Breakthrough haemolysis in paroxysmal nocturnal haemoglobinuria after COVID-19 infection and COVID vaccination: what is worse?

INTRODUCTION

Over the last year, an increasing number of reports have highlighted the crucial role of complement activation in the pathogenesis and severity of COVID-19, the disease caused by the SARS-CoV-2 virus. The latter is part of the immunological hyper-response to the virus, and may contribute to endothelial damage, blood clotting, and cytopenia. Trials with complement inhibitors in severe COVID-19 are currently ongoing with promising results¹. Paroxysmal nocturnal haemoglobinuria (PNH) is a rare acquired haematopoietic stem cell disease caused by a somatic mutation of the PIGA gene resulting in an increased susceptibility of PNH cells to complement damage. Clinically, PNH is marked by intravascular haemolysis, increased thromboembolic risk, and bone marrow failure, and benefits from treatment with complement inhibitors. PNH patients have a higher risk of infection due to therapy with complement inhibitors and, in many cases, to the presence of concomitant bone marrow failure. Moreover, infections are well known triggers of haemolysis, together with drugs, trauma or surgical interventions. Breakthrough haemolysis (BTH) has been defined as a reactivation of haemolysis in patients treated with complement inhibitors. There is growing evidence to suggest that both COVID-19 and COVID vaccination may precipitate haemolysis in therapy-naïve PNH patients and may trigger BTH in those on anti-complement therapy. Here we describe the case of a PNH patient on therapy with the C5 inhibitor eculizumab, who developed BTH following both SARS-CoV-2 infection and COVID vaccination.

CASE REPORT

A 27-year-old man was diagnosed with classical haemolytic PNH in 2010 (93% granulocyte PNH clone ). Due to transfusion-dependent anaemia and abdominal pain, he was started on eculizumab 900 mg every 14 days, achieving good control of intravascular haemolysis. Over the following years, the patient experienced a total of 5 severe BTH episodes with transfusion requirement, usually triggered by an infection. The last episode occurred in October 2020 after cholecystectomy due to gallstones. At the beginning of March 2021, 12 days after the last eculizumab infusion, a routine SARS-CoV-2 nasopharyngeal swab performed to allow in-hospital administration of therapy resulted positive and the patient was quarantined. Two days later, he presented to the emergency room complaining of asthenia and dark urine. Blood tests were consistent with severe BTH (haemoglobin 6.1 g/dL, lactate dehydrogenase (LDH) 9 times the upper limit of normal [ULN]), and computed tomography scan showed interstitial pneumonia involving approximately 10% of lung parenchyma. The patient was hospitalised and started on intravenous antibiotics, steroids, and low molecular weight heparin (LMWH). Furthermore, he received the scheduled eculizumab dose (with a delay of one day) and was transfused with 3 units of red blood cells (RBC) after which his clinical condition progressively stabilised. The patient was then discharged and continued on fortnightly eculizumab.

In July 2021, 2 days after eculizumab administration, he received the first dose of the mRNA Moderna COVID vaccine and the following day experienced a single episode of fever, accompanied by asthenia, nausea and loss of appetite. Nine days later, blood tests were performed revealing moderate anaemia with a change in haemolytic markers (haemoglobin 8.3 g/dL, LDH 2 times the ULN, indirect bilirubin 3.5 mg/dL). Therefore, he was transfused with 2 units of RBC and the next eculizumab dose was anticipated by 2 days. The clinical picture promptly improved and returned to normal within a few days. The patient received the second dose of the Moderna vaccine in January 2022 experiencing only a mild (10%) decrease in Hb and no other signs of BTH.

DISCUSSION

In the COVID-19 era, the virus itself has been recognised as an additional trigger of haemolytic flares/BTH, and it is essential that patients are educated to recognise early symptoms and to refer possible contagions in order to allow proper surveillance. Haemolysis in this context may be due to a direct virus-mediated complement activation² that overwhelms the effect of ongoing complement inhibition, as also reported for other infectious agents.

In fact, our patient had experienced several para-infectious BTH, not dissimilar to that occurring after COVID-19. Other proposed mechanisms for the exacerbation of haemolytic disorders include molecular mimicry for autoimmune haemolytic anaemia, given the high homology among Spike protein and erythrocyte antigens³, and complement deposition on erythrocyte surface following the exposure of phosphatidyl serine during peroxidative stress typical of severe COVID-19⁴. Eight reports of 19 PNH patients contracting COVID-19 infection have been published, 13 of whom experienced haemolytic flares (68%), and 7 who experienced BTH (36%) (Table I). Despite the limited numbers reported, BTH seemed to occur equally in patients on eculizumab or ravulizumab. Patients were mainly handled with supportive therapy and transfusions; only 2 required admission to an intensive care unit and intubation, and one died. The latter was a 43-year-old man comorbid with type 2 diabetes mellitus, who developed respiratory failure requiring mechanical ventilation. In our case, the patient experienced a moderate COVID-19 infection but a severe BTH requiring transfusions. Importantly, LMWH prophylaxis was administered, given the known thrombotic risk of BTH and that associated with COVID-19 itself.

Table I.

COVID-19 in paroxymal nocturnal haemoglobinuria (PNH) patients

StudySupplementary references*	N. of patients, therapy for PNH	Severity of SARS-CoV-2 infection	Haemolysis type	Outcome
Kulasekararaj et al.s1	4 patients
	• Pt 1, ravulizumab	• Mild	• None	• Alive
	• Pt 2, eculizumab	• Mild	• BTH	• Alive
	• Pt 3, therapy naive	• Moderate	• Intravascular haemolysis	• Alive
	• Pt 4, therapy naive	• Moderate	• Intravascular haemolysis	• Alive
Pike et al.s2	4 patients
	• Pt 1, ravulizumab	• Severe	• BTH	• Alive
	• Pt 2, eculizumab	• Mild	• None	• Alive
	• Pt 3, eculizumab	• Mild	• BTH	• Dead
	• Pt 4, eculizumab	• Critical illness	• BTH	• Alive
Araten et al.s3	3 patients
	• Pt 1, ravulizumab	• Mild	• None	• Alive
	• Pt 2, eculizumab	• Moderate	• None	• Alive
	• Pt 3, eculizumab	• Milde	• None	• Alive
Genthon et al.s4	• 1 patient, eculizumab	• Critical illness	• BTH	• Alive
Schüller et al.s5	• 1 patient, eculizumab	• Mild	• BTH	• Alive
Otieno et al.s6	• 1 patient, therapy-naive	• Mild	• Intravascular hemolysis	• Alive
Barcellini et al.s7	4 patients
	• Pt 1, eculizumab	• Mild	• None	• Alive
	• Pt 2, eculizumab	• Mild	• None	• Alive
	• Pt 3, ravulizumab	• Mild	• BTH	• Alive
	• Pt 4, therapy-naive	• Mild	• None	• Alive
Sokol et al.s8	• 1 patient, eculizumab	• Moderate	• BTH	• Alive

^*See Online Supplementary Content. Severity of COVID-19 according to NIH guidelines⁵.

BTH: breakthrough haemolysis; pt: patient.

The impact of COVID vaccines on PNH is unusual. A total of 94 patients have been reported, of whom 7 experienced significant haemolytic flares (6 BTH), 10 with only a mild decrease in Hb levels, and one developing a microvascular thrombosis after mRNA vaccines (Table II). Most cases were mild, were manageable with supportive measures, and recovered spontaneously. Complement activation driven by the immune response to vaccine seems the prominent trigger. In this regard, Gerber et al. showed that the addition of Spike protein subunit 1 did not increase lysis of PNH erythrocytes in vitro and did not appear to bind RBC, weakening the hypothesis of a direct vaccine effect on erythrocyte damage⁶.

Table II.

Anti-SARS-CoV-2 vaccine and paroxymal nocturanl haemoglobinuria (PNH) patients

StudySupplementary references*	N. of patients, therapy for PNH	atype	Outcome
Gerber et al.s9	5 patients
	• pt 1, therapy-naive	• Microvascular bowel thrombosis
	• pt 2, ravulizumab	• BTH
	• pt 3, ravulizumab + danicopan	• BTH	• Alive
	• pt 4, ravulizumab	• No
	• pt 5, ravulizumab	• BTH
Fattizzo et al.s10	• 1 patient ravulizumab	• BTH	• Alive
Portuguese et al.s11	• 1 patient, therapy-naive	• Intravascular haemolysis	• Alive
Giannotta et al.s12	87 patients		• Alive
	• 70/87 on complement inhibition therapy	• 12/87 experiencing BTH
	• 46/87 eculizumab	• 3/87 patients displaying clinical BTH
	• 15/87 ravulizumab	• pt 1, ravulizumab: BTH
	• 3/87 crovalimab	• pt 2, eculizumab: BTH
	• 3 of whom in combination with oral anti-factor D inhibitor	• pt 3, eculizumab: BTH
	• 3/87 on anti-factor B single agent

^*See Online Supplementary Content.

BTH: breakthrough haemolysis; pt: patient.

The question arises as to which is “worse”: COVID-19 infection itself or COVID vaccination? In our patient, the BTH after vaccination was milder than that after COVID-19 infection. Moreover, our review of the literature seems to confirm that management of haemolytic flares after vaccination is not problematic, further supporting the view that the advantages of vaccination outweigh the risk of SARS-CoV-2 infection. Finally, vaccines, in contrast to infections, are predictable triggers, and may, therefore, be easily monitored in order to intercept BTH.

In conclusion, in PNH patients, haemolytic flares may be triggered both by COVID-19 and by COVID vaccination, the latter apparently causing a milder clinical picture. Indeed, last year, several new variants of SARS-CoV-2 emerged with different morbidity and lethality as compared to the original and delta variants. Moreover, the vaccination campaign has probably had an effect on the severity of COVID-19 in recent months. Finally, it is not known if patients who have already encountered the virus may be at a higher or lower risk for haemolytic crisis after vaccination. Although it is difficult to clearly estimate the risk of vaccine-related haemolysis in this rapidly evolving scenario, PNH patients should be encouraged to get vaccinated and educated to recognise symptoms of haemolysis, in order to allow timely management.