Polymyalgia rheumatica and giant cell arteritis following COVID-19 vaccination: Results from a nationwide survey

ABSTRACT

We conducted a national in-depth analysis including pharmacovigilance reports and clinical study to assess the reporting rate (RR) and to determine the clinical profile of polymyalgia rheumatica (PMR) and giant cell arteritis (GCA) in COVID-19-vaccinated individuals. First, based on the French pharmacovigilance database, we estimated the RR of PMR and GCA cases in individuals aged over 50 who developed their initial symptoms within one month of receiving the BNT162b2 mRNA, mRNA-1273, ChAdOx1 nCoV-19, and Ad26.COV2.S vaccines. We then conducted a nationwide survey to gather clinical profiles, therapeutic management, and follow-up data from individuals registered in the pharmacovigilance study. A total of 70 854 684 COVID-19 vaccine doses were administered to 25 260 485 adults, among which, 179 cases of PMR (RR 7. 1 cases/1 000 000 persons) and 54 cases of GCA (RR 2. 1 cases/1 000 000 persons) have been reported. The nationwide survey allowed the characterization of 60 PMR and 35 GCA cases. Median time to the onset of first symptoms was 10 (range 2–30) and 7 (range 2–25) days for PMR and GCA, respectively. Phenotype, GCA-related ischemic complications and -large vessel vasculitis as well as therapeutic management and follow-up seemed similar according to the number of vaccine shots received and when compared to the literature data of unvaccinated population. Although rare, the short time between immunization and the onset of first symptoms of PMR and GCA suggests a temporal association. Physician should be aware of this potential vaccine-related phenomenon.

KEYWORDS:

• Polymyalgia rheumatica
• giant cell arteritis
• COVID-19 vaccination
• pharmacovigilance
• clinical study

Introduction

Giant cell arteritis (GCA) and polymyalgia rheumatica (PMR) are two closely related inflammatory diseases that primarily affect individuals aged 50 years and older. GCA is a systemic large vessel vasculitis (LVV), while PMR is characterized by pain in the shoulders and pelvic girdles.¹ There is a 50% overlap in features between GCA and PMR cases and a 20% overlap within PMR cases. The prevalence of GCA varies among ethnic groups, ranging from 1.5 per 100,000 in Japan to 250 per 100,000 in the United Kingdom.^2,3 PMR follows a similar distribution pattern as GCA but is 3 times more frequent overall. While the exact triggers of both diseases remain unknown, it is likely that genetic and environmental factors both play a role in their development. The introduction of coronavirus disease 2019 (COVID-19) vaccines represents the most promising opportunity to control the global pandemic. When the vaccination campaign started in Europe, 4 types of vaccines utilizing 2 different immunological processes (mRNA and viral vector vaccines) became available: BNT162b2 mRNA (Pfizer-BioNTech), mRNA-1273 (Moderna), ChAdOx1 nCoV-19 (Oxford-AstraZeneca), and Ad26.COV2.S (Johnson & Johnson/Janssen). Each COVID-19 vaccine employs a harmless version of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike protein as an antigen to stimulate the production of antibodies targeted against the spike protein, thereby preventing the virus from entering host cells.⁴ There have been reports of some serious adverse events, like myocarditis following mRNA vaccinations, or thrombosis following non-replicating viral vector vaccinations.⁵ However, there is limited knowledge regarding cases of PMR and GCA occurring after COVID-19 vaccination.

Some individuals experiencing the onset of PMR and GCA symptoms shortly after receiving COVID-19 vaccination have raised concerns about potential adverse events associated with immunization.^6,7 Intriguingly, a recent PMR/GCA-focused pharmacovigilance study utilizing an international database and conducted after the initiation of COVID-19 vaccination campaigns indicated a possible safety signal.⁸ However, both types of studies faced limitations regarding access to clinical data and follow-up information. To address this context, we initiated a national in-depth analysis including pharmacovigilance reports and clinical study among medical experts of PMR and GCA. First, based on the French pharmacovigilance database, we aimed to estimate the reporting rate (RR) of PMR and GCA cases in individuals aged over 50 who developed their initial symptoms within one month of receiving the BNT162b2 mRNA (Pfizer-BioNTech), mRNA-1273 (Moderna), ChAdOx1 nCoV-19 (Oxford-AstraZeneca), and Ad26.COV2.S (Johnson & Johnson/Janssen) vaccines. In the second part of the study, we conducted a nationwide survey to gather comprehensive clinical profiles, therapeutic management, and follow-up data from individuals registered in the pharmacovigilance database and an overview by worldwide literature.

Patients and methods

Pharmacovigilance study

We conducted a retrospective national population-based pharmacovigilance study of PMR and GCA cases using the well-established French pharmacovigilance system coordinated by the French drug Agency.⁹ Only French reports of PMR and GCA, who developed their symptoms within one month of receiving COVID-19 vaccination (BNT162b2 mRNA (Pfizer- BioNTech), mRNA-1273 (Moderna),ChAdOx1 nCoV-19 Oxford-AstraZeneca), and Ad26.COV2.S (Johnson & Johnson/Janssen), between December 27th, 2020, and August 30th, 2022 were included in the study. All reports underwent a comprehensive pharmacological, clinical, and biological assessment conducted by a pharmacologist at the Regional Center. The main outcomes were the national RR of PMR and GCA cases per million vaccinated individuals and per million doses administered in adults over 50 years old. To calculate the national RR, we used the total number of COVID-19 vaccinated individuals or the total number of administered doses over the study period as the denominator (available at https://solidarites-sante.gouv.fr/grands-dossiers/vaccin-covid-19/article/le-tableau-de-bord-de-la-vaccination). Secondary outcomes included the RR of PMR and GCA following the first (1st), second (2nd), and third (3rd) injections of COVID-19 vaccines, as well as variations based on the type of COVID-19 vaccine administered.

Nationwide survey

A retrospective nationwide survey was conducted to gather de novo cases of PMR and GCA registered in the pharmacovigilance database. These cases were obtained from the French National Society of Internal Medicine (SNFMI) and the French Study Group for Large Vessel Vasculitis (GEFA). The diagnosis of PMR and GCA was confirmed when patients met the ACR/EULAR classification criteria for PMR and GCA, respectively.^10,11 Patients with positive anti- citrullinated protein antibodies (ACPA) were excluded. This study received approval from the Institutional Review Board of the Assistance Publique – Hôpitaux de Marseille (Health Data Hub number F20211207134309) and was conducted in accordance with the Declaration of Helsinki. A standardized electronic form was sent to each physician who declared a patient. Anonymized data were thus collected in a centralized database. We retrieved the following for all of the included patients: demographics, cardiovascular risk factors, type of vaccination and immediate clinical reaction, time to first symptoms, clinical manifestations at diagnosis, laboratory tests (including C-reactive protein (CRP), erythrocyte sedimentation rate, hemoglobin and platelet levels), HLA DRB typing by sequence-specific primers (PCR-SSP), temporal artery biopsy (TAB) status and the results from imaging (including articular echography, temporal arteries echography, computed tomography angiography (CTA) scan and fluorodeoxyglucose positron emission tomography with computed tomography (PET-CT) scan) were reported when available. The outcomes and follow-up data for all the patients were retrieved: clinical status, dose of corticosteroids (CS) per kilogram of body weight and requirement for immunosuppressive drugs (mainly methotrexate (MTX) or tocilizumab (TCZ)) at treatment initiation and at months 3, 6 and 12 (for PMR), at months 3, 6, 12 and 18 (for GCA). We defined relapse as a reoccurrence of symptoms and/or inflammatory parameters on laboratory findings without any other cause identified than PMR or GCA and that required a sustained increase in the CS dose and/or the addition of a CS-sparing agent.

Literature review

We conducted a review of the MEDLINE database (National Library of Medicine, Bethesda, MD) from 2021 to 2023, utilizing and combining the following keywords: PMR, GCA, Horton’s disease, COVID-19 vaccination, BNT162b2 mRNA(Pfizer-BioNTech), mRNA-1273 (Moderna), ChAdOx1 nCoV-19 (Oxford-AstraZeneca), and Ad26.COV2.S (Johnson & Johnson/Janssen) vaccines. Only confirmed cases of PMR and GCA with pertinent clinical data available were included in the analysis.

Statistical analysis

Categorical variables were presented as numbers (%), and quantitative variables were expressed as medians [range]. The three clinical groups, categorized by the number of vaccine shots received (1, 2, or 3), were compared with respect to patient characteristics, biological and imaging parameters, administered treatment, and outcomes using the Kruskal-Wallis test with False Discovery Rate (FDR) correction for multiple comparisons. Statistical analyses were performed using GraphPad Prism 5.0, and p-values <.05 were considered statistically significant.

Results

Pharmacovigilance study

During the study period, a total of 70,854,684 COVID-19 vaccine doses were administered, comprising 50,587,266 doses of BNT162b2 mRNA (Pfizer-BioNTech), 11924,026 doses of mRNA-1273 (Moderna), 7,390,842 doses of ChadOx1 nCoV-19 (Oxford-AstraZeneca), and 952,550 doses of Ad26.COV2.S (Johnson & Johnson/Janssen) to 25,260,485 adults over the age of 50. Pharmacovigilance centers received reports of 179 cases of PMR with a RR of 7.1 cases per 1,000,000 persons and 54 cases of GCA with a RR of 2.1 cases per 1,000,000 persons. The details of these cases are summarized in Table 1.

Table 1. Reporting rate of PMR and GCA in patients following COVID-19 vaccination.

	Shot number	Number of injected doses	Number of PMR	PMR RR (/10^6 doses)	Number of GCA	GCA RR (/10^6 doses)
Overall vaccination	All	70 854 684	179	2.53	54	0.76
	1	25 223 505	61	2.42	26	1.03
	2	24 611 543	87	3.53	18	0.73
	3	21 019 636	31	1.47	10	0.48
BNT162b2 mRNA (Pfizer-BioNTech)	All	50 587 266	128	2.53	39	0.77
	1	17 554 205	37	2.11	18	1.03
	2	17 851 086	65	3.64	13	0.73
	3	15 181 975	26	1.71	8	0.53
MRNA-1273 (Moderna)	All	11 924 026	22	1.84	6	0.5
	1	2 731 426	7	2.56	2	0.73
	2	3 354 939	10	2.98	2	0.6
	3	5 837 661	5	0.86	2	0.34
ChadOx1 nCoV-19 (Oxford-AstraZeneca)	All	7 390 842	27	3.65	9	1.22
	1	3 989 488	15	3.76	6	1.5
	2	3 401 354	12	3.53	3	0.88
Ad26.COV2.S (Johnson&Johnson/Janssen)	All	952 550	2	2.1	0	0
	1	948 386	2	2.11	0	0
	2	4164	0	0	0	0

PMR: Polymyalgia rheumatica; GCA: Giant cell arteritis.

RR: Reporting Rate.

Considering all cases of PMR and GCA, regardless of the type of vaccine administered, a national RR of 2.53 (range 0–3.76) and 0.76 (range 0–1.5) per 1,000,000 vaccine doses was observed, respectively. The highest RR for PMR (3.76) and GCA (1.5) occurred after the 1st dose of the ChadOx1 nCOV-19 (Oxford-AstraZeneca) vaccine. For mRNA vaccines, the highest RR was consistently observed after the 2nd shot for PMR cases (BNT162b2 mRNA (Pfizer-BioNTech): 3.64; mRNA-1273 (Moderna): 2.98) and after the 1st dose for GCA cases (BNT162b2 mRNA (Pfizer-BioNTech): 1.03; mRNA-1273 (Moderna): 0.73). Regarding non-mRNA vaccines, the highest RR occurred after the 1st shot for PMR (ChadOx1 nCoV-19 (Oxford-AstraZeneca): 3.76; Ad26.COV2.S (Johnson & Johnson/Janssen): 2.11) and for GCA cases (ChadOx1 nCoV-19 (Oxford-AstraZeneca): 0.88). No GCA was reported after Ad26.COV2.S (Johnson & Johnson/Janssen) vaccine.

Nationwide survey

PMR

The nationwide survey facilitated the identification of 60 PMR cases (22 (36.6%) occurring after the 1st shot, 28 (46.6%) after the 2nd shot, and 10 (16.8%) after the 3rd shot), median age of 71 years (range 52–94), 26 women (43.3%), whose characteristics are summarized in Table 2. The primary cardiovascular risk factor identified was hypertension, mentioned in 30 (50%) cases. Among the patients, 54 (90.1%) were vaccinated with mRNA vaccines, while 6 (9.9%) received viral vector vaccines. The median time to the onset of the first symptoms was 10 (range 2–30) days. All patients experienced girdle pain. Distal arthralgia was reported by 24 (40%) patients. Systemic signs were reported by 49 (81.6%) patients. Unusual symptoms were reported by 6 (10%) patients, including neck pain (n = 3), distal edema (n = 1), hemicrania (negative TAB)) (n = 1), and exudative pleuritis (n = 1). All patients exhibited a biological inflammatory syndrome, with a median CRP level of 55 mg/L (range 14–224). Among the 32 (53.3%) patients who underwent a PET-CT scan, 29 (90.6%) exhibited the typical hypermetabolism associated with PMR. No significant differences in terms of phenotype, biological markers, or PET-CT imaging were detected between the subgroups identified based on the number of vaccine shots received.

Table 2. Characteristics of the whole cohort of PMR patients following COVID-19 vaccination.

Patients	All patients n = 60	1st vaccine n = 22	2nd vaccine n = 28	3rd vaccine n = 10	p-value
Demographics
Age	71 [52–94]	71 [52–94]	72 [54–87]	67 [59–74]	0.701
<70 years	22 (36.6)	7 (31.9)	11 (39.2)	4 (40)	0.812
70–80 years	28 (46.6)	9 (40.9)	14 (50)	5 (50)	0.621
>80 years	10 (16.8)	6 (27.2)	3 (10.8)	1 (10)	0.445
Women	26 (43.3)	7 (31.8)	16 (57.1)	3 (30)	0.543
Cardiovascular risk factors*
No	15 (25)	6 (27.2)	9 (32.1)	0 (0)	0.201
1	13 (21.6)	3 (13,8)	9 (32.1)	1 (10)	0.321
≥2	32 (53.3)	13 (59)	10 (35.7)	9 (90)	0.102
Hypertension	30 (50)	10 (45.4)	12 (42.8)	8 (70)	0.490
Diabetes mellitus	9 (15)	3 (13.6)	5 (17.8)	1 (10)	0.803
Dyslipidemia	22 (36.6)	9 (40.9)	6 (21.4)	7 (70)	0.414
Tobacco use	5 (8.3)	2 (9)	2 (7.1)	1 (10)	0.608
Overweight	29 (48.3)	11 (50)	11 (39.2)	7 (70)	0.504
Vaccination
mRNA BNT162b2	50 (83.5)	16 (73)	24 (85.7)	10 (100)	0.801
mRNA-1263	4 (6.6)	2 (9)	2 (7.1)	0 (0)	0.304
Viral vector ChAdOxnCoV-19	4 (6.6)	2 (9)	2 (7.1)	0 (0)	0.904
Viral vector Ad26.COV2.S	2 (3.3)	2 (9)	0 (0)	0 (0)	0.609
Immediate reaction	27 (45)	11 (50)	13 (46.4)	3 (30)	0.232
Asthenia	23 (38.3)	10 (45.4)	10 (35.7)	3 (30)	0.105
Arthromyalgia	17 (28.3)	7 (31.8)	8 (28.5)	2 (20)	0.642
Fever	5 (8.3)	3 (13.6)	1 (3.5)	1 (10)	0.908
Headaches	5 (8.3)	1 (4.5)	3 (10.7)	1 (10)	0.453
Duration (d)	1 [1–3]	2 [1–3]	1 [1–3]	2 [1–2]	0.101
Time to first symptoms (d)	10 [2–30]	10 [1–30]	8 [2–21]	10 [4–21]	0.202
Time to diagnosis (d)	56 [8–196]	42 [8–196]	56 [8–147]	56 [36–112]	0.302
Clinical features
ACR/EULAR criteria
Morning stiffness	55 (83.3)	21 (95.4)	27 (96.4)	7 (70)	0.304
Hip pain	47 (78.3)	15 (68.1)	23 (82.1)	9 (90)	0.289
Absence of peripheral joint pain	31 (51.6)	13 (59)	12 (42.8)	6 (60)	0.349
Bursitis (one shoulder)	32/43 (74.4)	12/18 (66.6)	13/16 (81.2)	7/9 (77.7)	0.909
Bursitis (both shoulders)	17/43 (39.5)	4/18 (22.2)	8/16 (50)	5/9 (55.5)	0.689
Articular signs
Girdle pain	60 (100)	22 (100)	28 (100)	10 (100)	1
Distal arthralgia	24 (40)	6 (27.2)	15 (53.5)	3 (30)	0.069
Distal arthritis	15 (25)	6 (27.2)	8 (28.5)	1 (10)	0.090
Rib cage pain	5 (8.3)	2 (9)	3 (10.7)	0 (0)	0.203
Systemic signs	49 (81.6)	19 (86.3)	23 (82.1)	7 (70)	0.396
Fever	7 (11.6)	2 (9)	4 (14.2)	1 (10)	0.189
Asthenia	49 (81.6)	18 (81.8)	23 (82.1)	8 (90)	0.432
Anorexia	18 (30)	11 (50)	5 (17.8)	2 (20)	0.067
Weight loss (>10%)	6 (10)	4 (18)	1 (3.5)	1 (10)	0.347
Others	6 (10)	4 (18)	2 (7)	0 (0)	0.132
Laboratory tests
C-reactive protein (mg/L)	55 [14–224]	65.5 [17–160]	55 [15–224]	21 [14–74]	0.099
ESR (mm)	52 [5–110], n = 20	65 [5–110], n = 7	34 [30–72], n = 9	63 [22–80], n = 4	0.476
Hemoglobin level (g/dL)	11.7 [9.3–14.1]	11.3 [9.3–13.1]	11.6 [9.7–13.6]	12 [11.8–14.1]	0.998
Platelet count (G/mm^3)	389 [190–610]	369 [210–610]	399 [190–585]	392 [228–465]	0.988
PET-CT
Positive	29/32 (90.6)	10/11 (90.9)	12/14 (85.7)	7/7 (100)	0.203
Shoulder/hips	29 (90.6)	10 (90.9)	12 (85.7)	7 (100)	0.435
Sternoclavicular joints	5 (15.6)	2 (18.1)	1 (7.1)	2 (28.5)	0.190
Ischial tuberositis	10 (34.4)	6 (54.5)	2 (14.2)	2 (28.5)	0.545
Spinous process	13 (44.8)	6 (54.5)	6 (42.8)	1 (14.2)	0.234
Symphysis	3 (10.3)	2 (18.1)	1 (7.1)	0 (0)	0.541
Subtrochanteric bursae	7 (24.1)	2 (18.1)	2 (14.2)	3 (42.8)	0.345

Values are displayed as absolute number (%) or median [range], * excluding age.

PET-CT: Fluorodeoxyglucose positron emission tomography with computed tomography.

P-value: comparison between number of vaccine subgroup (pairwise comparisons).

^a1st vaccine Vs 2nd vaccine; ^b2nd vaccine Vs 3rd vaccine; ^c1st vaccine Vs 3rd vaccine.

GCA

The nationwide survey led to the identification of 35 GCA cases (14 (40%) occurring after the 1st shot, 13 (37.1%) after the 2nd shot, and 8 (22.9%) after the 3rd shot), median age of 72 years (range 50–99), 22 women (62.8%) whose characteristics are summarized in Table 3. The primary cardiovascular risk factors identified included hypertension and overweight, both mentioned in 13 (37.1%) patients. Among the patients, 31 (88.5%) were vaccinated with mRNA vaccines, while 4 (11.5%) received viral vector vaccines. The median time to the onset of the first symptoms was 7 (range 2–25) days.

Table 3. Characteristics of the whole cohort of GCA patients following COVID-19 vaccination.

Patients	All patients n = 35	1st vaccine n = 14	2nd vaccine n = 13	3rd vaccine n = 8	p-value
Demographics
Age	72 [50–99]	74 [61–99]	70.5 [50–83]	71 [60–84]	0.603
<70 years	12 (34.2)	3 (21.4)	6 (46.1)	3 (37.5)	0.189
70–80 years	15 (42.8)	6 (42.8)	5 (38.4)	4 (50)	0.067
>80 years	8 (23)	5 (35.8)	2 (15.5)	1 (12.5)	0.476
Women	22 (62.8)	10 (71.4)	9 (69.2)	3 (37.5)	0.435
Cardiovascular risk factors*
No	9 (25.7)	2 (14.2)	4 (30.7)	3 (37.5)	0.503
1	17 (48.5)	8 (57.4)	4 (30.7)	5 (62.5)	0.402
≥2	9 (25.8)	4 (28.4)	5 (38.6)	0 (0)	0.304
Hypertension	13 (37.1)	9 (64.2)	3 (23)	1 (12.5)	0.409
Diabetes mellitus	5 (14.2)	2 (14.2)	3 (23)	0 (0)	0.532
Dyslipidemia	8 (22.8)	2 (14.2)	4 (30.7)	2 (25)	0.160
Tobacco use	4 (11.4)	0 (0)	3 (23)	1 (12.5)	0.342
Overweight (kg)	13 (37.1)	6 (42.8)	6 (46.1)	1 (12.5)	0.367
Vaccination
mRNA BNT162b2	25 (71.4)	11 (78.6)	9 (69.2)	5 (62.5)	0.203
mRNA-1263	6 (17.1)	0 (0)	3 (23)	3 (37.5)	0.409
Viral vector ChAdOxnCoV-19	4 (11.5)	3 (21.4)	1 (7.8)	0 (0)	0.112
Viral vector Ad26.COV2.S	0 (0)	0 (0)	0 (0)	0 (0)	0.232
Immediate reaction	16 (45.7)	5 (35.7)	6 (46.1)	5 (62.5)	0.109
Asthenia	12 (34.2)	4 (28.6)	4 (30.7)	4 (50)	0.301
Arthromyalgia	8 (22.8)	3 (21.4)	3 (23)	2 (25)	0.102
Fever	5 (14.3)	1 (7.1)	1 (7.8)	3 (37.5)	0.100
Headaches	4 (11.5)	1 (7.1)	1 (7.8)	2 (25)	0.342
Duration (d)	1 [1–2]	1 [1–2]	1 [1–3]	1 [1–2]	0.485
Time to first symptoms (d)	7 [2–25]	7.5 [2–20]	10 [2–25]	6.5 [2–8]	0.101
Time to diagnosis (d)	42 [8–168]	42 [8–64]	56 [18–168]	42 [28–84]	0.304
Clinical features
Cephalic signs	30 (85.7)	11 (78.5)	11 (84.6)	8 (100)	0.209
Headaches	25 (71.4)	12 (85.7)	8 (61.5)	5 (62.5)	0.102
Scalp tenderness	18 (51.4)	8 (57.1)	4 (30.7)	6 (75)	0.309
Jaw claudication	13 (37.1)	8 (57.1)	2 (15.4)	3 (37.5)	0.243
Sensitive temporal artery	21 (60)	12 (85.7)	4 (30.7)	5 (62.5)	0.079
Ophthalmologic signs	7 (20)	3 (21.4)	4 (30.7)	0 (0)	0.122
Polymyalgia rheumatica	17 (48.5)	4 (28.5)	11 (84.6)	2 (25)	0.089
Extracephalic vascular signs	0 (0)	0 (0)	0 (0)	0 (0)	1
Systemic signs	30 (85.7)	9 (64.2)	13 (100)	8 (100)	0.988
Fever	14 (40)	5 (35.7)	4 (30.7)	6 (75)	0.234
Asthenia	28 (80)	7 (50)	13 (100)	8 (100)	0.098
Anorexia	18 (51.4)	6 (42.8)	7 (53.8)	5 (62.5)	0.112
Weight loss (>10%)	8 (22.8)	3 (21.4)	3 (23)	2 (25)	0.203
Laboratory tests
C-reactive protein (mg/L)	84 [5–286]	100 [5–237]	76 [8–286]	84.5 [17–106]	0.421
ESR (mm)	72 [36–120], n = 18	77 [50–120], n = 9	71 [36–100], n = 6	63 [38–63], n = 3	0.902
Hemoglobin level (g/dL)	12.1 [9.6–13.6]	12.5 [9.6–12.9]	12 [9.8–13.6]	11.9 [10.1–12.1]	0.394
Platelet count (G/mm^3)	389 [223–518]	399 [287–500]	349 [223–478]	384 [298–518]	0.809
Positive TAB	19/30 (63.3)	10/11 (90.9)	6/11 (54.5)	3/8 (37.5)	0.100
Positive TA Doppler ultrasound	4/7 (57.1)	3/4 (75)	0/1 (0)	1/2 (50)	0.598
Positive LVV imaging	18/32 (56.2)	5/11 (45.4)	8/13 (61.5)	5/8 (62.5)	0.793
Positive CTA	3/10 (30)	1/4 (25)	0/1 (0)	2/5 (40)	0.209
Positive PET-CT	15/22 (68.2)	4/7 (57.1)	8/12 (66.6)	3/3 (100)	0.932 0.234
Total number of vascular territories involved	2 [1–4]	1.5 [1–3]	2 [1–4]	3.5 [3–4]
Aorta	11/32 (34.3)	4/11 (36)	5/13 (38.4)	2/8 (25)	0.478
Subclavian arteries	1 (3.1)	2 (18)	2 (15.4)	2 (25)	0.342
Carotid arteries	5 (15.6)	1 (9)	2 (15.4)	2 (25)	0.121
Upper limb arteries	2 (6.2)	0 (0)	1 (7.7)	1 (12.5)	0.231
Ilio-femoral arteries	2 (6.2)	0 (0)	1 (7.7)	1 (12.5)	0.988
Lower limb arteries	1 (3.1)	0 (0)	1 (8.7)	0 (0)	0.433

Values are displayed as absolute number (%) or median [range], *excluding age.

CTA : computed tomography angiography; PET-CT: Fluorodeoxyglucose positron emission tomography with computed tomography; TAB: Temporal artery biopsy; LVV: Large vessel vasculitis.

P-value: comparison between number of vaccine subgroup (pairwise comparisons).

^a1st vaccine Vs 2nd vaccine; ^b2nd vaccine Vs 3rd vaccine; ^c1st vaccine Vs 3rd vaccine.

Cephalic signs were reported by 30 (85.7%) patients, among which 25 (71.4%) had headache, 18 (51.4%) had scalp tenderness, 13 (37.1%) had jaw claudication, and 21 (60%) had a sensitive temporal artery. Ophthalmologic signs were mentioned by 7 (20%) patients, including binocular diplopia (n = 3) and unilateral arteritic anterior ischemic optic neuropathy (AAION) (n = 4), with one case associated with central artery occlusion.

Associated PMR was mentioned in 17 (48.5%) patients. Systemic signs were reported by 30 (85.7%) patients. Other symptoms like pleuro-pericarditis and cough were reported in one and 3 patients, respectively. All patients, except one, exhibited a biological inflammatory syndrome, with a median CRP level of 84 mg/L (range 5–286). TAB was performed in 30 (85.7%) patients, and 19 (63.3%) of them confirmed the diagnosis of GCA. LVV was detected in 18 out of 32 (56.2%) patients, with involvement of the aorta and carotid arteries being predominant (aorta: 11 (34.3%), carotid arteries: 5 (15.6%)). No significant differences in terms of phenotype, biological markers, or PET-CT imaging were detected among the subgroups identified based on the number of vaccine shots received.

HLA study

HLADRB1 typing data were available for 15 patients with PMR and 11 patients with GCA, and these findings are summarized in Table 4. HLADRB1*04, a genetic marker associated with an increased susceptibility to PMR and GCA, was reported in 3 (20%) of PMR cases and 2 (18.1%) of GCA cases.¹

Table 4. Results of HLA DRB1 typing of 26 patients with PMR or GCA following COVID-19 vaccination.

	Diagnosis	HLA DRB1	Diagnosis	HLA DRB1
1	PMR	DRB1*13:01/*14:54	GCA	DRB1*06:03/*14:36
2	PMR	DRB1*03:01/*13:03	GCA	DRB1*12:01/*14:02/*15:01
3	PMR	DRB1*13:01/*15:01	GCA	DRB1*04:01/*12:23
4	PMR	DRB1*04:01/*11:02	GCA	DRB1*12:04/*14:28
5	PMR	DRB1*01:03/*04:01/*15:01	GCA	DRB1*03/*04
6	PMR	DRB1*04:01/*15:01	GCA	DRB1*01:02/*11:02
7	PMR	DRB1*03:01/*15:01	GCA	DRB1*04:03/*15:02
8	PMR	DRB1*04:01/*13:02	GCA	DRB1*04 :02/*15:01
9	PMR	DRB1*06:01/*12:04	GCA	DRB1*04:04/*11:01
10	PMR	DRB1*03:02/*13:02	GCA	DRB1*07:01
11	PMR	DRB1*01:04/*11:32	GCA	DRB1*07 :01/*11 :04
12	PMR	DRB1*01:03/*11:18/*15:01
13	PMR	DRB1*11 :01/*16 :01
14	PMR	DRB1*03/*12
15	PMR	DRB1*13/*14

Bold characters: HLA DRB1 typing previously associated with PMR and GCA.

Treatments and outcomes

PMR

The median follow-up period was 16 (range 12–20) months. Details of treatment and follow-up are provided in Table 5. At the initiation of treatment, the median oral corticosteroids (CS) dose was 0.31 mg/kg/day [range 0.14–1.09]. Over time, the CS doses were tapered to a median of 0.17 (range 0.05–0.33), 0.09 (range 0.02–0.25), and 0.08 (range 0.05–0.13) mg/kg/day at months 3, 6, and 12, respectively. Discontinuation of CS treatment was achieved in 15 (25.4%) patients at month 6 and 44 (74.5%) patients at month 12, with a median treatment duration of 8 (range 4–17) months. Six (10%) patients experienced a relapse with a median time after CS initiation of 6 (range 6) months. Among these, 5 (8.3%) patients received steroid-sparing treatment, including methotrexate (MTX) in 3 cases and tocilizumab (TCZ) in 2 cases.

Table 5. Treatment and outcomes in the whole cohort of patients with PMR following COVID-19 vaccination.

Patients	All patients n = 60	1st vaccine n = 22	2nd vaccine n = 28	3rd vaccine n = 10	p-value
Glucocorticoids dose (mg/kg)
Starting dose	0.31 [0.14–1,09]	0.3 [0.14–0.71]	0.33 [0.23–1.09]	0.30 [0.24–1]	0.988
At month 3	0.17 [0.05–0.33]	0.16 [0.05–0.33]	0.18 [0.06–0.27]	0.15 [0.08–0.26]	0.765
At month 6	0.09 [0.02–0.25]	0.08 [0.02–0.12]	0.1 [0.06–0.21]	0.12 [0.07–0.13]	0.654
At month 12	0.08 [0.05–0.13]	0.1 [0.09–0.13]	0.07 [0.06–0.08]	0.05 [0.04–0.05]	0.234
Outcomes
Discontinuation of glucocorticoids
At month 6	15 (25)	4 (18.1)	7 (25)	4 (40)	0.342
At month 12	44 (73.3)	19 (86.3)	19 (67.8)	6 (60)	0.435
Duration of glucocorticoids (months)	8 [4–17]	8.25 [4–17]	8 [5–16]	6 [4–13]	0.069
Relapse	6 (10)	0 (0)	4 (14.2)	2 (20)	0.102
Steroid-sparing agents	5 (8.3)	0(0)	3 (10.7)	2 (20)	0.212
MTX	3/5 (60)	0 (0)	2/3 (66.6)	1/2 (50)	0.988
TCZ	2/5 (40)	0 (0)	1/3 (33.3)	1/2 (50)	0.907
COVID-19 new vaccination	24 (40)	12 (54.5)	11 (39.2)	1 (10)	0.232
Relapse after vaccination	1 (1.6)	1 (4.5)	0 (0)	0 (0)	0.978

MTX: methotrexate; TCZ : tocilizumab.

P-value: comparison between number of vaccine subgroup (pairwise comparisons).

^a1st vaccine Vs 2nd vaccine; ^b2nd vaccine Vs 3rd vaccine; ^c1st vaccine Vs 3rd vaccine.

No significant differences in terms of therapeutic management or relapse rates were observed among the subgroups identified based on the number of vaccine shots received. Overall, 24 (40%) patients were re-challenged with mRNA vaccines after their initial diagnosis. Among them, one (1.6%) experienced a relapse 10 days after re-challenge, exhibiting a GCA phenotype with pathological confirmation via TAB, and was subsequently included in the GCA survey.

GCA

The median follow-up duration was 18 (range 18–24) months. An overview of treatment and follow-up is presented in Table 6. Treatment was initiated with a median oral CS dose of 0.7 mg/kg/day (range 0.3–1.1). Subsequently, CS doses were tapered to a median of 0.3 (range 0.08–1), 0.2 (range 0.07–0.75), 0.1 (range 0.05–0.21), and 0.1 (range 0.01–0.25) mg/kg/day at months 3, 6, 12, and 18, respectively. Discontinuation of CS treatment was achieved in 5 (14.3%) patients at month 12 and 20 (57.1%) patients at month 18, with a median treatment duration of 16 (range 6–23) months.

Table 6. Treatment and outcomes in the whole cohort of patients with GCA following COVID-19 vaccination.

Patients	All patients n = 35	1st vaccine n = 14	2nd vaccine n = 13	3rd vaccine n = 8	p-value
Glucocorticoids
Bolus	5 (14.2)	1 (7.1)	3 (23)	1 (12.5)	0.201
Starting dose (mg/kg)	0.7 [0.3–1.1]	0.7 [0.65–1]	0.7 [0.3–1.1]	0.7 [0.7–1]	0.989
At month 3	0.3 [0.08–1]	0.28 [0.18–0.57]	0.29 [0.08–1]	0.29 [0.18–0.63]	0.803
At month 6	0.2 [0.07–0.75]	0.17 [0.07–0.56]	0.18 [0.10–0.4]	0.18 [0.12–0.75]	0.766
At month 12	0.1 [0.05–0.21]	0.1 [0.04–0.15]	0.1 [0.05–0.21]	0.07 [0.07–0.21]	0.203
At month 18	0.1 [0.01–0.25]	0.04 [0.01–0.09]	0.10 [0.01–0.25]	0	0.201
Antiplatelet agent	9 (25.7)	4 (28.5)	3 (23)	2 (25)	0.809
Outcomes
Discontinuation of glucocorticoids
At month 12	5 (14.3)	1 (7.1)	2 (15.3)	2 (25)	0.203
At month 18	20 (57.1)	5 (35.7)	7 (53.8)	8 (100)	0.049^b, c
Duration of glucocorticoids (months)	16 [6–23]	19.5 [10–22]	16 [6–23]	15 [8–17]	0.069
Relapse	14 (40)	5 (35.7)	5 (38.4)	4 (50)	0.302
Steroid-sparing agents	16 (45.7)	5 (35.7)	7 (53.8)	4 (50)	0.304
MTX	5/16 (31.2)	2/5 (40)	3/7 (42.8)	0/4 (0)	0.105
TCZ	12/16 (75)	3/5 (60)	5/7 (71.4)	4/4 (100)	0.566
Others	1/16 (6,25)	0 (0)	1/7 (14.2)	0 (0)	0.876
COVID-19 new vaccination	10 (28.6)	7 (50)	3 (23)	0 (0)	0.877
Relapse after vaccination	0 (0)	0 (0)	0 (0)	0 (0)	1

MTX: methotrexate; TCZ: tocilizumab.

P-value: comparison between number of vaccine subgroup (pairwise comparisons).

^a1st vaccine Vs 2nd vaccine; ^b2nd vaccine Vs 3rd vaccine; ^c1st vaccine Vs 3rd vaccine.

The number of patients who were free of CS at month 18 was significantly higher in the group that developed the disease after receiving the 3rd shot compared to the other two groups (3rd: 8/8 (100%) vs. 2nd: 7/13 (53.8%), 1st: 5/14 (35.7%), p < .05). A relapse occurred in 14 (40%) patients, with a median time after CS initiation of 6 (range 2–12) months. Sixteen (45.7%) patients received steroid-sparing agents, including MTX in 5 cases and TCZ in 12 cases. One patient sequentially received MTX (at month 6), TCZ (at month 9), and infliximab (at month 12) due to uncontrolled disease. In total, 10 (28.6%) patients were re-vaccinated with mRNA vaccines after their diagnosis, and none of them experienced a relapse.

Literature review

A systematic literature review allowed us to identify an additional 43 cases of PMR and 7 cases of GCA, the characteristics of which are described in Supplemental Tables 1 (PMR) and 2 (GCA).^6,7,12–30 MRNA vaccines were received by 31 (72%) PMR patients and 6 (85.7%) GCA patients. The median time to the onset of the first symptoms was 8 (range 1–20) days for PMR and 2 days (range 2–10) for GCA, respectively. For PMR cases: all had hip/girdle pain associated with a biological inflammatory syndrome (median CRP level: 85 mg/L (range 29–180)). CS were initiated in 36 (83.7%) patients, with 3 (6.3%) cases also receiving MTX. Regarding GCA cases, cephalic and systemic signs were reported in 6 (85.7%) and 5 (71.4%) patients, respectively. All patients, except one, exhibited a biological inflammatory syndrome (median CRP level: 63.2 mg/L (range 8.2–272)). Four (57.1%) patients had a positive TAB, while the remaining 3 (50%) showed LVV on imaging. All patients received CS treatment, with 2 cases (40%) also receiving steroid-sparing agents (MTX in one case and TCZ in another). No data on the outcomes and follow-up were available for these patients, regardless of their pathology.

Discussion

While PMR and GCA are two well-known inflammatory diseases affecting adults over the age of 50, there have been reports of de novo cases where individuals developed symptoms shortly after receiving the COVID-19 vaccination. In 2021, a case/non-case study on the international pharmacovigilance database VigiBase, conducted after a couple of months of COVID-19 vaccination, identified an elevated odds ratio for PMR (2.3) and GCA (2.7) when compared to other drugs suggesting a potential safety signal.⁸

We initially conducted a national pharmacovigilance study encompassing all cases of PMR and GCA in individuals whose symptoms began within one month after receiving the COVID-19 vaccination. Over the study period, more than 70 million doses were administered to 25 million adults aged over 50. Regardless of the type and number of vaccine shots received, we identified 179 cases of PMR and 54 cases of GCA, resulting in a RR of 7.1 and 2.1 per 1,000,000 adults, and 2.53 and 0.73 per 1,000,000 injected doses for PMR and GCA, respectively. The elevated RR for both PMR and GCA was observed after the first dose of the ChadOx1 nCoV-19 vaccine (Oxford – AstraZeneca); however, the number of patients and injected doses was significantly lower compared to mRNA vaccines, primarily due to suspected coagulation side effects.³¹ When combining the RR of both PMR and GCA per adult and per injected dose, it raises suspicion of a lower incidence of these diseases after COVID-19 vaccination compared to the unvaccinated population, making the interpretation of pharmacovigilance studies in this context more challenging.^1,32

We provided a more comprehensive characterization of these patients through a nationwide survey. This survey identified 60 cases of PMR and 35 cases of GCA. Additionally, our literature review, which was based on one retrospective series and several case reports, enabled us to identify an additional 43 cases of PMR and 7 cases of GCA. However, the available data for these cases were limited, and the follow-up was relatively short. In our survey, it was observed that PMR and GCA were more likely to occur after the 1st or 2nd vaccine shot, with a lesser occurrence after the 3rd shot. This trend was expected, given the predominant use of mRNA vaccines, which resulted in a significantly higher number of patients receiving mRNA vaccines. Similar to previous large case series of unvaccinated PMR/GCA, patients in our study were primarily from the eighth decade of life, with 70 to 75% having cardiovascular risk factors. Notably, in contrast to the usual 2–3:1 female-to-male ratio reported in unvaccinated cohorts, patients with PMR after vaccination were predominantly male.^32–34

Aside from systemic manifestations, the phenotype of PMR and GCA cases following COVID-19 vaccination did not appear to differ from the unvaccinated PMR/GCA population and was not influenced by the number of vaccine shots received. Systemic manifestations were reported in 80% of COVID-19-vaccinated PMR/GCA cases, compared to 40 to 50% in the unvaccinated PMR/GCA population.³⁵ In addition, systemic manifestations are the most commonly reported COVID-19 vaccine-related adverse reaction suggesting a confounding factor, since we included patients who started their symptoms quickly after the immunizations.³⁶

In terms of ischemic complications, 20% of vaccinated GCA cases exhibited ophthalmologic manifestations, particularly AAION, which resembled those seen in unvaccinated GCA cases.³⁷ Importantly, although we had a limited number of vaccinated GCA patients in our study, no cerebrovascular or myocardial events were reported, that was in line with the low prevalence of 4% and 2.5% of cerebrovascular and myocardial ischemic complications reported in unvaccinated GCA, respectively.^38,39 Imaging for LVV was conducted in over 90% of vaccinated GCA cases, predominantly using PET-CT since a 2015-published meta-analysis has reported a sensitivity and specificity of 90% in detecting LVV with PET-CT.⁴⁰ Combining both PET- CT and CTA imaging, LVV was identified in more than 50% of vaccinated GCA cases, with 35% exhibiting aortitis – a proportion slightly lower than the 45 to 65% of cases detected in the unvaccinated population.⁴¹ Collectively, these data suggest a largely similar phenotype of PMR/GCA cases occurring after COVID-19 vaccination, particularly GCA-related ischemic complications and -LVV, when compared to the unvaccinated population. CS serve as the fundamental medical treatment for both PMR and GCA, although the initial dose and tapering schedule differ between these 2 diseases. In our study, the starting dose for PMR and GCA aligned with previous recommendations.^42–44 However, in PMR patients, the tapering schedule appeared to be somewhat shorter than recommended, as 75% were able to discontinue CS after 12 months, with a median duration of CS use being 8 months. During the follow-up period, 10% of vaccinated PMR cases and 40% of vaccinated GCA cases experienced relapses, which seemed slightly lower compared to unvaccinated cases. Some studies have reported relapse rates of 20–55% in PMR patients during the first year and 50% in GCA cases during the first 2 years.^45,46

In line with society recommendations, steroid-sparing agents were initiated to manage most relapses, with MTX and TCZ being the primary choices, especially TCZ in the case of GCA.^39,40 These findings suggest that vaccinated PMR and GCA cases did not exhibit a higher risk of relapse, prolonged CS use, or increased requirements for steroid-sparing treatments, indicating a medical management approach similar to that of unvaccinated cases.

The median time between the first reported symptoms of PMR/GCA and immunization was short, from 7 to 10 days suggesting a temporal association that is considered a primary contributing factor in identifying and defining environmentally associated disorders.⁴⁷ Furthermore, among 34 patients who underwent re-challenge with vaccination, one PMR patient exhibited a GCA-like phenotype 10 days after receiving a new immunization, which was subsequently confirmed pathologically, adding another noteworthy element to consider. A descriptive analysis conducted across three international databases revealed that PMR accounted for 9.2% of reported cases following immunization, with a higher likelihood of occurrence following influenza vaccination, whereas GCA constituted 3% of all reported cases.⁴⁸ However, detailed clinical data were not available. Interestingly, another study reported ten cases of GCA that occurred shortly after influenza vaccination, indicating a prevalence of 2.8% for these cases within a hospital-based recruitment context.⁴⁹

Thus far, while no specific components of COVID-19 vaccines have been strictly identified as responsible for these rare inflammatory-mediated adverse events, one hypothesis that could be considered is consistent with the concept of autoimmune/inflammatory syndrome induced by adjuvant (ASIA).⁵⁰ Although the association of HLA DRB1 haplotypes has been described as a minor criterion for diagnosing ASIA, our study found HLA DRB1*04 in 20% of PMR/GCA patients, suggesting a genetic predisposition previously documented in PMR and GCA, which may contribute to the development of both diseases.¹

Interestingly, the lipid nanoparticles used in COVID-19 mRNA vaccines have been found to induce significant inflammatory interleukin (IL)-1β/IL-6 secretion in a murine model, potentially enhancing the adjuvant immunogenicity of the vaccine.⁵¹ COVID-19 vaccines encode the SARS-CoV- 2 spike protein, which plays a crucial role in binding the virus to host cells expressing its receptor, angiotensin-converting enzyme 2 (ACE2). A recent study elegantly demonstrated that the spike protein promotes an ACE2 type 1 receptor-mediated signaling cascade, induces transcriptional regulation involving nuclear factor-kappa B (NF-κB) and AP- 1/c-FOS via Mitogen-activated protein kinases (MAPK) activation, and increases the release of IL-6, a key cytokine in the pathogenesis of PMR and GCA.⁵² Considering the well-established involvement of cellular immunity in GCA pathogenesis, a Th1 profile was also detected by day 8 after COVID-19 vaccine immunization, similar in magnitude to those observed after natural infection, which could contribute to the disease process.⁵³

This study has several limitations. Some relevant cases may have been missed in the national in-depth analysis since reports have been declarative. In addition, some clinical data were not available owing to the retrospective design and the heterogeneity of management of these patients. However, we used stringent inclusion criteria regarding the diagnosis as well as the timeline from the immunization to reduce selection and confusion bias. Finally, we were not able to perform statistical comparisons regarding the phenotype presentation, the treatment strategy and the follow-up between vaccinated and unvaccinated patients, since we did not have control-cohort group of unvaccinated PMR/GCA patients.

In conclusion, although cases of PMR or GCA following COVID-19 vaccination appear to be rare, the short time between immunization and the onset of initial symptoms suggests a temporal association. The phenotype, ischemic and vascular complications, as well as the therapeutic management of these patients, resemble those of PMR or GCA in the unvaccinated population, strongly supporting the beneficial effects of COVID-19 vaccination in the context of widespread circulation of SARS-CoV-2. While a clear causative link between immunization and the onset of the disease remains unclear, various hypotheses have been proposed.

Physicians should be aware of this vaccine-related phenomenon, and re-challenging may be considered while maintaining vigilance.

Author contributors

PAJ collected data, conducted statistical analysis and wrote the manuscript. PAJ, AM, JM, DS and GK designed the study, and wrote the manuscript. SO, SC, KHS, EL, SP, MM, BT, PEG, LT, VL, JK, DR, BB, TW, DR, AP, CC, VG, MH, MP, CW, LK, PG, TR, NB, VP, EB, AL, JH, NS, NS, JMD, BC, AP, RA, BG, PG, HL, PJW collected data. PG, SG, VG, JM conducted the pharmacovigilance study. All authors provided care for study patients. PAJ and GK are guarantors. They accept full responsibility for the work and of the conduct of the study, had access to the data, and controlled the decision to publish.

Patient consent for publication

Consent obtained directly from patient(s).

Acknowledgments

We thank the French Network of Regional Center of Pharmacovigilance for their help in the study.

Disclosure statement

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

Data availability statement

Data are available upon reasonable request.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website at https://doi.org/10.1080/21645515.2024.2334084

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.