Atypical thrombosis associated with VaxZevria® (AstraZeneca) vaccine: Data from the French Network of Regional Pharmacovigilance Centres.

Abbreviations

adverse drug reactions

French Medicines Agency

coronavirus disease 2019

French Regional Pharmacovigilance Network

cerebral venous thrombosis

disseminated intravascular coagulation

European Medicines Agency

severe acute respiratory coronavirus 2 syndrome

splanchnic thrombosis

thrombosis with thrombocytopenia syndrome

vaccine-induced prothrombotic immune thrombocytopenia

Starting in late 2019, the initial cases of a previously unknown form of pneumonia, now referred to as coronavirus disease 2019 (COVID-19), led to a global pandemic. In response, most countries have sought to curb the spread of the virus by imposing periods of lockdown as a function of the national infection rates. By the end of 2020, the advent of vaccines against this severe acute respiratory coronavirus 2 syndrome (SARS-CoV-2) prompted new hope in the global fight against the COVID-19 pandemic. In Europe, mRNA vaccines and adenovirus vector vaccines have received conditional marketing authorizations for active immunization against SARS-CoV-2 in individuals aged 16 and over.

On January 29th, 2021, the European Medicines Agency (EMA) authorized VaxZevria®, the AstraZeneca adenovirus vector vaccine directed against SARS-CoV-2 and in France, the campaign officially started on February 6, 2021.

These new vaccine technologies are now considered to be the best option of countering the COVID-19 pandemic. Given the high level of population likely to be exposed to these drugs, vaccine safety is a critical issue. In order to promptly and accurately identify potential new signal, the French Medicines Agency (ANSM) oversees the assessment of vaccine safety and has initiated a specific strengthened surveillance system for adverse drug reactions (ADRs) related to COVID-19 vaccines in France. This system is based on the collaboration between the Regional Pharmacovigilance Network (CRPV) and the expert council of the specific ANSM/CRPV monitoring committee for vaccines [1].

In this letter, we describe and discuss the VaxZevria® associated-atypical thrombosis specific signal identified by this committee.

In France, VaxZevria® ADRs reporting was initially dominated by flu-like syndromes. In late February 2021, the first report of serious, unexpected, thrombotic events associated with coagulation disorders namely thrombocytopenia and disseminated intravascular coagulation (DIC) was identified by the two CRPV in charge of the survey, which alerted the French authorities.

This potential signal, also observed in other European countries, was confirmed by the EMA on March 18, 2021 and definitely validated on April 7th, 2021 [2]. Initially, the at-risk population, thought to be limited to young women, prompting member states to adapt their vaccination policy accordingly. Since then, various attempts have been made to define this new atypical thrombosis entity, with different entry points according to the presence of thrombocytopenia (thrombosis with thrombocytopenia syndrome [TTS] as proposed by the Brighton collaboration [3]), thrombosis (vaccine-induced prothrombotic immune thrombocytopenia [VIPIT] [4]) or anti-PF4 antibodies [5].

In the context of this signal, the ANSM/CRPV specific monitoring committee on vaccines considered 4 categories of interest:

cerebral venous thrombosis (CVT) or splanchnic thrombosis (ST);

multi-site thrombosis whether or not associated with thrombocytopenia or coagulation disorders;

any other thrombosis associated with thrombocytopenia or coagulation disorders;

isolated DIC.

In France, 11,206 ADR reports of which 2811 were serious have been received up to April 15th, 2021 from healthcare professionals and patients for a total of 3,263,188 injections of VaxZevria®. Of these, 360 mentioned venous and/or arterial thrombosis. According to the above defined categories, 27 cases fulfilled the criteria of the atypical thrombosis, i.e. a notification rate [95% confidence interval] of 0.8 [0.54–1.20] per 100,000 doses. These cases involved 13 women and 14 men, and the median (range) age was 60 years (21-74) (Table 1 ). There was no particular history or risk factor apart from long term well tolerated oral estroprogestative contraception in 4 patients. The median (range) time to onset was 11 days (2 to 35). Of the 16 patients tested for 12 were positive for anti-PF4 antibodies. There were 8 fatal issues, giving a mortality rate of 30%.

Table 1

Details of the 27 reported cases of atypical thrombosis following administration of the VaxZevria vaccine against COVID-19 between February 6th and April 15th, 2021.

N	Sex	Age	Onset period	Time to onset (days)	Cerebral venous thrombosis	Splanchnic thrombosis	Other thrombosis	Thrombocytopenia	Disseminated intravascular coagulation	Anti-PF4 antibodies	Causality assesment [6]
Brighton collaboration level 1
1a	M	41	W15	17	Yes	Yes	Yes	Yes	Yes	Yes	I5
2a	M	63	W11	11	Yes	Yes	Yes	Yes	Yes	Yes	I5
3	F	21	W12	15	Yes	Yes	Yes	Yes	No	Yes	I6
4a	F	69	W14	12	Yes	No	Yes	Yes	Yes	Yes	I5
5	F	26	W8	9	No	Yes	Yes	Yes	Yes	Yes	I6
6	M	73	W15	12	No	Yes	Yes	Yes	No	Yes	I6
7a	F	61	W14	13	No	Yes	Yes	Yes	Yes	Yes	I5
8a	F	38	W12	8	Yes	Yes	No	Yes	Yes	No	I3
9	F	74	W14	15	No	No	Yes	Yes	Yes	Yes	I6
10	M	23	W12	9	Yes	No	No	Yes	Yes	Yes	I6
11	F	44	W12	9	Yes	No	No	Yes	Yes	Yes	I6
12	M	60	W14	11	Yes	No	No	Yes	No	Yes	I6
13a	M	60	W14	11	No	Yes	No	Yes	No	No	I3
14a	M	67	W14	8	No	Yes	No	Yes	Yes	No	I3
Other categories of interest
15	F	67	W13	11	No	No	No	Yes	Yes	ND	I3
16	F	73	W12	10	No	No	No	Yes	Yes	Yes	I4
17	F	24	W11	35	Yes	No	Yes	No	No	ND	I2
18	M	51	W11	6	Yes	No	No	No	No	No	I2
19	F	53	W12	18	Yes	No	No	No	No	ND	I2
20	M	54	W14	25	No	Yes	No	No	No	ND	I2
21	M	56	W14	2	No	Yes	No	No	No	ND	I2
22	F	61	W15	17	No	Yes	No	No	No	ND	I2
23	M	68	W15	11	No	Yes	No	No	No	No	I2
24	M	73	W15	30	No	Yes	No	No	No	ND	I2
25a	M	24	W11	7	No	Yes	No	No	No	ND	I2
26	M	61	W12	2	No	Yes	No	No	No	ND	I2
27	F	58	W14	8	No	Yes	No	No	No	ND	I2

ND: not determine; W: calendar week.

Fatal issue.

Fourteen of the 27 cases met the level 1 criteria of the Brighton collaboration, with a median (range) trough platelet count of 25 G/L (9–61), and 13 cases corresponded to other categories of interest defined by the ANSM/CRPV specific monitoring committee.

Of the 14 “level 1” cases, 9 corresponded to multiple thrombosis and thrombocytopenia and/or DIC, and 5 to isolated CVT (n  = 3) or ST (n  = 2) with thrombocytopenia. The 13 last cases included CVT (n  = 3) and ST (n  = 8) with no evidence of thrombocytopenia, and isolated DIC (n  = 2).

These atypical thrombosis were not reported during the clinical trials of VaxZevria® [7], generally not powerful enough to identify rare ADRs, which emphasizes the importance of a close real time safety monitoring and scientific analysis by pharmacovigilance experts, as the risk was not identified after the first uses outside Europe.

Among the several possible explanations for these extremely rare thrombotic complications associated with thrombopenia and occurring within 1 to 2 weeks after vaccination with VaxZevria®, an autoimmune heparin-induced-like thrombocytopenia is the most frequently discussed. This well-known prothrombotic disorder is caused by platelet-activating antibodies that bind to multimolecular complexes between cationic PF4 and anionic heparin. However, anti-PF4 antibodies can be induced by substances other than heparin [5], [8]. According to the German Society of Thrombosis and Haemostasis Research, vaccination is likely to induce the formation of antibodies against platelet antigens as a part of the inflammatory reaction and immune stimulation [9]. In line with previous reports from Norway [8] and Germany [5], 75% of our tested patients were positive for anti-PF4 antibodies. Screening must be performed under specific conditions, since, in our series, 6 of the 12 positive patients were initially seronegative with rapid screening assays.

Nevertheless, despite an adapted technique, in some of our patients sharing similar clinical and radiological pictures of those described as VIPIT/TTS, the anti-PF4 antibodies remained negative and in some of those cases no thrombocytopenia or coagulation disorders were observed suggesting other potential pathophysiological mechanisms justifying further investigation. Likewise, it is important to identify the causal determinant of these reactions:

factors linked to the vaccine itself;

factors linked to the induced immune reaction;

factors linked to the patient himself.

To date, no particular risk factor has been identified in patients and although the incidence of cases was initially higher in young women [3], in the most recent data, atypical thrombosis equally affected men and patients above 60 years although the incidence of notification remains proportionately higher in younger patients.

Hence, all these points are essential to establish a diagnostic strategy allowing rapid identification of cases and to determine the most appropriate therapeutic attitude, notably on the choice of anticoagulants and the role of immunosuppressive therapies.

The existence of a similar confirmed signal for Johnson & Johnson vaccine [10], [11], another adenovirus vector vaccines, may help guide further researches to answer those questions and more broadly on VaxZevria's place in COVID-19 immunization policy, even though this clinical picture remains rare and do not currently cast doubt on this vaccine risk/benefit ratio.

Disclosure of interest

The authors declare that they have no competing interest.