The classic recommendation for initial treatment of temporal arteritis has been to commence with prednisone at a dosage of 40 to 80 mg/day, though 60 mg/day or above has been more commonly described than lower doses. This relatively high dose is usually maintained for 3 to 4 weeks, following which the dose is gradually reduced over many weeks to months, watching for reemergence of clinical symptoms (headache, jaw claudication, etc.) or rising erythrocyte sedimentation rate. Sometimes side effects require that the dose be reduced earlier than 4 weeks, but this must be individualized to suit each patient's special needs. Patients with polymyalgia rheumatica but no headache or other specific signs of temporal arteritis can be safely managed with 15 mg/day or less prednisone from the outset and seem not to be liable to get ischemic complications. If headache or other signs of temporal arteritis emerge during the course of polymyalgia rheumatica, however, it is important to increase the prednisone dosage to the levels recommended for initial treatment of temporal arteritis.
Temporal arteritis tends to remain active for at least 1 year, and usually longer, an average of 3 to 4 years in some series, but with a with wide range among the various studies (Kyle and Hazleman 1993), so tapering the dose of prednisone too early commonly results in recurrence of headache and other symptoms that signal the threat of ischemic complications. Recurring symptoms can usually be brought under control by reestablishing the original high dose followed by another tapering interval. This is, however, cumbersome and does not afford full protection from ischemic complications. It is better to taper slowly and avoid reactivation of the disease. It is important to note that alternate-day steroid therapy has never been considered acceptable for these patients. Return of headache on the off day suggests that the patient is not adequately protected by this regimen.
Since the introduction of high-dose corticosteroid therapy for temporal arteritis in the 1950s by Shick and colleagues (Shick et al 1950), new visual loss has only infrequently been reported in patients who had been taking the recommended high initial dose for at least 7 days. Because corticosteroid treatment has been so markedly effective in preventing complications in temporal arteritis, it will be difficult if not impossible to conduct controlled clinical trials even of promising other medications, even though the high frequency of steroid side effects would make alternative treatment highly desirable.
Delecoeuillerie and colleagues suggested that previous dosage recommendations were unnecessarily high even for initiation of therapy (Delecoeuillerie et al 1988). They correlated the clinical outcome with starting dosage of corticosteroids in 132 patients with polymyalgia rheumatica and 78 with temporal arteritis. Their findings were in accord with others that doses higher than 15 mg/day are unnecessary in polymyalgia rheumatica. As for primary temporal arteritis, they found that 5 patients initially treated with less than 20 mg/day of prednisone had no greater visual or other ischemic morbidity than 53 patients treated with standard high-dose prednisone at the outset. As might be expected, the lower dose resulted in fewer steroid side effects and complications. Over the course of treatment, which averaged 31 months, 15 patients (19%) had visual or neurologic complications. Myles and colleagues analyzed the course of 96 patients with temporal arteritis, of whom 51 also had polymyalgia rheumatica (Myles et al 1992). They tabulated the initial dosage of prednisolone and the dose being administered at the time of visual symptom onset in 6 temporal arteritis patients and 3 polymyalgia patients having complications and found no correlation between either dosage and the presence or absence of ischemic visual symptoms. Only 1 patient (with temporal arteritis) had lasting visual loss; the others all had "amaurosis fugax." These recommendations for low-dose safety were made based on the clinical course of only 5 patients in the earlier study and 1 in the later study (Myles et al 1992).
Kyle and Hazleman found that of 35 patients with temporal arteritis, all but 2 were successfully treated with 40 mg/day of prednisone as initial dose (Kyle and Hazleman 1989). After 5 days at 40 mg/day the patients were divided into a "low-dosage" group (taking 20 mg/day for 4 weeks, 15 mg/day for another 2 weeks, and 10 mg/day for another 2 weeks) and a "high-dosage" group (taking 40 mg/day for 4 weeks, 30 mg/day for 2 weeks, and 20 mg/day for another 2 weeks). The temporal arteritis cohort consisted of 18 patients with only temporal arteritis and another 17 with symptoms of both temporal arteritis and polymyalgia (35 total).
The same authors later reported on the long-term (median 60 weeks) follow-up of the same cohort of patients, tabulating relapses of both polymyalgia and temporal arteritis (Kyle and Hazleman 1993). It is notable (and atypical of other reports) that only 1 long-term follow-up patient lost vision permanently in 1 eye. The only other patient to lose vision did so early in the course and went on to die in the eighth week of treatment. Episodes of transient visual symptoms were recorded a total of 14 times, 10 in patients with symptoms of both temporal arteritis and polymyalgia and 1 in a patient with only temporal arteritis, but 3 instances were found in a patient (or patients) with polymyalgia alone. It is useful to note that most of the relapses occurred during the first year and 50% between the third and the sixth month. Only 50% of relapses occurred after reduction of the steroid dosage.
Nesher and colleagues compared outcomes in 77 temporal arteritis patients divided into 3 prednisone initial dosage groups: (1) group A started at 30 to 40 mg/day, (2) group B started at 40 to 60 mg/day and (3) group C started at doses greater than 60 mg/day (Nesher et al 1997). All 3 groups fared equally well as measured by cumulative cure rates at years 1, 2, and 3 as well as by relapse rate (though group C had more relapses during the first year than the other groups). Group A had significantly lower incidence of steroid side effects (36%) as compared with groups B (78%) and C (88%).
There have been proponents of intravenous pulse methylprednisolone therapy at doses of 1 to 2 g/day to initiate treatment in temporal arteritis in cases where visual loss has occurred in 1 or both eyes. Occasional cases have demonstrated marked visual return with this type of treatment, although a secondary goal is simply to prevent visual involvement of the second eye (Matzkin et al 1992). In a randomized, double-blind, placebo-controlled study in which 14 patients with biopsy-positive temporal arteritis received IV methylprednisolone (15 mg/kg of ideal body weight/day) and 13 other patients received IV saline for 3 consecutive days together with long-term oral prednisone starting at a dose of 40 mg/day, follow-up for a total of 78 weeks showed highly significant long-term benefit for the group that had received pulse IV corticosteroid initially. Those who had been treated initially with pulse IV corticosteroids 1) had a higher percentage at 36 through 78 weeks taking oral prednisone at a dose of less than or equal to 5 mg/day (P = 0.0003); 2) required a cumulative dose of oral prednisone, excluding the IV methylprednisolone dose, of 5636 mg compared with 7860 mg in the IV saline-treated group (P = 0.001); and 3) had a higher number of sustained remissions after discontinuation of treatment and a lower median daily dose of prednisone at 78 weeks (P = 0.0004) (Mazlumzadeh et al 2006).
Some have claimed that treatment with corticosteroids improves vision in eyes already affected by ischemic damage. This contention was not supported by Hayrey and colleagues in their survey of 114 eyes in 84 patients treated with IV pulse followed by oral corticosteroids or high-dose oral corticosteroid alone, in which only 4% showed improvement of visual acuity > 2 lines Snellen) and central visual field (Hayreh et al 2002).
There has been interest in adding immunosuppressant drugs to the treatment regimen in an effort to spare steroid dose requirements. Krall and colleagues reported that addition of methotrexate in doses between 7.5 and 12.5 mg/week controlled symptoms in 3 patients who relapsed as prednisone was tapered. Van der Veen and colleagues, however, submitted the addition of methotrexate 7.5 mg/week to a double-masked placebo-controlled trial carried out on 40 patients with polymyalgia rheumatica, 6 of whom also had temporal arteritis. When comparing values of initial remission, relapse frequency, and prednisone dose necessary to maintain remission, there was no significant difference between the group randomly selected to receive placebo and the group selected to receive methotrexate (van der Veen et al 1996). In addition, a multicenter, randomized, double-blind, placebo-controlled trial of adjuvant methotrexate for treatment of temporal arteritis could not demonstrate any advantage for the drug over placebo in addition to corticosteroids. Outcome measures included frequency of relapse and cumulative corticosteroid dose in the 2 groups (Hoffman et al 2002).
It has also been claimed that azathioprine spares steroid requirements in temporal arteritis (De Silva and Hazleman 1986). A survey of hospital-based consultants in England found that 6 of 19 (32%) were using azathioprine to lessen the steroid requirement in patients with temporal arteritis (Chakravarty et al 1994).
Intravenous cyclophosphamide pulses of 0.5 to 1.0 g given weekly for 3 weeks allowed a significant reduction in concomitant corticosteroid requirement in 4 patients during the initial treatment of temporal arteritis (de Vita et al 1992).
The effectiveness of oral cyclophosphamide in daily doses of 1.5 to 2.0 mg/kg per day along with medium- to high-dose glucocorticoid (0.3 to 1.0 mg/kg per day) was studied in 19 patients with giant cell arteritis. Patients received cyclophosphamide until clinical symptoms remitted and laboratory inflammatory markers normalized and for a maximum of 12 months. Fifteen of the patients started cyclophosphamide after the symptoms and labs proved refractory to glucocorticoids (5 patients) or after clinical symptoms returned during glucocorticoid tapering (10 patients). Complete remission of symptoms and normalization of inflammatory lab markers was the definition of efficacy for cyclophosphamide and this was achieved in 15 of the 19 patients (78.9%). Cyclophosphamide treatment was sustained for a mean of 5.2 months with discontinuation because of either inefficacy or side effects. Of the 15 patients for whom cyclophosphamide was efficacious, 13 were observed for at least 6 months after cessation of cyclophosphamide therapy, during which time methotrexate 10 to 20 mg/week was administered to 14 of the patients. Four of these 13 patients had clinical relapses; 1 relapse occurred 1 month after stopping the drug and 3 relapses occurred more than 24 months after cessation of cyclophosphamide. Cyclophosphamide was effective in 2 relapsed patients who were treated with this agent again (Quartuccio et al 2012).
Another 31 patients with giant cell arteritis who had persisting symptoms despite taking glucocorticoids and either methotrexate or azathioprine for at least 3 months, and could not reduce the corticosteroid dose below 10 mg/day prednisolone equivalent, were treated with cyclophosphamide administered either intravenously at various intervals or orally on a daily basis. Twenty-eight of the 31 patients were “responders” in that there was sustained cessation of symptoms and normalization of erythrocyte sedimentation rate and C-reactive protein. Relapses occurred in 12 patients after a median of 20.5 months after cessation of the first cyclophosphamide treatment and all responded to retreatment with the same drug. Adverse effects of cyclophosphamide included transient leucopenia, infections, and 1 case of hemorrhagic cystitis (Loock et al 2012).
In the mid-1980s reports began to appear suggesting that Dapsone could be used either in conjunction with corticosteroid or alone to treat temporal arteritis. Although this drug seems to allow dose reduction of corticosteroids, its toxicity is prohibitive. Anemia, neuropathy, toxicoderma, and agranulocytosis have all been encountered.
Infliximab, a tissue necrosis factor inhibitor, proved ineffective in reducing the corticosteroid dose required to prevent relapse in a study of 44 patients with newly diagnosed, biopsy-positive patients with temporal arteritis. The outcome was assessed after 22 weeks of treatment in 16 patients randomized to receive corticosteroids and placebo compared with 16 patients who received corticosteroids and infliximab. The authors admit that the study is too small to preclude some benefit from infliximab, but they judge that the benefit would likely be relatively minor if subjected to a larger study (Hoffman et al 2007). However, another tissue necrosis factor inhibitor, etanercept, was shown to reduce the cumulative dose of corticosteroid required to keep temporal arteritis in clinical remission over the course of a 1-year study, and a larger proportion of patients given etanercept were able to remain in remission without corticosteroid at 1 year compared with a placebo control group. The advantages of etanercept did not reach statistical significance, but the effect was considered substantial enough to warrant evaluation in a larger study (Martinez-Taboada et al 2008).
Gabriel and others at the Mayo Clinic reviewed the morbidity of long-term corticosteroid and nonsteroidal antiinflammatory drugs used to treat a cohort of 232 patients with polymyalgia rheumatica, of whom 30 also had temporal arteritis. Though the dosage used was probably lower than would be the case for a group of patients with temporal arteritis, the figures give a good indication of the significant drug-related problems for these patients. It was found that the risks for diabetes mellitus and various bone fractures were 2 to 5 times greater among the polymyalgia rheumatica patients as compared with age-matched controls. Proportional hazards modeling showed that the following 3 factors independently increased the number of adverse events: (1) higher age at diagnosis, (2) cumulative dose of prednisone greater than or equal to 1800 mg, and (3) female sex.
One of the most serious side effects of long-term corticosteroid treatment is osteoporosis. Various authors have generally recommended prophylactic treatment for osteoporosis using calcium and phosphate compounds, but there has been little proof of their effectiveness. Mulder and Struys found that pulse treatment with etidronate resulted in significantly greater bone density in a group of postmenopausal women being treated with long-term corticosteroids for temporal arteritis, as compared with a control group on comparable doses of corticosteroids alone (Mulder and Struys 1994).
The erythrocyte sedimentation rate is the most common laboratory parameter to follow during follow-up to signal relapses. Kyle and Hazleman found that the erythrocyte sedimentation rate correlated better than C-reactive protein with relapses, and that only "a minimal increase in alpha 1-antichymotrypsin was seen in 27% of clinical relapses" (Kyle and Hazleman 1993). Pountain and colleagues later reported that alpha 1-antichymotrypsin levels remained elevated during treatment for up to 18 months, well after the erythrocyte sedimentation rate and the C-reactive protein had normalized; a significant secondary rise in erythrocyte sedimentation rate or C-reactive protein did not accompany relapses. They found that if the alpha 1-antichymotrypsin was less than or equal to 0.7 g/L at 18 months follow-up, there was a significantly reduced risk of subsequent relapse (P=-0.006) (Pountain et al 1994).
Studies have identified some other markers that respond in a sensitive way to successful treatment of temporal arteritis. Johansen and colleagues reported on YKL-40, a mammalian member of the family 18 glycosyl hydrolases, which is secreted by activated macrophages. They found that serum YKL-40 levels were increased over controls patients with temporal arteritis, and the levels returned to normal during prednisolone treatment. The serum levels were not elevated in patients with polymyalgia rheumatica (Johansen et al 1999). Weyand and colleagues studied 25 patients with biopsy proven temporal arteritis prospectively and demonstrated plasma interleuken-6 (IL-6) to be a much more sensitive measure of disease activity than erythrocyte sedimentation rate. These authors also found that the IL-6 levels did not return to normal when all clinical indicators suggested favorable response; they suggested that incomplete suppression of arterial inflammation in most patients puts them at risk for progressive vascular disease (Weyand et al 2000).
The effects of aspirin and corticosteroids (dexamethasone) were studied in severe combined immunodeficiency mice into which inflamed temporal arteries were engrafted. Corticosteroids repressed nuclear factor kappaB, but aspirin was much more effective in suppressing interferon gamma production by T-cell clones, suggesting that there may be synergistic relation between the 2 agents in treating temporal arteritis (Weyand et al 2002). This has yet to be tested clinically.