Left untreated, temporal arteritis (TA) frequently results in blindness. Treatment requires immediate high dose steroids. Risk factors for initial and progressive visual loss, despite appropriate treatment, have not been extensively characterised, making selection of patients at risk for progressive visual loss difficult.

We retrospectively reviewed clinical findings in patients with biopsy proven TA in order to assess risk factors for visual loss during the first days of steroid therapy.

METHODS

A retrospective chart review of 341 patients with suspected TA undergoing temporal artery biopsy during 15 consecutive years at the National Hospital for Neurology and Neurosurgery, London, UK, was performed.

Biopsies were performed on the side of predominant symptoms. Biopsies were deemed positive for TA if the histological specimen demonstrated arteritis, characterised by mononuclear cell arterial wall infiltration and interruption of the internal lamina elastica. Additional evidence included media degeneration or intima thickening.1 Visual acuity was expressed as a decimal (20/20 = 1.00; finger counting = 0.012; hand motion = 0.006; light perception = 0.001; no light perception = 0).2 In order to compare steroids among patients, the doses in patients receiving a steroid other than hydrocortisone were converted according to the following dosing scheme: 1 mg dexamethasone = 30 mg hydrocortisone; 1 mg methylprednisolone = 5 mg hydrocortisone, 1 mg prednisone = 4 mg hydrocortisone.

SPSS10.0 was used for statistical analysis. A p value of <0.05 was considered statistically significant.

RESULTS

Ninety-three (27%) out of 341 biopsy results confirmed TA. Ninety patients (67 females; mean age 74.6 (SD 7.8); range 59–93) were included. The remaining three notes were lost.

Onset of symptoms prior to admission ranged from 2 days to 7 years (median 125 days). It took a median of 8 days (0–242 days) from referral until admission. Median duration of hospital stay was 11 days (0–53). Ninety-one per cent of biopsies were performed within 5 days of admission; 71% were biopsied within 5 days after starting steroids. Five patients underwent temporal artery biopsy prior to initiation of steroids. All 90 biopsies were suggestive of TA. Fifty-six (62%) biopsies showed giant cells.

Eighty-nine patients received corticosteroids. One patient presented 3 weeks after blindness had occurred and was not treated. Thirty-three patients initially received intravenous corticosteroid treatment (hydrocortisone, methylprednisolone or prednisolone). Fifty-six patients were initially treated with prednisone by mouth. Doses were converted to hydrocortisone strength for comparison. The average initial corticosteroid dose consisted of 495 mg of hydrocortisone (range 0–5000). The average time until first dose taper was 21 days. In 31 patients, initial corticosteroid dose was increased in order to control symptoms completely.

Twenty-one (23%) patients showed progressive visual symptoms despite corticosteroid therapy. A total of 20 patients suffered visual loss. Thirteen patients suffered persistent loss of acuity or field loss and one additional patient presented with a hemiparesis and diplopia (table 1). Deterioration occurred usually within the first days of treatment. Seven additional patients had transient visual impairment.

Of the seven patients with transient visual impairment despite corticosteroids, four patients had transient visual loss, one patient had an episode of left sided weakness and left visual loss, one patient complained about visual teichopsia and one patient developed disc swelling. All symptoms resolved after an increase in steroid dose.

The presence of disc swelling, history of hypertension and older age were significantly associated with visual loss on presentation. A trend of higher odds for initial visual loss was also seen in patients with polymyalgia, diabetes and jaw claudication (table 2). Patients with visual loss on presentation had a higher systolic blood pressure on presentation (table 3).

Risk factors for progressive visual loss included older age, elevated C reactive protein (CRP) and disc swelling (tables 4 and 5). Fourteen patients with early progressive visual loss despite corticosteroid therapy were, on average, 79.4 years old. Seventy-six patients without progressive visual loss were, on average, 73.7 years old and therefore significantly younger (p<0.05, t test value 2.2). The CRP of patients with progressive visual loss was, on average, significantly higher (57.4; n = 7) than that of patients without visual loss (22.1; n = 12; p<0.02, t test value −3.0). Other factors significantly (p<0.05) associated with progressive visual loss were disc swelling (odds ratio (OR) 5.3 (95% CI 1.4 to 20.7)) and administration of steroids intravenously (OR 5.6 (95% CI 1.6 to 19.9)) (tables 4 and 5). Men had reduced odds of visual loss progression (OR 0.24 (95% CI 0.05 to 1.2)), approaching statistical significance (p = 0.06) (tables 4 and 5).

Median follow-up duration was 7 months (range 11 days to 16 years). Two year follow-up was available for 28 patients. Four patients without previous visual symptoms had visual loss (unilateral permanent visual loss in two and unilateral temporary visual loss in two). Visual loss occurred, on average, 24 months after initial presentation. All four patients were on a tapering dose of oral prednisone at that time (on average 14 mg daily).

DISCUSSION

Continuous visual symptoms despite steroid therapy were seen in 23% of patients, and 16% suffered visual deterioration during therapy. Risk factors for visual loss on presentation were disc swelling and hypertension. Risk factors for progressive visual loss included older age, elevated CRP and disc swelling.

Fifty-eight patients (3.1%) with visual loss after initiation of corticosteroid therapy were found among 1296 patients with TA in the literature (table 6) ranging from 0%3 to 38.9% in prospective series.4 In a meta-analysis of 39 retrospective4^–24 and prospective studies (table 6), we found a highly significant correlation (Pearson’s correlation coefficient 0.604; p<0.0001) between the percentage of patients with visual loss on presentation and visual loss under corticosteroid therapy (fig 1). Visual loss on presentation may therefore predict visual deterioration under corticosteroid therapy, as also seen in our patients.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Visual deterioration in relationship to visual loss on presentation, in per cent, based on retrospective and prospective series with complete data from the literature.

Visual deterioration occurs in two peaks. The first peak manifests as progression of the ongoing flare on an unchanged steroid dose, typically during the first 6 days.25 The second peak occurs after weeks or months of tapering treatment. Relapses increase with reduction of corticosteroid therapy and were seen in 19% of patients within 1 year.26

Reasons for progression of visual loss despite treatment may include hypoperfusion of the optic disc, treatment delay, inadequate steroid dose, quick taper or hypercoagulability with retinal artery infarction, possibly due to steroid therapy. Continuation of arteritis despite adequate corticosteroid dose may be considered part of the spectrum of TA or may even be a separate disease entity.

Differential diagnoses mimicking TA include systemic lupus erythematodes, Sjögren’s syndrome, rheumatoid arthritits, Behcet’s disease, antiphospholipid antibody syndrome, polyarteritis nodosa, Churg–Strauss syndrome, Wegener’s granulomatosis and other rheumatic conditions presenting with granulomatous vasculitis such as Takyasu arteritis.27 Sarcoid, primary angiitis of the central nervous system, non-arteritic AION, neoplastic conditions as well as viral infections (varicella zoster, human parvo virus B19, human herpes virus 6, herpes simplex) and nocardiosis should also be considered.

Risk factors for initial visual loss include transient visual ischaemic symptoms, increased platelet count, jaw claudication and HLA-DRB1 phenotype.20 28 Constitutional symptoms and elevated liver enzymes are associated with a lower risk of visual loss.28 29 Risk factors for permanent visual loss include amaurosis fugax and cerebrovascular accidents.20

Risk factors for progressive visual loss may include occlusive strokes, possibly due to steroid therapy itself.30 Late recurrence of visual loss was associated with female sex, older age, worse initial visual acuity, oral (as compared to intravenous) initial steroid treatment and higher erythrocyte sedimentation rate.24 HLA DRB1 alleles were also associated with progressive symptoms.31

Intravenous or high dose oral corticosteroids remain the standard of care for patients at risk for visual loss.25 A retrospective review of 166 patients demonstrated better outcome in patients on low dose aspirin at the time of symptom onset.32 Aspirin may therefore decrease the rate of visual loss and strokes in patients with TA. Further research is on the way to determine effectiveness. Some authors also use heparin, in particular in patients with progressive visual loss.25 Additional steroid sparing agents during the long term treatment period have been tried but no positive randomised placebo controlled prospective trials are available.

Limitations of our study include the retrospective design and referral bias at a tertiary treatment centre. Our literature review attempted to compensate for these variations by comparing our data with previous studies in different settings.

The 1990 American College of Rheumatology (ACR) Criteria for the Classification of Giant Cell (Temporal) Arteritis1 were not applied to all 341 patients seen in our clinic because of the retrospective study design. We limited our study population to the gold standard of diagnosis prior to establishment of these criteria, a positive temporal artery biopsy. Nevertheless, all included patients met the 1990 ACR criteria. By making the positive biopsy a prerequisite, we likely applied more sensitive inclusion criteria. Stricter inclusion criteria may also have led to a selection bias towards more active cases. Our data are therefore valid for patients over 50 years old and in whom the biopsy is positive and at least one additional diagnostic ACR criterion is present. A prospective study on progressive visual loss according to ACR standards is needed.

CONCLUSION

Progression of visual loss despite steroid therapy occurs in a significant minority of patients with TA. In most patients, deterioration occurs within the first 3 days after initiation of steroid therapy. Individual risk anticipating treatment strategies might improve visual prognosis in TA.