Article Text

Research paper
Stroke associated with giant cell arteritis: a population-based study
  1. Maxime Samson1,2,3,
  2. Agnès Jacquin4,
  3. Sylvain Audia1,2,3,
  4. Benoit Daubail4,
  5. Hervé Devilliers5,
  6. Tony Petrella6,
  7. Laurent Martin6,
  8. Jérôme Durier4,
  9. Jean-François Besancenot5,
  10. Bernard Lorcerie1,
  11. Maurice Giroud4,
  12. Bernard Bonnotte1,2,3,
  13. Yannick Béjot4
  1. 1Service de Médecine Interne et Immunologie Clinique, CHU de Dijon, Dijon, France
  2. 2INSERM, UMR1098, Besançon Cedex, France
  3. 3Faculté de Médecine, Université de Bourgogne, Dijon, France
  4. 4Dijon Stroke Registry, EA4184, Department of Neurology, University Hospital and Medical School of Dijon, University of Burgundy, France
  5. 5Service de Médecine Interne et maladies systémiques, CHU de Dijon, Dijon, France
  6. 6Service d'anatomie et cytologie pathologiques, CHU de Dijon, Dijon, France
  1. Correspondence to Dr Yannick Béjot, Dijon Stroke Registry, Department of neurology, CHU, 3 Rue du Faubourg Raines, Dijon 21000, France; ybejot{at}yahoo.fr

Abstract

Background Giant cell arteritis (GCA) is the most common vasculitis in people ≥50 years and can be associated with stroke. We aimed to evaluate the epidemiology and characteristics of stroke in patients with GCA.

Methods All patients with a biopsy-proven diagnosis of GCA were identified among residents of the city of Dijon, France (152 000 inhabitants), between 2001 and 2012 using a prospective database. Among these, patients who suffered from stroke were retrieved by crossing data from the population-based Dijon Stroke Registry. Demographics and clinical features were recorded. We considered that the stroke was GCA-related if the stroke revealed GCA or occurred between the onset of symptoms and 4 weeks after the start of treatment.

Results Among the 57 biopsy-proven patients with GCA (incidence rate 10.9/100 000/year in individuals ≥50 years), 4 (7.0%) experienced a GCA-related stroke. Three were men and all had ≥2 vascular risk factors and were ≥80 years. The stroke was vertebrobasilar for 3/4 patients and undetermined for the remaining one. The incidence rate of GCA-related stroke in patients ≥50 years was 0.76/100 000/year (95% CI 0 to 2.47), 1.36/100 000/year in men (95% CI 0 to 3.63) and 0.33/100 000/year (95% CI 0 to 1.45) in women.

Conclusions This population-based study demonstrated that GCA-related stroke essentially affects the vertebrobasilar territory and mainly occurs in old men with associated vascular risk factors. Although rare, GCA symptoms must be searched for in elderly patients with stroke, and optimal vascular prevention must be conducted carefully in patients with GCA with a high vascular risk before initiating GCA treatment.

  • Stroke
  • Epidemiology
  • Rheumatology
  • Vasculitis

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Giant cell arteritis (GCA) is a granulomatous large vessel vasculitis that usually affects the aorta and/or its major branches, especially the branches of the carotid arteries. Due to the progressive aging of the population in Western countries, GCA has become the most common vasculitis in persons over 50 years of age. Peak incidence is observed in those aged between 70 and 80 years.1–3 Histopathological lesions are observed in all layers of the artery and lead to segmental and focal panarteritis with a polymorphic cell infiltrate that includes T cells, macrophages and multinucleated giant cells, a fragmented internal elastic lamina and intimal hyperplasia.4 ,5 The pathophysiology of GCA is complex and not fully understood, but knowledge about the disease is improving.6–10 In 27–56% of cases, GCA is associated with polymyalgia rheumatica (PMR), which has a common pathogenesis.1 ,2 ,9

In both PMR and GCA, systemic symptoms (fever, asthenia, anorexia and weight loss) are the consequence of chronic inflammation. However, the typical features of GCA result from the arterial topography of the inflammatory process, which involves vessel scarring and narrowing and eventually occlusion or severe stenosis. Vascular remodelling is responsible for ischaemic symptoms such as headache, scalp tenderness, jaw claudication, visual loss, scalp or tongue necrosis and also central nervous system ischaemic complications. Visual loss is usually caused by acute ischaemia of the optic nerve (AION), which is one of the most severe complications of GCA, occurring in about 30% of patients and leading to permanent loss of vision in 15% of cases.1 ,11 ,12 Patients with GCA may also experience a stroke in the carotid or vertebrobasilar territories.1 ,13 ,14 It usually occurs in the period of active disease and mainly affects the vertebrobasilar territory.1 ,15 ,16 This topography of stroke is more common in patients with active GCA than in the general population of the same age.16 Nevertheless, stroke remains an uncommon complication of GCA. It was observed in 2.8% of a large series of 287 biopsy-proven patients with GCA between the onset of symptoms and 4 weeks after the beginning of the treatment.15 However, no population-based study is currently available to determine the exact incidence of stroke associated with GCA.

Therefore, based on data from the Dijon Stroke Registry, we aimed to evaluate the epidemiology and characteristics of stroke in patients with GCA.

Methods

Study population and case ascertainment

So as to identify overall cases of stroke associated with GCA in a geographically defined area, this population-based study relied on crossing data prospectively obtained from two different databases.

Ascertainment of cases of GCA

Cases of GCA were identified using a database that collected all patients who underwent a temporal artery biopsy (TAB) between 1 January 2001 and 31 December 2011 at the University Hospital of Dijon, France, and at one private practice of the city. These two centres are responsible for the management of all TABs performed in patients living in Dijon (2007 census: 151 543 inhabitants). TABs are routinely performed in all patients with clinical manifestations of GCA. The TAB was taken from the side where symptoms were predominant. When possible, the TAB was guided by Doppler sonography of the temporal artery. When clinical symptoms suggested GCA but the first TAB was negative, a second TAB could be performed but was not done in every case. Biopsies were also considered for patients with isolated PMR, even without specific manifestations of GCA, especially if general symptoms were severe and/or erythrocyte sedimentation rate (ESR) >50 mm/h. The two pathology departments used a similar protocol for microscopic examination of TABs, by dividing the artery, when sufficiently long, into several segments of 2 mm, paraffin embedding and transverse cutting of 3–5 mm thick sections stained with H&E. TABs were then analysed by expert pathologists (LM and TP).

GCA was diagnosed in accordance with the American College of Rheumatology (ACR) criteria.17 Patients included in this study were considered to have biopsy-proven GCA when the TAB showed a pathology compatible with the diagnosis of GCA: mononuclear cell infiltration of the arterial wall, with or without the presence of granulomas and/or multinucleated giant cells.2

Ascertainment of stroke cases

All cases of stroke (including ischaemic stroke, spontaneous intracerebral haemorrhage and undetermined stroke) defined according to WHO recommendations18 were identified among residents of the city of Dijon using the Dijon Stroke Registry. This population-based registry complies with the criteria recommended for the running of ‘ideal’ stroke incidence studies,19 ,20 and case-collection procedures have been described elsewhere.21–23 Briefly, so as to ensure the exhaustiveness of case ascertainment, multiple overlapping sources of information were used to identify fatal and non-fatal stroke in hospitalised and non-hospitalised patients: (1) a review of medical records from the emergency rooms, and all the clinical and radiological departments of Dijon University Hospital, with a diagnosis of stroke made by one of the neurologists of the department of neurology, where the Stroke Registry is located. (2) A review of medical records from the emergency rooms and all of the clinical departments of the three private hospitals of the city and its suburbs, with diagnosis made by neurologists working in these establishments. (3) A review of computerised hospital diagnostic codes of Dijon University Hospital. The International Classification of Diseases, tenth revision (ICD-10), was used. The following codes are initially searched for: I60 (subarachnoid haemorrhage), I61 (intracerebral haemorrhage), I62 (non-traumatic intracranial haemorrhage), I63 (ischaemic stroke), I64 (non-determined stroke), G45 (vascular syndromes), G46 (transient ischaemic attack) and G81 (hemiplegia). Study investigators then consulted the medical records of identified patients to confirm or not the reported diagnosis or to reclassify the patients if a misclassification was noted. (4) A review of computerised hospital diagnostic codes of the private hospitals with the same procedure as described above. (5) Collaboration with the general practitioners to identify patients with stroke managed at home or in nursing homes, with the diagnosis assessed by public or private neurologists from outpatient clinics. (6) A review of the medical records of patients identified from a computer-generated list of all requests for imaging to the private radiological and Doppler ultrasound centres of the city and its suburbs. (7) Finally, regular checking of death certificates obtained from the local Social Security Bureau that is responsible for the registration of deaths in the community, particularly fatal strokes outside hospital. The quality and the completeness of the registry are certified every 4 years by an audit from the National Institute for Health and Medical Research and the National Public Health Institute. Ischaemic strokes were classified according to the trial of ORG 10172 in acute stroke treatment (TOAST) classification.

So as to identify potential cases of stroke preceding the diagnosis of GCA or occurring soon after, the Dijon Stroke Registry database was searched for stroke cases identified among patients with GCA between 1 January 1985 and 28 January 2012.

Definition of GCA-related stroke

We considered that the stroke was CGA-related if it revealed GCA or if a stroke occurred between the onset of GCA symptoms and 4 weeks after the beginning of the treatment.15

Data collected

The following vascular risk factors were systematically collected21–23: hypertension (high blood pressure noted in a patient's medical history or patients under antihypertensive treatment), diabetes mellitus (glucose level ≥7.8 mmol/L reported in the medical record or patients on insulin or oral hypoglycaemic agents), hypercholesterolaemia (total cholesterol level ≥5.7 mmol/L reported in the medical history or patients treated with lipid-lowering therapy), atrial fibrillation, history of coronary heart disease, heart failure, peripheral artery disease (PAD), alcohol intake (≥3 units a day in men and ≥2 in women), smoking, cancer and previous stroke or transient ischemic attack. Treatments prior to the stroke including antiplatelet agents, anticoagulants, antihypertensive treatments and statins were recorded. Clinical features at stroke onset were also reported. Data about the clinical characteristics of GCA and its treatment and outcome were retrospectively reviewed using medical charts.

Statistical analysis

Proportions were compared using Fisher’s exact test. Incidence rates were calculated using the average population of Dijon as denominators. These rates were age and sex adjusted to the 2006 European population using the direct method.24 CIs for a Poisson distribution were calculated. p Values <0.05 were considered statistically significant. STATA V.10.0 software (StataCorp LP, College Station, Texas, USA) was used for statistical analyses.

Ethics

The Dijon Stroke Registry was approved by the National Ethics Committee and the French Institute for Public Health Surveillance.

Results

Between 1 January 2001 and 31 December 2011, a biopsy-proven diagnosis of GCA was made in 57 consecutive patients (39 females and 18 males) living in Dijon. All these patients met the ACR criteria. The corresponding annual incidence of GCA was 10.9/100 000/year (95% CI 4.4 to 17.3) in individuals aged ≥50 years old and 13.4/100 000/year (95% CI 6.3 to 20.6) in those aged ≥55 years old.

Nine of these patients were identified in the Dijon Stroke Registry as having suffered from a stroke between 1 January 1985 and 28 January 2012. For three of these patients, the stroke occurred before the first symptoms of GCA (5, 15 and 113 months before), and for two, the stroke occurred >4 weeks after the start of treatment for GCA (24 and 38 months later). The characteristics of these patients for whom stroke was not considered to be GCA-related are shown in online supplementary table S1. Of note, for patient 2, who had a stroke only 5.1 months before GCA was diagnosed, no GCA-related symptoms were observed at stroke onset or during the acute stage. Apart from one patient for whom data were lacking, none of these patients had a stroke involving the vertebral territory, 4 were women and 4 had hypertension in their medical history (see online supplementary table S1).

The remaining four patients (7.0% of patients with GCA living in the city of Dijon) had thus suffered from GCA-related stroke according to our criteria (figure 1). The characteristics of these four patients are summarised in table 1. One patient had stroke 2 weeks after glucocorticoids had been started. In this patient, heparin was also given at the time glucocorticoids were started because of central retinal artery occlusion and AION. Heparin was administered for 10 days and then stopped and replaced with aspirin (100 mg/day) 3 days before occurrence of the stroke, which was related to small vessel occlusion (TOAST 3). Of note, at the time of the stroke in this patient, diabetes was well controlled even though the patient was being treated with steroids. For the other three patients, GCA was revealed by the stroke. In patient 3, GCA symptoms (weakness, anorexia, weight loss and acute phase syndrome) had started 1 month before the stroke, making clinicians to perform a TAB that subsequently revealed GCA.

Table 1

Characteristics of the four patients who had a GCA-related stroke among the 57 biopsy-proven patients with GCA

Figure 1

Flow chart of the study. *In Dijon between 1 January 2001 and 31 December 2011. **In people ≥50 years old living in Dijon between 1 January 2001 and 28 January 2012. GCA, giant cell arteritis.

All of the patients had at least two risk factors in addition to age for atherosclerosis. All suffered from arterial hypertension, which was already treated before stroke onset in all the cases. Of note, 3/4 cases (75%) were male. In the biopsy-proven GCA cohort of patients, 3/18 (16.7%) males had stroke versus 1/39 (2.6%) females (p=0.088).

The topography of stroke was vertebrobasilar in 3/4 (75%) cases. In patient 2, no CT scan was performed but symptoms could be compatible with a massive or brainstem stroke. Of note, this patient was already receiving an effective dose of heparin for atrial fibrillation before GCA was confirmed and the stroke occurred (table 1).

Between 1 January 2001 and 28 January 2012, 2305 strokes that occurred in people ≥50 years old were registered in the Dijon Stroke Registry, including 1966 ischaemic, 285 haemorrhagic, and 54 undetermined strokes. Among these, four (0.17%) were directly related to GCA. The crude incidence rate of stroke associated with GCA in patients aged ≥50 years old was 0.76/100 000/year (95% CI 0 to 2.47). Corresponding figures were 1.36/100 000/year in men (95% CI 0 to 3.63) and 0.33/100 000/year (95% CI 0 to 1.45) in women. The European standardised rate was 0.57/100 000/year (95% CI 0 to 2.05).

Discussion

To our knowledge, this report is the first population-based study that assessed the frequency of stroke among patients with GCA and provided the incidence rate of stroke associated with GCA.

Herein, we confirmed that stroke is a rare complication of GCA, occurring in 7.0% of biopsy-proven GCA. This is consistent with other studies, in which the authors reported cerebrovascular complications in 6.1–7.2% of patients with GCA.13 ,25 A lower frequency of stroke was reported by Gonzalez-Gay et al,15 who reported that 8 of their 287 (2.8%) biopsy-proven patients with GCA experienced a stroke between the onset of GCA symptoms and 4 weeks after the beginning of glucocorticoids. Because of the small absolute number of recorded events, direct comparisons between series are difficult. However, it could be hypothesised that we observed more cases of stroke in our study because of the prospective population-based setting, which allowed the assessment of all cases of hospitalised and non-hospitalised patients with stroke. This design is the main strength of our study and ensures the accuracy of data collection.

The risk factors for developing stroke during GCA have already been described.15 Men are more at risk than women of developing stroke. We observed the same trend in our work, though the difference failed to reach the level of significance because the number of events was too small. Other risk factors for stroke during GCA are a history of visual loss, smoking, arterial hypertension and a high haemoglobin level.15 It is worth mentioning that all of the patients who had a GCA-related stroke in our study had ≥2 major vascular risk factors.

For three of our patients, the clinical symptoms clearly suggested posterior stroke involving the vertebrobasilar territory. For the remaining patient, because the clinical presentation was severe and rapidly led to death, no brain CT scan was performed and it was not possible to determine the exact topography of the lesion, even though the clinical symptoms were compatible with a massive or a brainstem stroke, which could have been ischaemic or haemorrhagic as this patient was being treated with heparin for atrial fibrillation. Of note, apart for one patient for whom data were lacking, strokes that were not considered GCA-related did not involve the vertebrobasilar territory, but were related to multiple lacunes or carotid atherosclerosis, and essentially occurred in women. This finding is consistent with the fact that stroke predominantly affects women and that atherosclerosis and small vessel disease are the leading cause when all patients aged >50 years old included in the Dijon Stroke Registry are considered.26 Even though we cannot exclude the possibility that the four GCA-related strokes we recorded in this study could have resulted from thromboembolic or cardioembolic diseases, the epidemiological differences we recorded between GCA-related and GCA-unrelated strokes argue for a specific link between GCA and strokes and suggest that the vascular inflammation triggered by GCA could synergise with other cardiovascular risk factors to induce strokes in the vertebrobasilar territory. Consistent with this hypothesis, the high frequency of vertebrobasilar territory involvement in patients with GCA-related stroke has already been reported in the literature.14 ,15 Interestingly enough, Pfadenhauer et al evaluated by colour duplex sonography the frequency of vertebrobasilar involvement in 93 prospectively followed patients with GCA. Vertebrobasilar ischaemia was detected in 4.3% of these patients, being associated with proximal vertebral artery occlusive disease in all of them and hypoechogenic mural thickening of the proximal segments V0/V1 of the vertebral artery in half of them.27 This study underlines how helpful could be the use of colour duplex sonography for GCA diagnosis in case of stroke involving the vertebral arteries. Furthermore, it suggests the potential interest of US scan to identify patients at risk of stroke during GCA.

The major strength of our study was the population-based setting, which allowed the exhaustive and prospective assessment of all cases of stroke occurring in a geographically defined area. Our study design made it possible to determine the incidence of GCA-related stroke, which appeared to be low. GCA-related stroke was found to affect fewer than eight individuals aged ≥50 years old per million per year. However, several limitations must be acknowledged. Even though our results and those from others suggest that GCA-related strokes have a typical topographic distribution,14 ,15 it is impossible to confirm this hypothesis because of the small absolute number of events we recorded in the present study. First, although all TABs performed in patients living in Côte d'Or were retrieved since only two pathology labs performed such analyses, it cannot be excluded that some patients refused the biopsy or that some TABs were too small or not performed in cases of isolated PMR. Second, the sensitivity of TAB ranges from ∼70% to >90%, and GCA can be diagnosed on clinical grounds alone in the absence of histological evidence, even if biopsy confirmation of GCA is important to prevent unnecessary glucocorticoid therapy over long periods.1 ,17 As a result, the true incidence of GCA-related stroke could have been underestimated in our study, but, like others,15 ,28 we chose to focus on biopsy-proven patients with GCA in order to analyse patients for whom there was no doubt concerning the diagnosis. Of note, the annual incidence rate of GCA observed in our study was consistent with the findings from a French study that reported a prevalence of both biopsy-proven and not proven GCA of 9.4/100 000/year in individuals aged ≥55 years old,29 suggesting that our case-collection procedure limited to biopsy-proven GCA certainly did not influence our results excessively. Even though a gradient between northern and southern countries concerning the incidence of GCA has been reported,30 thus making it difficult comparisons, a Danish study evaluated the proportion of biopsy-proven and not proven GCA cases, and showed that the incidence rate for GCA in the population >50 years was 20.4/100 000 overall, whereas it was 15.1/100 000 when only considering biopsy-proven GCA.31

This population-based study demonstrated that GCA-related stroke essentially affects the vertebrobasilar territory and mainly occurs in old men with associated vascular risk factors. Although rare, GCA symptoms must be searched for in elderly patients with stroke, and optimal vascular prevention must be conducted carefully in patients with GCA with a high vascular risk.

Acknowledgments

We thank Philip Bastable for editorial assistance.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

Footnotes

  • Contributors All authors contributed to the article by drafting/revising the manuscript for content, collecting and analysing data, and approving the final manuscript. MS, MG, BB and YB additionally contributed to the conception and study design. MS and YB also contributed to statistical analysis.

  • Funding The Dijon Stroke Registry is supported by the French Institute for Public Health Surveillance (InVS) and INSERM. The study sponsors had no role in either the study design or the collection, analysis and interpretation data, the writing of the report or in the decision to submit the paper for publication.

  • Competing interests None.

  • Ethics approval The Dijon Stroke Registry was approved by the Comité National des Registres (French National Committee of Registers) and the InVS (French Institute for Public Health Surveillance). Authorisation of the CNIL (Commission Nationale de l'Informatique et des Libertés; National Commission for the Protection of the Privacy of Electronic Data) was obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.