COVID-19 vaccination prevents a more severe course and treatment with complement inhibitors reduce worsening hemolysis during the Omicron pandemic in patients with PNH: a single-center study

Abstract

Objective

Paroxysmal nocturnal hemoglobinuria (PNH) is a rare acquired disease characterized by chronic complement-mediated hemolysis. The concentrated outbreak of coronavirus disease 2019 (COVID-19) in China after 6 December 2022, provided an opportunity to observe the disease course of PNH during an active Omicron infection epidemic.

Patients and method

Patients diagnosed with PNH at Peking Union Medical College Hospital (PUMCH) before 6 December 2022, were followed up until 10 April 2023. Clinical data related to coronavirus infection and hemolysis were recorded. Factors influencing the infection and severity rate of Omicron, as well as hemolysis provocation, were analyzed.

Results

In total, 131 patients with PNH were included in this retrospective analysis; 87.8% were infected with Omicron. Among them, 15.7% met the criteria for severity, and 1 patient died (0.87%). No protective factors were identified against Omicron infections. However, patients with severe Omicron infection (n = 18) had a lower vaccination rate than those with non-severe infection (n = 97; p = 0.015). Among those infected (n = 115) with Omicron, there was a significant increase in lactate dehydrogenase (LDH) levels compared with those in the uninfected group (n = 16, p = 0.000). Patients with severe infections (n = 18) had even higher LDH increase rates than those without severe infections (n = 97; p = 0.002). 10 (37.0%) patients treated with complement inhibitors developed breakthrough hemolysis (BTH). Patients treated with complement inhibitors (n = 27) exhibited less severe hemolysis than treatment-naïve patients (n = 104; p = 0.003).

Conclusions

Omicron infection exacerbates hemolytic attacks in patients with PNH. Vaccination helps mitigate the severity of Omicron infection, and using complement inhibitors reduces hemolysis exacerbation.

Keywords:

• Paroxysmal nocturnal hemoglobinuria
• omicron infection
• Covid-19
• hemolysis aggravation
• complement inhibitor
• vaccination

Introduction

Paroxysmal nocturnal hemoglobinuria (PNH) is a clonal hematopoietic stem cell disease in which patients develop somatic phosphatidylinositol glycan (PIG-A) gene mutations in hematopoietic stem cells, resulting in impaired biosynthesis of glycosylphosphatidylinositol (GPI) anchors. This results in the deletion of proteins attached to the cell surface by GPI anchors. The absence of the two important complement regulatory proteins, CD55 and CD59, can lead to the destruction of red blood cells following complement activation, resulting in hemolysis [1–4]. Thus, PNH hemolysis is driven by complement activation. The clinical manifestations of PNH include intravascular hemolysis, bone marrow failure, and thrombosis [5]. In 2007, the United States Food and Drug Administration (US FDA) approved the first-generation complement inhibitor eculizumab, which greatly improved the related symptoms caused by intravascular hemolysis, prolonged the survival of patients, and enhanced their quality of life [5]. In recent years, newer generation complement inhibitors, such as crovalimab and iptacopan, have been used in premarket clinical studies in China, demonstrating promising efficacy [6–8].

Since December 2019, the outbreak of the novel coronavirus pneumonia (COVID-19) has been recognized by the World Health Organization as a major emergency public health event owing to its strong pathogenicity and transmissibility, causing a global impact on human health and mortality [9]. China initiated a large-scale COVID-19 vaccination campaign in 2021, followed by sporadic vaccinations. Patients with PNH are more susceptible to hemolytic attacks under stress, infections, and other complement-activating factors. In China, there was a concentration of omicron infections from December 2022 to April 2023.

Our center routinely followed up many patients with PNH in the outpatient clinic. Some patients received COVID-19 vaccinations before December 6. In addition, some patients participated in several complement inhibitor clinical trials that began in October 2021 and continued until the time of this paper’s submission. To investigate the infection rate, severity of Omicron disease, and hemolysis-related symptoms following infection, we conducted a retrospective analysis of patients with PNH during this specific period. Patients willing to provide medical records were included in the analysis.

Patients and methods

Data collection

Data were collected from successive patients diagnosed with PNH before 6 December 2022 who were regularly followed up at Peking Union Medical College Hospital (PUMCH) until 10 April 2023. During that period, 150 patients were followed up and 131 patients were included in this study. Patients infected with pathogens other than Omicron were excluded. Demographic information, medical history, physical examination, hemoglobin, LDH, hepatic and renal function, ferritin, PNH clone size (calculated by neutrophil FLAER-negative cell ratio), disease complications, COVID-19 vaccination status, and previous treatment were documented.

The above information and the Omicron-related manifestations during the pandemic were also collected. Hemolysis tests for infected patients were conducted during active Omicron infections, usually within 2 weeks after the diagnosis of Omicron infection. Hemolysis tests for uninfected patients were performed during the same Omicron pandemic period as part of regular follow-up. Data within 1 month before 6 December 2022, were recognized as the baseline data.

Most of the above information was obtained from the medical records of our hospital, some patients’ laboratory examinations came from local hospitals, and follow-up information was obtained from medical records or telephone follow-ups. All patients signed an informed consent form before providing their medical history. This study was approved by the Ethics Committee of the Peking Union Medical College Hospital.

Criteria for the diagnosis and classification of PNH

PNH diagnosis was based on clinical manifestations combined with the absence of GPI-anchored proteins (GPI-Aps) on ≥2 lineages [1,5,10,11]. PNH was further classified as PNH with or without concomitant bone marrow failure (BMF). PNH clone size was determined using the proportion of FLAER-negative neutrophils.

The criteria of diagnosis and efficacy

Diagnostic criteria for COVID-19

Patients were required to meet the following criteria for the diagnosis of COVID-19: (1) Present COVID-19 infection-related clinical manifestations; (2) Have one or more of the following etiological and serological test results: (a) positive nucleic acid test for novel coronavirus, (b) test positive for COVID-19 antigen, (c) positive isolation or culture of the novel coronavirus, and (d) the level of novel coronavirus-specific IgG antibody increased four times or more in the convalescent phase than in the acute phase [12,13].

Criteria for severe COVID-19 (severe and/or critical)

Severe cases met any of the following criteria:

• Oxygen saturation <90% on room air

• Severe pneumonia

• Signs of severe respiratory distress (adult use of auxiliary muscles, inability to speak complete sentences, and respiratory rate >30 times/min [12].

Critical case

Critical cases involve acute respiratory distress syndrome (ARDS), sepsis, septic shock, or other conditions usually requiring life-sustaining treatment, such as invasive or noninvasive mechanical ventilation or pressors.

Definition of hemolysis-related manifestations

Hemolysis aggravation: worsening hemolysis and/or new thrombotic events. Worsening hemolysis: an aggravation of hemolysis-related manifestations (hemoglobinuria, jaundice), an increase of >200 U/L in LDH, a decrease of 20 g/L in hemoglobin (HGB) from baseline, or a change from non-transfusion-dependent to transfusion-dependent without other causes.

Thrombotic events are diagnosed by imaging [14].

Definition of breakthrough hemolysis

Since no consensus was reached for the definition of breakthrough hemolysis in patients with PNH, we adopted the criteria from previous reports [15–17]: at least one new or worsening symptom or sign of intravascular hemolysis (fatigue, hemoglobinuria, abdominal pain, shortness of breath [dyspnea], anemia [hemoglobin < 10 g/dL], MAVE [including thrombosis], dysphagia, or erectile dysfunction) in the presence of elevated LDH ≥2 × ULN after prior LDH reduction to <1.5× ULN on treatment.

Statistical analysis

Data analysis was conducted using the SPSS software. The chi-square test was used for categorical variables, and the rank-sum test was used for rank data. An independent samples t-test was used for continuous variables. The chi-square test was used to screen for influencing factors, followed by a binary logistic regression analysis. Statistical significance was set at p < 0.05.

Results

Baseline (data before 6 December 2022) characteristics of patients with PNH

The final analysis included 131 patients, including 66 males (50.4%). The median patient age was 42 years (range 17–81 years). The median time from the diagnosis of PNH to 6 December 2022, was 72 months (range 12–504 months). The final follow-up date was 10 April 2023. The median follow-up duration was 3 months (range 2–4 months).

Before 6 December 2022, the median frequency of hemoglobinuria was two episodes per year (range 0–10). Further, 7 patients (5.3%) had persistent hemoglobinuria and 39 (29.8%) had a history of thrombosis; 20 cases (51.2%) of thrombus occurred in the abdominal vein. Other cases involved deep vein thrombosis (DVT), intracranial venous sinus thrombosis, myocardial infarction, and cerebral infarction. A total of 24 patients (61.5%) received anticoagulant therapy, such as warfarin, low-molecular-weight heparin, and rivaroxaban, and 15 (38.5%) did not receive anticoagulants due to contraindications.

Of all the patients in this study, 27 patients were treated with complement inhibitors, including 11 patients (40.7%) treated with crovalimab (a C5 complement inhibitor) for 9 months (range 6–14 months), 10 patients (37%) treated with iptacopan (Factor B complement inhibitor) for 9 months (range 6–14 months) and 6 patients (22.2%) treated with CAN106 (a C5 complement inhibitor) for 7 months (range 6–9 months). A total of 41 patients (31.3%) received COVID-19 vaccinations, and the median time from the last vaccination to 6 December 2022, was 9 months (range 1–18 months). The vaccines included the Beijing Biological Vaccine (45%) and the Kexing Vaccine (55%); both are inactivated vaccines. All patients received two shots. The baseline characteristics of the patients are shown in Table 1.

Table 1. Clinical features of patients with PNH before 6 December 2022.

Clinical features	All (n = 131)	Infected (n = 115)			Not-infected (n = 16)	p^▲
		Severe (n = 18)	Non-severe (n = 97)	p*
Median age (range), yrs	42 (17–81)	45.5 (32–58)	42 (17–75)	0.968	43 (21–81)	0.633
Male ratio (%)	50.4%	50%	50.5%	0.968	43.8%	0.571
PNH type, No. (%)				0.694		0.922
With BMF	64 (48.9%)	8 (44.4%)	48 (49.5%)		8 (50%)
Without BMF	67 (51.1%)	10 (56.6%)	49 (50.5%)		8 (50%)
With thrombosis, No. (%)	39 (29.8%)	7 (38.9%)	24 (24.7%)	0.341	8 (50.0%)	0.110
Laboratory tests
HGB (g/L)	79 (36–132)	77 (36–122)	75 (42–132)	0.501	80 (45–126)	0.546
ANC (×10^9/L)	2.54 (0.49–4.5)	2.39 (0.49–4.0)	2.1 (1.2–4.2)	0.613	2.63 (0.86–4.5)	0.354
PLT (×10^9/L)	146 (64–352)	132 (66–339)	154 (70–281)	0.357	140 (64–352)	0.268
Ret (×10^9/L)	55 (21–89)	54 (22–87)	55 (31–89)	0.441	58 (21–80)	0.551
SGPT (U/L)	17 (4.6–36)	15.5 (5.88–27)	14 (6.1–33)	0.538	17 (4.6–36)	0.746
Tbil (μmol/L)	28 (11.3–111.8)	26.1 (15.5–107)	27 (18.2–111.8)	0.356	29.6 (11.3–106)	0.245
Dbil (μmol/L)	10.1 (4.5–13.1)	9.05 (4.5–12.6)	10.2 (5.2–13.1)	0.437	10.9 (6.2–12.8)	0.631
Cr (μmol/L)	55 (36–139)	56.5 (42–134)	57 (40–128)	0.543	54 (36–139)	0.645
LDH (U/L)	762 (223–2572)	798 (432–2572)	743 (231–2203)	0.291	801 (223–2450)	0.289
SF (μg/L)	106 (20–261)	110 (20–252)	112 (25–243)	0.627	102 (35–261)	0.461
PNH clone size (granulocyte flare)	82 (60–100%)	82 (61–100%)	84 (63–100%)	0.745	83 (60–100%)	0.874
COVID-19 vaccination status				0.015		0.252
Vaccinated, No. (%)	41 (31.3%)	1 (5.6%)	33 (34%)		7 (43.8%)
Unvaccinated, No. (%)	90 (68.7%)	17 (94.4%)	64 (66%)		9 (56.2%)
The median duration from vaccination to 6 December 2022 (range), month	9 (1–18)		12 (1–18)	0.601	9 (3–18)	0.354
Baseline treatment, No. (%)
Complement inhibitor	27 (20.6%)	2 (11.1%)	21 (21.6%)	0.480	4 (25.0%)	0.894
glucocorticoid	15 (11.4%)	3 (16.7%)	12 (12.4%)	0.908	0 (0%)	0.264
CsA	21 (16%)	4 (22.2%)	14 (14.4%)	0.630	3 (18.8%)	1.000
androgen	37 (28.2%)	5 (27.8%)	26 (26.8%)	1.000	6 (37.5%)	0.561
TPO-RA	3 (2.3%)	1 (5.6%)	2 (2.1%)	0.403	0 (0%)	1.000
blood transfusions	23 (17.6%)	4 (22.2%)	17 (17.5%)	0.887	2 (12.5%)	0.828
untreated	17 (13%)	4 (22.2%)	12 (12.4%)	0.460	1 (6.3%)	0.647

PNH: paroxysmal nocturnal hemoglobinuria; BMF; bone marrow failure; CsA: cyclosporine A; TPO-RA: thrombopoietin receptor agonist; HGB: hemoglobin; ANC: absolute neutrophil count; PLT: platelet; Ret: reticulocyte; SGPT: serum glutamic-pyruvic transaminase; Tbil: total bilirubin; Dbil: direct bilirubin; Cr: creatinine; LDH: lactate dehydrogenase; SF: serum ferritin; infected: Omicron infected; severe: severe Omicron infection.

p*: <0.05 indicated a statistically significant difference between severe and non-severe infections.

p^▲: <0.05 indicated a statistically significant difference between infected and uninfected individuals.

A total of 115 patients (87.8%) were newly diagnosed with COVID-19 (Omicron). Of these, 86 (74.8%) infections were confirmed via antigen and nucleic acid tests, while 29 (25.2%) were based on close contact history, clinical manifestations during the peak of Omicron infection, and other positive COVID-19 tests. Diagnoses were made between 6 December 2022 and 5 February 2023. No significant differences were found in clinical characteristics, including COVID-19 vaccination status and use of complement inhibitors, between those infected or with Omicron (Table 1).

COVID-19 related symptoms

All observed Omicron infections occurred within a month after 6 December 2022, and most hemolysis cases occurred within 2 weeks after Omicron infections. A total of 101 (87.8%) patients exhibited fever with a median maximum body temperature of 38.5 °C (range 37.1–40.5) and an average duration of 2 days (range 1–12). Other symptoms included sore throat (n = 46, 40%), cough (n = 75, 65.2%), phlegm (n = 59, 51.3%), muscle and joint pain (n = 45, 39.1%), fatigue (n = 84, 73%), headache (n = 11, 9.6%), nausea and vomiting (n = 7, 6.1%), nasal congestion (n = 4, 3.5%) and dizziness (n = 4, 3.5%), loss of smell and taste (n = 4, 3.5%), diarrhea (n = 3, 2.6%), shortness of breath in chest (n = 3, 2.6%), loss of appetite (n = 2, 1.7%), abdominal pain (n = 1, 0.9%), sweating (n = 1, 0.9%), bleeding (n = 1, 0.9%), toothache (n = 1, 0.9%), chest pain (n = 1, 0.9%), Lumbago (n = 1, 0.9%), insomnia (n = 1, 0.9%), orbital pain (n = 1, 0.9%), and dry eyes (n = 1, 0.9%). The median duration of COVID-19 infection was 8 days (range 3–16).

Severe infections and possible risk factors

Of all the infected patients, 18 (15.7%) were considered severe. Baseline clinical features were compared between patients with and without severe infection. Only patients who had received vaccination exhibited a significant reduction in severe COVID-19 symptoms (p = 0.015, Tables 1 and 2). We further categorized patients who received complement inhibitors as those with proximal or terminal complement inhibitors. Of the 17 patients treated with terminal complement inhibitors (crovalimab and CAN106), 14 (82.4%) were infected with Omicron, and 2 (14.3%) were severe cases. Of the 10 patients treated with a proximal complement inhibitor (iptacopan), 9 (90%, p = 1.000 compared with those with terminal complements) were infected with Omicron, but no serious cases were observed (p = 0.502, Figure 1).

Figure 1. Comparison of proximal and terminal complement inhibitors in protecting BTH. Patients treated with terminal (crovalimab and CAN106) or proximal (iptacopan) complement inhibitors were also compared. No differences were found in infection or severe infection rates. However, higher BTH levels were observed in patients treated with terminal complement inhibitors (p = 0.003). Infection rate: Omicron infection rate; severe infection rate: severe Omicron infection rate; BTH: breakthrough hemolysis. p*: <0.05 indicated a statistically significant difference.

Table 2. Comparison of patients with or without vaccination.

Clinical features	The whole cohort			Patients with Omicron infection
	Vaccinated (n = 41)	Unvaccinated (n = 90)	p Value	Vaccinated (n = 34)	Unvaccinated (n = 81)	p Value
Median age (range), yrs	41 (19–81)	45.5 (17–75)	0.332	39.5 (19–75)	44 (17–75)	0.101
Male ratio (%)	53.7%	48.9%	0.613	58.8%	48.1%	0.254
PNH type, No. (%)	20 (48.8%)	44 (48.9%)	0.991	15 (44.1%)	41 (50.6%)	0.525
With BMF
Without BMF	21 (51.2%)	46 (51.1%)		19 (55.9)	40 (49.4%)
Infected with COVID-19, No. (%)			0.252
with	34 (82.9%)	81 (90%)		—	—
without	7 (17.1%)	9 (10%)		—	—
Severe COVID-19 infection, No. (%)						0.015
with	—	—		1 (2.9%)	17 (21%)
without	—	—		33 (97.1%)	64 (79%)
Hemolysis aggravation, No. (%)			0.004			0.018
with	18 (43.9%)	63 (70%)		18 (52.9%)	61 (75.3%)
without	23 (56.1%)	27 (30%)		16 (47.1%)	20 (24.7%)

PNH: paroxysmal nocturnal hemoglobinuria; BMF: bone marrow failure.

Statistical significance was set at p < 0.05.

PNH-related manifestations during the COVID-19 pandemic

Hemolysis aggravation

From 6 December 2022, to the end of follow-up, 81 (61.8%) patients experienced hemolysis aggravation, among them, 2 patients had new thrombotic events. Of the 81 patients, 79 patients (97.6%) were infected, and only 2 (2.4%) were uninfected with Omicron (p = 0.000). In the infected group, all severe cases (100%) had hemolysis aggravation, while 62.9% of non-severe cases had hemolysis aggravation (p = 0.002).

Changes in laboratory parameters

During the COVID-19 pandemic, which started on 6 December 2022 hemolytic-related parameters were significantly worse in infected patients, especially those with severe infections. The HGB level was 65 (43–138) g/L, which was significantly lower in infected patients than in non-infected patients [80 (68–136) g/L, p = 0.010]. Severe cases had lower HGB levels than non-severe cases [56 (43–86) g/L vs. 77 (45–138) g/L, p = 0.016]. Similar results were observed for platelet (PLT) counts. PLT counts were lower in infected patients than in non-infected patients [112.5 (26–251) ×10⁹/L vs 151 (75–356) ×10⁹/L, p = 0.025], and even lower in severe cases than in non-severe cases [97 (26–157) ×10⁹/L vs 158 (63–251) ×10⁹/L, p = 0.013). Absolute neutrophil count (ANC) was significantly lower in infected patients than non-infected patients [1.81 (0.18–6.52) ×10⁹/L vs 3.62 (1.01–6.45) ×10⁹/L, p = 0.040], and even lower in severe cases than non-severe cases [1.35 (0.18–3.59) ×10⁹/L vs 2.68 (0.8–6.52) ×10⁹/L, p = 0.013]. Among those infected (n = 115) with Omicron, there was a significant increase in LDH levels compared to the uninfected group (n = 16, 68.7% vs 12.5%, p = 0.000), and individuals with severe infections (n = 18) had an even higher LDH increase rate than those without (n = 97; 100% vs 62.9%, p = 0.002).

PNH outcome

At the end of follow-up, 53 patients (65.4%) with aggravated hemolysis returned to the baseline level, 27 patients (33.3%) worsened, and 1 patient (1.2%) died of COVID-19. Infected patients had a lower recovery rate than uninfected patients (64.5 vs 100%, p = 0.542), and severe cases had worse recovery rates than non-severe cases (27.8 vs 78.7%, p = 0.000). One severe patient died from an omicron infection, with a death rate of 0.87%.

Predictors for aggravation of hemolysis and thrombosis

For the entire cohort, a comparison between those with and without hemolysis aggravation showed that the use of complement inhibitors [odds ratio (OR) (95% confidence interval (CI)): 0.311 (0.117–0.822), β < 0; p = 0.019] was independent factors for less hemolysis aggravation and omicron infection [OR (95%CI): 16.016 (3.334–76.946), β > 0; p = 0.001] was independent factors for hemolysis aggravation (Table 3).

Table 3. Clinical features before 6 December 2022, for all patients with or without hemolysis aggravation.

Clinical features	The whole cohort			Patients with Omicron infection
	With (n = 81)	Without (n = 50)	p Value	With (n = 79)	Without (n = 36)	p Value
Median age (range), yrs	43 (17–75)	42 (18–81)	0.914	42 (17–75)	42 (18–72)	0.971
Male ratio (%)	51.9%	48%	0.668	50.6%	52.8%	0.831
PNH type, No. (%)			0.382			0.309
With BMF	42 (51.9%)	22 (41.7%)		41 (51.9%)	15 (41.7%)
Without BMF	39 (48.1%)	28 (58.3%)		38 (48.1%)	21 (58.3%)
With thrombosis, No. (%)	26 (32.1%)	13 (26.0%)	0.458	24 (30.4%)	7 (19.4%)	0.220
baseline laboratory tests
HGB (g/L)	78 (36–105)	92 (42–132)	0.241	78 (36–105)	92 (42–132)	0.241
ANC (×10^9/L)	2.47 (0.85–4.2)	2.54 (0.49–4.5)	0.796	2.45 (0.85–4.2)	2.23 (0.49–3.7)	0.717
PLT (×10^9/L)	176 (66–339)	143.5 (64–352)	0.404	176 (66–339)	145 (74–265)	0.267
Ret (×10^9/L)	50 (22–85)	58 (21–89)	0.526	54 (22–89)	50 (26–85)	0.653
SGPT (U/L)	18 (6–33)	13 (4.6–36)	0.790	18 (6–33)	11.5 (5.88–30)	0.554
Tbil (μmol/L)	27.2 (15.5–93.3)	28 (11.3–111.8)	0.752	27.2 (15.5–93.3)	28.38 (18.9–111.8)	0.913
Dbil (μmol/L)	9.05 (4.5–13.1)	10.1 (4.7–12.8)	0.975	9.05 (4.5–13.1)	8.55 (4.7–12.4)	0.820
Cr (μmol/L)	51 (40–134)	67 (36–139)	0.434	51 (40–134)	69.5 (49–97)	0.521
LDH (U/L)	1146 (297–2364)	480 (223–2570)	0.520	1146 (297–2364)	436 (231–2572)	0.436
SF (μg/L)	106 (20–256)	110 (31–261)	0.630	110 (20–252)	112 (26–245)	0.568
PNH clone size (granulocyte flare)	82.5 (60–100%)	85 (61–100%)	0.742	84 (65–100%)	80 (61–100%)	0.459
COVID-19 vaccination status			0.004			0.018
Vaccinated, No. (%)	18 (23.2%)	23 (46%)		18 (22.8%)	16 (44.4%)
Unvaccinated, No. (%)	63 (77.8%)	27 (54%)		61 (77.2%)	20 (55.6%)
The median duration from vaccination to 6 December 2022 (range), month	12 (1–18)	7.5 (1–18)	0.216	12 (1–18)	7 (1–17)	0.348
Baseline treatment, No. (%)
Complement inhibitor	10 (12.3%)	17 (34.0%)	0.003	10 (12.7%)	13 (36.1%)	0.004
glucocorticoid	12 (14.8%)	3 (6.0%)	0.124	12 (15.2%)	3 (8.3%)	0.475
CsA	12 (14.8%)	9 (18.0%)	0.692	12 (15.2%)	6 (16.7%)	0.840
androgen	24 (29.6%)	13 (16.0%)	0.654	23 (29.1%)	8 (22.2%)	0.440
TPO-RA	1 (1.2%)	2 (4.0%)	0.558	1 (1.3%)	2 (5.6%)	0.230
blood transfusion	19 (23.5%)	4 (8.0%)	0.024	18 (22.8%)	3 (8.3%)	0.063
untreated	9 (11.1%)	8 (16.0%)	0.367	9 (11.4%)	7 (19.4%)	0.247
Infected with COVID-19, No. (%)			0.000
with	79 (97.6%)	36 (72.0%)		—	—
without	2 (2.4%)	14 (28.0%)		—	—
Severe COVID-19 infection, No. (%)						0.002
with	—	—		18 (22.8%)	0 (0%)
without	—	—		61 (77.2%)	36 (100%)

PNH: paroxysmal nocturnal hemoglobinuria; BMF: bone marrow failure; CsA: cyclosporine A; TPO-RA: thrombopoietin receptor agonist; HGB: hemoglobin; ANC: absolute neutrophil count; PLT:platelet; Ret: reticulocyte; SGPT: serum glutamic-pyruvic transaminase; Tbil: total bilirubin; Dbil: direct bilirubin; Cr: creatinine; LDH: lactate dehydrogenase; SF: serum ferritin; with: with hemolysis aggravation; without: without hemolysis aggravation.

Statistical significance was set at p < 0.05.

Univariate analysis showed that, for the infected patients, being vaccinated, receiving complement inhibitors, and having non-severe infections significantly reduced hemolysis aggravation (p = 0.018, p = 0.004, and p = 0.002, respectively). Only receiving complement inhibitors was the independent factor for protecting hemolysis aggravation on multivariate regression analysis [OR (95%CI): 0.318 (0.118–0.857), β < 0; p = 0.024, Table 3]. In total, 10 (58.8%) patients treated with terminal complement inhibitors developed BTH, whereas none of the patients who received proximal complement inhibitors developed BTH (p = 0.003, Figure 1). However, patients with proximal complement inhibitors had a statistical higher vaccination rate than those with C5 inhibitors (p = 0.014), meanwhile, there were 2 severe infections in the C5 complement inhibitor.

Discussion

The COVID-19 pandemic has significantly impacted public health both in China and globally. Patients with PNH are generally more susceptible to various infections, and the use of complement inhibitors increase susceptibility to encapsulated organisms, namely meningococcal and gonococcal infections, potentially leading to different clinical outcomes during a pandemic. Although a few reports have investigated PNH in other countries, their patients have different backgrounds from ours. First, most patients with hemolysis were treated with complement inhibitors or, on rare occasions, had no prior exposure to such complement inhibitors. Owing to different management policies, most countries have experienced a gradual occurrence of COVID-19.

Most importantly, the current publications often involve few participants [18–20]. In Asia, a report from Japan described 5 patients with PNH undergoing complement inhibitor treatment, all of whom had contracted COVID-19. Among them, 4 had mild COVID-19, and 1 had moderate COVID-19. They all encountered new hemolytic events during infection [21].

In this study, we included a relatively large number of patients with various disease types, some of whom were vaccinated or treated with various complement inhibitors (clinical trials), had concentrated outbreaks of infections during the winter of 2022–2023, and were newly infected with COVID-19.

Based on data from 131 cases, the infection rate of COVID-19 was 87.8%, which was similar to our assessment of the general population at the same time in China, although accurate reports have not yet been published. Findings from a South Korean study involving 40 patients and a Chinese study involving 3,715 patients with Omicron disease showed that fever remained the most prevalent symptom, followed by mild dry cough and sore throat [22,23]. In the present study, the most frequently reported symptoms were fever, sore throat, cough, phlegm, muscle and joint pain, fatigue, and headache. Unlike the symptoms from the initial COVID-19 strain [24–26], those associated with the Omicron variant were much milder [27–29]. In addition, the emerging literature concerning the outcome of Omicron, which appeared long after the initial COVID-19 outbreak, showed a substantial reduction in hospitalizations and deaths in the general population [30–33]. In Shanghai, China, the number of severe Omicron infections in the general population was 4.43% between April 2022 and May 2022 [34]. However, there are no data for China on the general population from 6 December 2022 to 10 April 2023. In the context of PNH, we observed a 15.7% incidence of severe infections, higher than that reported in the general population between April 2022 and May 2022.

We next analyzed the factors that could influence the infection rate, severity, and PNH hemolysis-related attacks during this pandemic period. Unlike most reports from other countries that have demonstrated that vaccination can reduce the risk of COVID-19 infection [35,36], we did not identify any factors that may protect patients from Omicron infection. One major difference is the type of vaccine used. Most vaccines from other countries are mRNA vaccines administered concurrently with viral exposure, and China mainly employs inactivated vaccines. Most individuals, including patients with PNH, were vaccinated long before 6 December 2022, and the country adhered to stringent COVID-19 control measures before 6 December 2022.

In addition, most of our patients were younger, whereas many reports from other countries focused on older populations. Variances in vaccine types, timing, age groups, management approaches, and disease status across different countries may explain the different protective effects of vaccinations against different viral strains at different times. However, reports on the protective role of complement inhibitors are scarce. Barcellini et al. found that using complement inhibitors does not decrease susceptibility to COVID-19 [22], which is consistent with our findings.

Although these factors did not affect the infection rate, some affected the severity of infection. We found that vaccination reduced the incidence of severe COVID-19. Multiple studies have demonstrated that vaccination can significantly reduce the incidence of severe or critical diseases caused by COVID-19 [35,37,38]. For instance, a few studies have shown that higher serum-positive detection rates of vaccines in patients with multiple myeloma and hematopoietic stem cell transplantation may be associated with better prevention [39–42]. However, most of these reports were related to other hematological diseases. A previous study has shown that vaccination can achieve a good SARS-CoV-2 antibody response in patients with PNH [43]. Recent findings have highlighted the importance of neutralizing antibody levels as reliable predictors of protection against symptomatic COVID-19 [44–48].

Contrary to expectations, our study indicated that complement inhibitors do not substantially protect against infection severity. Numerous studies have confirmed the efficacy of complement inhibitors in reducing the severity of pneumonia and lowering the fatality rate in the general population [49–52]. Other studies have shown that complement inhibitors are beneficial both in vivo and in external models of novel coronavirus infection and may help reduce the clinical severity of a novel coronavirus infection [53,54]. This disparity may be attributed to differences in dosage, combined medications, and types of complement inhibitors. Our study used crovalimab, iptacopan, and CAN106, whereas other studies used eculizumab or AMY-101. However, these effects of case reports/series mentioned above were not seen in clinical trials. Although patients treated with different complement inhibitors had similar infection rates, those treated with iptacopan seemed to have less severe cases. However, the difference was not statistically significant, probably because of the limited number of cases.

In our study, COVID-19 infection exacerbated hemolytic-related events, as evidenced by more severe hemolytic attacks in infected patients, particularly those with severe infections. Correspondingly, hemolytic-related laboratory parameters, such as HGB and LDH, were worse in infected patients, especially those with severe infections. COVID-19 is related to aggravated PNH symptoms; this has been shown by other investigators. Pike reported 4 cases of PNH infected with COVID-19, 3 of which had breakthrough hemolysis, and noted that infections increased the risk of breakthrough hemolysis in PNH due to increased complement system activation, overwhelming the complement blockade [55]. Schüller and Cavallaro et al. also reported 2 cases of hemolysis and exacerbation of chronic anemia in patients with PNH due to COVID-19 infection [56,57]. Changes in ANC and platelet counts may stem from the infections themselves, as observed in other infectious diseases [58,59], or potentially from aggravated hemolysis.

Our data revealed that complement inhibitors could reduce hemolytic attacks. This finding is consistent with those reported in the literature. Kulasekararaj et al. reported that 4 COVID-19-infected patients with PNH with complement inhibitors experienced fewer hemolysis-related symptoms than their non-inhibited counterparts [60]. Other small-sample studies have similarly indicated the beneficial role of complement inhibitors during COVID-19 infections [6,61]. Some studies indicated that complement inhibitors can reduce the inflammatory response caused by COVID-19 in both PNH and non-PNH cases [51,60].

Further analysis demonstrated that patients treated with a proximal complement inhibitor (iptacopan) had a significantly lower risk of developing BTH than those treated with a terminal complement inhibitor (p = 0.003). Research on iptacopan showed that this proximal complement inhibitor can effectively control intra- and extravascular hemolysis, leading to transfusion-free and normal HGB levels [7,62]. In addition, iptacopan had a very low BTH level during the experimental period, which is consistent with our results. However, patients with iptacopan had a statistical higher vaccination rate than those with C5 inhibitors (p = 0.014), meanwhile, there were 2 severe infections in the C5 complement inhibitor. Considering the factors above, as well as the limited number of cases, the conclusion ‘proximal complement inhibitors may reduce BTH events in the setting of COVID-19’ needs further studies to be evaluated.

In our study, only 1 patient died from the infection, resulting in a mortality rate of 0.87%, significantly lower than that observed in the early days of the COVID-19 pandemic. The patient had severe COVID-19 and did not receive vaccination or complement inhibitor treatment.

Conclusions

Although the retrospective nature of this study introduced some bias, the prospective data collection remained challenging. With a relatively large sample size of consecutive patients and a short timeframe for exposure to the Omicron-variant strain, we described the characteristics of patients with PNH under specific conditions. Furthermore, our study showed that vaccination can reduce the severity of COVID-19, whereas COVID-19 infection exacerbates hemolytic episodes. Complement inhibitors did not significantly influence infection severity and seemed to mitigate the aggravation of hemolytic-related events in patients with PNH.

Author contributions

Bing Han and Chen Yang were responsible for the initial planning and study design. Leyu Wang, Qinglin Hu, Yuan Yang, Miao Chen, and Chen Yang were responsible for data collection, extraction, and statistical analyses. Leyu Wang was responsible for the data interpretation and drafting of the manuscript. Bing Han and Chen Yang are guarantors and have full access to all data, including statistical reports and tables, and take full responsibility for the integrity of the data and the accuracy of the data analysis.

Acknowledgement

The funding agencies played no role in the study design, collection, analysis, and interpretation of data, writing of the report, or the decision to submit the article for publication. The authors have no financial or non-financial conflicts of interest to disclose.

This study was supported by the National Natural Science Foundation [82370121], Beijing Natural Science Foundation [7232109], National High-Level Hospital Clinical Research Funding [2022-PUMCH-C-026, 2022-PUMCH-D-002, and 2022-PUMCH-B-046], CAMS, and the Innovation Fund for Medical Sciences [CIFMS 2021-I2M-1-003].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All the data are available in the main text. All the detailed metadata are available upon reasonable request from the corresponding author.

Additional information

Funding

This study was supported by the National Natural Science Foundation [82370121], Beijing Natural Science Foundation [7232109], National High-Level Hospital Clinical Research Funding [2022-PUMCH-C-026, 2022-PUMCH-D-002, and 2022-PUMCH-B-046], CAMS, and the Innovation Fund for Medical Sciences [CIFMS 2021-I2M-1-003].