Peripheral facial palsy post SARS-CoV-2 vaccine: A regional pharmacovigilance cases series

Abstract

Peripheral facial palsy (PFP) is a rare adverse reaction identified from clinical trials of coronavirus disease 2019 (COVID-19) vaccines (messenger ribonucleic acid [mRNA] and viral vector). Few data are available on their onset patterns and risk of recurrence after re-injection of a COVID-19 vaccine; the objective of this study was to describe PFP cases attributed to COVID-19 vaccines. All cases of facial paralysis reported to the Regional Pharmacovigilance Center of Centre-Val de Loire area between January and October 2021, in which the role of a COVID-19 vaccine was suspected, were selected. Based on initial data and following additional information requested, each case was reviewed and analyzed to include only confirmed cases of PFP for which the role of the vaccine could be retained. From the 38 cases reported, 23 were included (15 excluded because of diagnosis not retained). They occurred in 12 men and 11 women (median age of 51 years). The first clinical manifestations occurred with a median time of 9 days after COVID-19 vaccine injection, and the paralysis was homolateral to the vaccinated arm in 70%. The etiological workup, always negative, included brain imaging (48%), infectious serologies (74%) and Covid-19 PCR (52%). Corticosteroid therapy was prescribed for 20 (87%) patients, combined with aciclovir in 12 (52%). At 4-month follow-up, clinical manifestations had regressed completely or partially in 20 (87%) of the 23 patients (median time of 30 days). From them 12 (60%) received another dose of COVID-19 vaccine and none had a recurrence and the PFP regressed despite the second dose in 2 of the 3 patients not fully recovered at 4 months. The potential mechanism of PFP after COVID-19 vaccine, which don’t have a specific profile, is probably the interferon-γ. Moreover, the risk of recurrence after a new injection appears to be very low, which makes it possible to continue the vaccination.

Abbreviations

BP

Bell's palsy

HSV

herpes simplex virus

IFN-γ

interferon-γ

IRR

incidence rate ratio

mRNA

messenger ribonucleic acid

OR

odds ratio

PCR

polymerase chain reaction

PFP

peripheral facial palsy

RPVC

Regional Pharmacovigilance Center

TNF-α

tumor necrosis factor α

VZV

varicella-zoster virus

Introduction

The risk of peripheral facial palsy (PFP) was among the rare adverse reactions labelled when the Comirnaty® (tozinameran) and Spikevax® (elasomeran) messenger ribonucleic acid (mRNA) vaccines were first marketed. In the phase III clinical trials, the incidence of PFP was indeed higher in the vaccinated group than in the placebo group for these two vaccines (e.g., for tozinameran: 4/18,860 in the vaccinated group versus 0/18,846 in the placebo group) [1], [2]. By combining these studies, Ozonoff et al. [3] have estimated that the incidence of PFP in vaccinated patients was 3.5 to 7 times higher than that expected in the general population. For Vaxzevria® (ChAdOx1-S) and Jcovden® (Ad26.COV2-S), this adverse reaction was identified from clinical trials and post-marketing studies, respectively. Since then, many articles on this topic have been published, and included in a recent review of the literature [4]. However, few data are available on the onset patterns of these PFPs, how they evolve, and whether they recur after additional administration of a coronavirus disease 2019 (COVID-19) vaccine, although this is a frequent question from physicians and patients. The objective of this study was to describe the characteristics and evolution, especially in case of a re-injection, of PFP cases attributed to COVID-19 vaccines reported to the Regional Pharmacovigilance Center (RPVC) of the Centre-Val de Loire area (France).

Methods

All cases of facial paralysis reported to the RPVC of Centre-Val de Loire area between January 1, 2021 and October 6, 2021, in which the role of a COVID-19 vaccine was suspected, were selected. Based on initial data provided by the reporter, or following additional information requested by a RPVC physician, each case was reviewed and analyzed to include only confirmed cases of PFP for which the role of the vaccine could be retained [5]. For each included case, the following information was retrieved: age, sex, history of symptomatic COVID-19 infection, history of facial paralysis and diabetes, the name and shot rank of the vaccine used, the time to PFP onset, the initial clinical presentation, additional examinations performed (e.g., search for a viral infection, imaging,), therapeutic management, evolution at 4 months and existence of recurrence in case of re-injection of a COVID-19 vaccine. The clinical profile and course were compared according to the rank of vaccination and the existence of a history of symptomatic COVID-19.

Results

Between January 1 and October 6, 2021, 38 cases of facial paralysis occurring after COVID-19 vaccination were reported to the RPVC Centre-Val de Loire (23 cases reported by physicians and 15 by patients). After collection of additional information and analysis, 15 cases were excluded, as the diagnosis of PFP could not be retained: 2 because an identified etiology (presence of an auricular zoster or a jugal cyst), 1 central facial palsy, 9 cases for which the reported clinical manifestations did not allow to confirm the diagnosis of PFP (5 paresthesia of the hemiface contemporary with a transient edema of the face, 1 transient sensation of dental anesthesia, 1 fleeting facial paresis (less than one minute), 1 transient speech disorder and 1 trismus due to dental involvement) and 3 cases with insufficient data.

The 23 remaining selected cases occurred in 12 men and 11 women, with a median age of 51 years [range, 14–85]. Only one patient had a history of facial palsy (which occurred 15 years earlier and regressed in 3 weeks) and 2 (9%) had diabetes mellitus. Finally, one patient had been treated for several years with interferon-α for malignant melanoma.

The first clinical manifestations occurred with a median time of 9 days [1 hour–57 days] after vaccine injection. The paralysis was homolateral to the vaccinated arm in 16 (70%) patients and contralateral in 7 (30%). The etiological workup, always negative, included brain imaging for 11 (48%) patients, viral serologies (herpes simplex type 1 [HSV], varicella-zoster virus [VZV], Epstein-Barr virus, human immunodeficiency virus) or other infectious serologies (B. Burgdorferi, T. Pallidum) for 17 (74%) patients and COVID-19 polymerase chain reaction (PCR) for 12 (52%) patients. Corticosteroid therapy was prescribed for 20 (87%) patients, combined with aciclovir in 12 (52%) and/or physiotherapy in 13 (56%) patients.

At 4-month follow-up, clinical manifestations had regressed completely or partially in 20 (87%) patients, with a median time of 30 days [1–105]. For two patients, the recovery time was abnormally short (less than one week). Regression was total and without after-effects in 7 (30%) patients, but partial for 13 (70%) with mild after-effects such as ocular involvement in 9 patients (dry eye [n  = 2], sensation of ocular discomfort with fatigue [n  = 4] or less tonic eye [n  = 1], delay in blinking [n  = 1], lacrimation during meals [n  = 1]), facial asymmetry with fatigue in 3 patients and episodic pain in 3 patients (of the hemi-face [n = 2], of the upper hemi-lip [n  = 1]). For 3 (13%) patients, the PFP had not regressed after 4 months of evolution despite a treatment combining aciclovir, corticoids and physiotherapy. These patients were a 46-year-old man and 2 women aged 51 and 74 years, for whom paralysis occurred 8, 9 and 25 days after vaccination, respectively.

Concerning the vaccines, tozinameran and ChAdOx1-S was reported in 22 (96%) and 1 case (4%), respectively. PFP occurred after the first dose in 15 cases (65%) of which 4 (27%) in patients with a history of symptomatic COVID-19 in the previous 3 to 9 months, and after the second dose in 8 cases (35%). The comparison of characteristics according to injection rank and history of symptomatic COVID-19 is shown in Table 1 .

Table 1.

Characteristics of PFP according to injection rank and history of symptomatic COVID-19.

	Circumstances of occurrence
	Post 1st dose without history of COVID-19	Post 1st dose with history of COVID-19	Post 2nd dose
Patients (n)	11	4	8
Sex ratio	0.27	0.75	0.63
Age in years, median (range)	46 [14–85]	50 [37–57]	51 [32–74]
Time to onset in days, median (range)	3 [1–32]	3 [0–47]	15 [0–57]
Peripheral facial palsy, n (%)
Homolateral to the vaccinated arm	8 (73%)	3 (75%)	5 (63%)
Symptoms
Mouth deviation	11 (100%)	4 (100%)	8 (100%)
Incomplete eye closure	10 (91%)	4 (100%)	7 (87%)
Speech impediment	7 (64%)	4 (100%)	8 (100%)
Blurring of facial folds	8 (73%)	3 (75%)	7 (87%)
Eating/drinking disorder	8 (73%)	3 (75%)	5 (62%)
Mastoid pain	5 (45%)	3 (75%)	4 (50%)
Agueusia	4 (36%)	1 (25%)	2 (25%)
Hyperacusis	2 (18%)	0%	2 (25%)
Hypoesthesia concha ear	1 (9%)	0%	0%
Evolution at 4-month follow-up
Regression	10 (91%)	4 (100%)	6 (75%)
-total regression	3 (27%)	2 (50%)	2 (25%)
time in days, median [range]	38 [0–75]	28 [7–105]	30 [3–105]
-partial regression (slight sequelae)	7 (64%)	2 (50%)	4 (50%)

Fifteen of the 23 patients (65%) who developed PFP received another dose of vaccine, 11 (73%) with tozinameran and 4 (27%) with elasomeran, within a median of 177 days [30–393] after the first dose and 150 days [20–390] after the onset of PFP. Among them, 12 (80%) patients were fully recovered at the time of the second injection (median of 123 days [16–288]) and none had a recurrence. In the 3 patients who were not fully recovered (PFP evolving for 20, 38 and 40 days), the PFP regressed despite the second dose in 2 of them (over 20 and 25 days respectively), while in the third, there was no regression at 4-month follow-up. Two patients did not receive another dose of vaccine because they did not recover from PFP, and the three were lost of follow-up.

Discussion

Idiopathic peripheral facial palsy or Bell's palsy (BP) is a paralysis of the peripheral facial nerve, which affects the upper and lower parts of the face (erasure of the nasolabial fold, fall of the labial commissure, incomplete closure of the eye and erasure of the forehead wrinkles), which can be maximal in a few hours or rapidly progressive (less than 72 hours) and can be accompanied by retroauricular pain, hyperacusis or ageusia of the anterior two-thirds of the hemitongue [6]. It is the main cause of peripheral facial paralysis and its annual incidence is estimated at 20–30 cases per 100,000 inhabitants [6]. Management is based on early initiation of oral corticosteroid therapy and prevention of corneal damage. The first signs of recovery usually occur within 8 to 15 days, and full recovery within 3 to 4 months in most cases. In 5–10% of patients, recovery is incomplete, particularly in cases of severe initial damage [7], [8]. Recurrence occurs in 10% of these patients [9].

In our study on 23 cases of COVID-19 vaccine-associated PFP, the profile of the patients in terms of sex and age, as well as the evolution (87% cured or with mild after-effects at 4 months) are superimposed to the usual profile of PFP [7], [8]. The majority of the reported cases (96%) concerned tozinameran, which is expected since this was the vaccine most commonly used at the time of the collection. Similarly, the predominance of cases after the first dose (65%) is probably explained by the fact that during the inclusion period, the number of patients who received a second dose was still low. In fact, the risk is higher after the second dose as evidenced by the clinical trials, in which cases occurred almost exclusively after the second dose (6/7 cases). This increased risk of PFP was observed during the first 14 days after the second dose of tozinameran in both nested case-control (adjusted OR: 2.325, 95%CI: 1.414–3.821) and self-controlled case series studies (adjusted IRR = 2.44, 95%CI: 1.32–4.50) [10], [11].

The median onset time of 9 days is consistent with other studies, estimated between 9.3 and 14 days, in a review of the literature [4]. Furthermore, the pattern of cases in our study was broadly similar regardless of the rank of the injection and the existence of a history of COVID-19, except for a much longer median time of onset for cases occurring after the second dose compared to the first (15 days versus 3 days). This longer time of onset after the second dose was also observed in the study by Shemer et al. [11] although the difference was less marked (14 days ± 4 and 9 days ± 4).

To our knowledge, there is no published prospective follow-up to accurately estimate the risk of PFP recurrence after re-injection and, to date; only one case of recurrence after a second dose of vaccine has been published [12].

In our study, more than half of the patients had a vaccine re-injection (15 patients, i.e., 65% of cases), and none of the patients cured at the time of the second injection (n  = 12) had a recurrence of PFP. Moreover in patients whose PFP had not fully recovered, improvement continued despite the re-injection. These data are inconsistent with the study by Bertin et al. [4], which estimates the recurrence rate at 41% (20/49) based on positive reintroduction cases recorded in the international pharmacovigilance database (Vigibase) with four COVID-19 vaccines. However, as these are spontaneous reports, this rate is probably overestimated because a recurrence during a re-injection is much more often the subject of a new pharmacovigilance report than the absence of a recurrence, which is rarely reported. Moreover, unlike ours, the diagnosis of cases extracted from Vigibase has not been systematically validated, and it is impossible to confirm that they were indeed cases of PFP and not another pathology.

The pathogenic mechanisms can be considered as direct (release of pro-inflammatory cytokines involved in the genesis of PFP) or indirect when the vaccination is likely to induce events that are themselves a risk factor for PFP. Regarding indirect mechanisms, numerous HSV and VZV viral reactivations secondary to COVID-19 vaccination have been reported [13], and these infectious etiologies are at the forefront of the causes of PFP. Similarly, messenger ribonucleic acid (mRNA) vaccines against COVID-19 are associated with hypertensive flares, sometimes of high grade [14], one of the mechanisms described in drug-associated PFP [15].

COVID-19 vaccines induce a significant release of different pro-inflammatory cytokines such as interleukin-6, TNF-α or interferon-γ [16] that are overexpressed in patients with Bell's palsy [17], [18], [19]. Compared to the first injection, the more than 10-fold increase in interferon-γ after the second shot [16] could explain that this adverse effect occurs preferentially after the second dose.

This study has several limitations. First, as these cases were reported spontaneously by healthcare professionals or patients, they are not representative of the case profiles related to all COVID-19 vaccines. Cases reported to the RPVC are often more serious or have less favorable outcomes, especially when the adverse reaction is already labelled. Second, additional information on the initial clinical manifestations and evolution was collected, some of it with delay, due to the overload of activity. This delay may have contributed to the collection of less accurate data (symptoms, timeline). Third, the lack of use of an accurate classification of symptoms, such as House-Brackmann, may have led to a degree of subjectivity in the description of the initial condition and its evolution. Thus, for the persistence of the same symptom after several months, some patients declared themselves cured while others reported severe sequelae.

The main strengths of our study are the number of cases of PFP included, their description and most importantly the remote follow-up allowing us to describe the evolution. Another strength is the quality of the PFP cases analyzed, although they were pharmacovigilance reports.

Indeed, clinical validation and pharmacological expertise of the reported cases to analyze the link with the vaccine is an essential prerequisite for any study based on the basis of pharmacovigilance reports. Thus, of the 38 reports of facial palsy received in our center, if we exclude the 3 insufficiently documented cases, only 70% of them met the diagnostic criteria for PFP. Even for a report made by a health professional, the analysis of the reports by a clinical pharmacologist is a mandatory step before coding the cases, which is the only way to perform expert assessments on valid cases.

Thus, cases of PFP reported after COVID-19 vaccination have a profile that overlaps with other etiologies of PFP. The risk of recurrence after a new injection appears to be very low, which makes it possible to continue the vaccination, provided that the PFP has completely regressed.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Disclosure of interest

The authors declare that they have no competing interest.