SARS‐CoV‐2 vaccination induces breakthrough hemolysis in paroxysmal nocturnal hemoglobinuria on complement inhibitor

To the Editor:

The SARS‐CoV‐2 infection and vaccination have raised concerns in complement mediated hemolytic anemias (i.e., paroxysmal nocturnal hemoglobinuria, PNH, and cold agglutinin disease, CAD), particularly if on treatment with complement inhibitors. ¹ , ² , ³ Among complement amplifying triggers (infections, traumas, and surgery ⁴ , ⁵ ) an emerging causative agent may be represented by SARS‐CoV‐2 vaccines. In PNH patients on complement inhibitors, pharmacodynamic breakthrough hemolysis (BTH) has been defined as hemolysis reactivation with hemoglobin drop due to increased complement activity. ⁶ In CAD, where complement inhibitors are in clinical trials, a true BTH has not been defined yet, although hemolytic flares have been described under treatment. ⁷ Gerber et al. ⁵ recently reported six PNH patients, either untreated or on complement inhibitors, who experienced abrupt hemolytic crisis following SARS‐CoV‐2 vaccination (four Pfizer‐BioNTech and two Moderna vaccine). Five patients were on intravenous ravulizumab and developed BTH after either the first (N = 1), the second (N = 2), or after both doses (N = 2) of vaccine. Two patients were also receiving oral factor D inhibitor danicopan, and one experienced BTH only after the second dose, after interruption of danicopan (two missed doses) due to concerns about interaction with vaccine efficacy. Pérez‐Lamas et al. described a hemolytic crisis occurring 2 days after the first dose of an mRNA Covid‐19 vaccination in a CAD patient not under complement inhibition. ⁸ Here we describe a classic hemolytic PNH patient on C5 inhibitor, who developed severe BTH one day after the second dose of Moderna SARS‐CoV‐2 vaccine. This patient belongs to a cohort of 16 patients (13 PNH and three CAD) who received mRNA SARS‐CoV‐2 vaccine from March 2021. All of them were actively treated with complement inhibitors (five eculizumab, three intravenous ravulizumab, four subcutaneous ravulizumab, one factor B inhibitor iptacopan, and three C1s inhibitor sutimlimab) at our Hematology unit in Milan, Italy. The study was conducted in accordance with Helsinki Declaration and patients gave informed consent.

A 45‐year‐old lady suffering from PNH presented to the emergency room on April 21 the day after receiving the second dose of Moderna mRNA SARS‐CoV‐2 vaccine. The patient had a diagnosis of classic hemolytic PNH since 1996, and was on C5 complement inhibition from 2016 with suboptimal response and complete control of intravascular hemolysis (average Hb 9.5 g/dL with normal LDH, inadequate reticulocytosis). Since 2019 she has been enrolled in a clinical trial with subcutaneous ravulizumab, administered weekly. The patient had had no side effects following her first dose of the Moderna vaccine (March, 24). At presentation, she complained of hyperpyrexia (38.8°C) and several episodes of vomiting, abdominal pain and dark urine. Vital signs were normal and blood counts (Figure S1) displayed severe anemia with significant intravascular hemolysis (LDH 2.3 × ULN), consistent with BTH. Moderate neutropenia (0.83 × 10⁹/L), and mild thrombocytopenia (100 × 10⁹/L) were also noted, along with slightly increased C reactive protein (3.6 mg/dL), prolonged prothrombin time ratio (1.46), and increased D dimer (895 g/L). Molecular testing on nasopharyngeal swab for SARS‐CoV‐2 was negative, as well as chest X‐ray, blood and urine cultures. The patient was started on intravenous antibiotics, hydration, and paracetamol with progressive amelioration of symptoms, and admitted to the hematology ward. BTH persisted, and low molecular weight heparin was started. On day+4 from presentation, Hb dropped to 6.7 g/dL with suboptimal reticulocyte counts (106 × 10⁹/L). Recombinant human erythropoietin (rhEPO, epoetin alpha 40 000 IU subcutaneously) was administered with hematological improvement and repeated on day+14. Of note, subcutaneous ravulizumab was regularly continued along admission.

The occurrence of post‐vaccine BTH in PNH patients should not discourage vaccination of this population. In fact, whilst Covid‐19 may be fatal and difficult to handle, ¹ , ² BTH is more known and seems manageable with supportive care, as in the case reported. In our patient, neutropenia deserved hospital admission, and recombinant erythropoietin was helpful to improve anemia avoiding transfusions. Moreover, the prompt recognition of BTH is pivotal, since, besides anemia, active intravascular hemolysis puts patients at higher risk of thrombosis, through various mechanisms including free Hb release and nitric oxide depletion with consequent endothelial dysfunction. ⁶ Of note, increased thrombotic risk is also reported in CAD, although its management has not been clearly established. Thrombotic risk may be even higher during Covid‐19, and after certain SARS‐CoV‐2 vaccines that may mimic autoimmune heparin‐induced thrombocytopenia. ⁹ In our patient, prophylaxis with LMWH was therefore instituted and no thrombosis occurred.

Regarding the mechanisms of post‐vaccine BTH, it has been described that SARS‐CoV‐2 spike protein may bind on nucleated cells and amplify complement alternative pathway by interfering with factor H. ¹⁰ However, the addition of spike protein subunit 1 did not increase lysis of PNH erythrocytes ex vivo, ⁵ and did not appear to bind red cells, weakening the hypothesis of a direct vaccine effect on erythrocytes. On the contrary, complement amplification triggered by the inflammatory/immune response following vaccine seems prominent. In our patient, no BTH was reported after the first vaccine administration, suggesting a pivotal role of higher immune response following the “booster” dose. In keeping with this hypothesis, pharmacodynamic BTH has been observed even after ravulizumab that has a half‐life four times longer than eculizumab, and is associated with reduced pharmacokinetic BTH events. ¹¹ On the other hand, a protective effect of proximal complement inhibitors (i.e. C1 inhibitors, C3 inhibitors, factor D and factor B inhibitors) may be hypothesized, although definite data are not available. Consistently, no hemolytic events occurred after either vaccine doses in the three CAD patients on the C1s inhibitor sutimlimab at our center.

Concerning timing of administration, in the experience by Gerber et al., ⁵ all patients had received ravulizumab ≥4 weeks before vaccine, suggesting that administration of SARS‐CoV‐2 vaccination may be safer within 4 weeks from the last ravulizumab. However, our patient had received weekly subcutaneous ravulizumab the day immediately before both vaccine doses. It is therefore difficult to advise a proper timing of vaccine administrations, and close monitoring and patient education remain the only mandatory procedures.

In conclusion, this report shows that hemolytic flares may occur after SARS‐CoV‐2 vaccines and should be promptly recognized and managed through medical suspicion, patient education, and clinical monitoring.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

AUTHOR CONTRIBUTIONS

All authors followed patients, wrote the article and revised it for important intellectual content.

Supporting information

Figure S1. Laboratory values upon the first and second dose of SARS‐CoV‐2 vaccination. Hb, Hemoglobin; LDH, lactate dehydrogenase; LMWH, low molecular weight heparin; Ret, reticulocytes; rhEPO, recombinant human erythropoietin.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.