, the patient’s Hb level (figure 1) continued to decline with further episodes of haemoglobinuria. There was a significant worsening of her kidney function and white cell count (figure 2). Her level of consciousness decreased with a capillary sugar of 1.1. Emergency intravenous glucose replacement was initiated; however, a venous blood gas revealed a pH of 6.982, base excess of −25.0 mmol/L (bicarbonate 5.3 mmol/L), lactate >25 mmol/L and unrecordable Hb. Given these biochemical parameters, it was felt that this was not survivable even with intensive care support. Shortly after, the patient regrettably deceased.
Discussion
We describe a unique and fatal case of post-artemisinin therapy haemolytic anaemia in a patient who would not accept blood products. PADH has been defined as a>10% fall in Hb along with an increase in LDH concentration between day 7 and day 28 after initiation of artemisinin therapy.8 18 19 In this case, those conditions and time frame are met and the presence of positive DAT may suggest an element of immune-mediated response. One study reports that 73% of patients who suffered PADH following artemisinin derivates required blood transfusions.7 As this case reveals, the challenge this poses to the management of patients who may not accept blood products or who may not have access to safe blood products is significant.
The two multicentre prospective clinical trials that showed the survival benefit of artesunate compared with quinine in fact suggested that artemisinin derivative medications were well tolerated and had no significant adverse effects.3 4 Consequently, artemisinin derivatives have become the main stay of falciparum malaria management globally.2 However, as cases of delayed haemolysis began to be reported in returning travellers, studies have since shown it to be a predictable event that could be attributed to their effect known as ‘pitting’ (where erythrocytes expel parasite rings then reseal rather than lyse).8 The erythrocytes re-enter the circulation, parasite free, but with a reduced lifespan which leads to a delayed clearance of once infected red blood cells.8 20
It is debated as to whether there is also a mechanism of immune-mediated haemolysis in PADH. The literature includes cases with both positive and negative DAT.14 21–24 One with a positive DAT reports the successful management of PADH with steroid therapy and no blood products, strongly suggesting an autoimmune component to the haemolysis.14 In 2020, the the US Food and Drug Administration approved intravenous artesunate but recommended treatment of PADH with steroids in DAT positive cases.25 In our case, a positive DAT suggests commencing of steroid therapy was appropriate. However, decisions regarding the commencement of steroids were delayed by the need to also manage the ongoing septic process, indicated by the biochemistry, for which we had no source. The clinical decision making involved multidisciplinary approach and input both from the parent acute medical team and the haematology team as to balancing these risks.
The element of renal failure evidenced by the rising creatinine has also been documented in previous case reports.11 26 The mechanisms by which the haemolytic response can cause renal damage are varied. With the Hb recorded as low as 2.6 g/L in this case, it could be said that renal damage was inevitable due to the likely lack of oxygenation to the renal tissue in combination with multifactorial ‘haem pigment AKI’ reported in the literature.26
Most protocols that exist around caring for patients who refuse blood products are designed to optimise the patient in anticipation of a surgical procedure and prevent blood loss or cell salvage during the procedure. Intravenous iron replacement, folic acid, erythropoiesis stimulating agents and B12 supplementation were all administered in line with our trust protocol to offer the best chance for a physiological response to the ongoing haemolysis.27 Blood substitutes remain elusive and while glutaraldehyde-polymerised bovine Hb can raise the oxygen-carrying capacity, it is currently available only on a compassionate basis or as part of a clinical trial.27–29 In this case, the ongoing haemolytic process that led to rapid deterioration unfortunately meant attempts to optimise the patient had little time to bring about benefit and left little time for ‘compassionate’ options.
The General Medical Council has clarified that clinicians should respect a capacitous decision to refuse a treatment.30–33 This, however, should not limit the clinician from advising the patient based on their expert opinion, without judgement or coercion. In any given religion there are varying divergent beliefs and individuals can vary in their adherence to certain regulations.34 As such, addressing variability in what treatments/investigations patients are willing to accept (some patients will accept cryoprecipitate, albumin, etc) is important to treat them effectively.35 The UK Blood Transfusion and Tissue Transplantation Services Professional Advisory Committee suggests that clinicians must adhere to an advanced directive expressed competently.35 The only caveat is in severely ill patients lacking capacity where life-saving treatment should be administered unless there is clear evidentiary support of prior refusal.30 35
The key limitations of this case report first surround the lack of information about the presentation and management of this patients’ malaria in the Congo, and hence timeline is non-specific. We have not been able to determine the initial parasite count. However, we can confirm the absence of anaemia 1 year prior to this fatal episode and hence can infer that a chronic autoimmune haemolytic process is unlikely. A second limitation was that the number of tests we were able to run was limited. To preserve her dropping Hb level, repeated venepuncture was avoided for anything other than tests that would immediately alter disease management.
This case should naturally prompt rediscussion of alternative treatments of falciparum malaria to artemisinin derivatives. WHO, Center for Disease Control and Prevention and National Institute for Health and Care Excellence guidelines emphasise artemisinin combination therapies (ACTs) are the mainstay of uncomplicated falciparum malaria treatment, and intravenous artesunate is first line for complicated and severe falciparum malaria.2 36 37 The reasoning behind ACTs superiority is that artemisinin can rapidly reduce parasitaemia, the partner drug can clear residual parasitaemia, and the combination can help thwart the growing threat of resistance to artemisinin.38–40 Alternatives include atovaquone with proguanil hydrochloride, quinine with doxycycline/clindamycin and less commonly chloroquine and mefloquine.36 37 For a combination of reasons—including side effect profiles, efficacy and resistance development—ACTs are favoured.36–39 41 42 In cases of non-immune individuals travelling to malaria-endemic areas with poor access to medical care, ACTs can be issued as ‘standby emergency treatment for malaria’.43 As non-immune travellers are at risk of hyperparasitaemia, and hence PADH, this case demonstrates the possible severe consequences of ACT use in this particular patient group.9 44 Yet, this risk must be weighed against the known benefits of artemisinin derivates in rapidly clearing high parasite loads.
So far, the focus for clinical management of PADH has been through monitoring and follow-up.25 45 The severity of anaemia documented in this case suggests that monitoring should not only be encouraged for cases of severe malaria managed with intravenous artesunate, but also in those who receive oral artemether–lumefantrine, as this was not the first case report of PADH following oral therapy.46 Research should seek to explore whether this is truly an isolated case, or whether it is possible that in settings where access to blood products is limited—for example, areas that also carry high risks of malaria—or in cases such as this where patients do not accept blood products, PADH can in fact be fatal. The nature of artemisinin derivatives’ effectiveness through the mechanism of pitting is likely to be the very reason it poses a risk of PADH.8 In patient groups for whom blood products are not an option—by choice or by availability—the decision to use artemisinin’s may not be straight-forward and requires further discussion given this report.
Learning points
-
Globally, healthcare professionals should counsel patients they are treating with artemisinin therapies (including oral) on the risk and signs of postartemisinin delayed haemolysis to encourage patients to seek medical attention at earliest opportunities, even if lost to follow-up.
-
The safety of both oral and intravenous artesunate in patients who refuse blood products (notably those with hyperparasitaemia), or are in settings where safe blood products are not readily available, should be carefully weighed against the benefits of the medications’ antimalarial properties.
-
Anaemia in malaria patients who refuse blood products should be anticipated, so that optimisation and monitoring can begin early.
-
Clinicians should consider early use of steroid therapy in managing drug-induced haemolytic anaemia as this may prevent requirement for blood products.
Ethics statements
Patient consent for publication
Acknowledgments
We would like to thank the patient and their family for allowing us the opportunity to submit this case. The family especially have been very gracious and understanding.
Footnotes
-
Contributors LG and SJ drafted the initial paper. YZ edited the paper and created the figures. KPT finalised the paper. All authors agreed on the final draft.
-
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
-
Competing interests None declared.
-
Provenance and peer review Not commissioned; externally peer reviewed.
- © BMJ Publishing Group Limited 2021. No commercial re-use. See rights and permissions. Published by BMJ.
References
Use of this content is subject to our disclaimer