In a recent issue of Arthritis Care & Research, Salvarani and Hunder described the characteristics of 9 patients with temporal arteritis who had an erythrocyte sedimentation rate (ESR) < 40 mm/hour before treatment (1). They stated that none of the patients developed severe visual complications and all had a favorable outcome. We would like to present our personal experience based on a large series of patients with temporal arteritis. Our results support some of those presented by the authors, and may provide additional insight on the inflammatory response in such variant cases of vasculitis because we systematically measured the acute phase serologic response proteins in patients with temporal arteritis, regardless of the ESR value.
Since 1976, we diagnosed and followed in our department 216 patients with temporal arteritis, of whom 207 (173 biopsy-verified) had a pretreatment measurement of ESR by the Westergren method. Twelve patients (5.8%) were found to have an ESR ≤ 45 mm/hour (median 32 mm/hour) at diagnosis (Table 1). Temporal artery biopsy demonstrated giant cell arteritis in all but 2 patients, and none had other conditions that may affect ESR values. The mean delay to diagnosis from the vasculitis onset was 34 ± 23 days in 11 patients with prominent cephalic symptoms, whereas the vasculitis was later revealed by the occurrence of a subclavian artery occlusion and upper limb ischemia in the 12th patient.
|C-reactive protein, mg/l||5||2||3||5||2||5||31||52||28||29||64||20|
|Platelet count, × 109/liter||305||267||333||437||178||378||228||561||371||416||276||301|
The acute phase serologic response was studied further in all cases and an inflammatory response was demonstrated in half of these. We also analyzed, using the Spearman's rank correlation, the correlation between ESR and acute phase reactant protein levels, blood hemoglobin value, and the platelet count in 83 patients who had had all these measurements performed before glucocorticoid treatment. A correlation was found between ESR and hemoglobin (inverse correlation, R = 0.49, P < 0.0001), C-reactive protein (R = 0.37, P = 0.002) levels and the platelet count (R = 0.29, P = 0.01), but not haptoglobin, fibrinogen, and orosomucoid levels. We did not find any correlation between ESR and other parameters in patients with an ESR ≤ 45 mm/hour, possibly because of the small sample size.
It should be noted that 2 of 6 patients with low acute phase serologic response developed permanent blindness from acute ischemic optic neuropathy, which was irreversible in one. One patient died of treatment-induced gastrointestinal bleeding, 2 are currently receiving glucocorticoid therapy, and the other 9 recovered from giant cell arteritis after a median duration of glucocorticoid therapy of 21.2 months and a median followup duration of 52 months.
We therefore confirm that a low ESR in active temporal arteritis is not a rare event, being observed in approximately 6% of our cases. In our series, most patients with a low ESR presented with typical cranial symptoms, especially headache (92%), scalp tenderness (58%), and clinically abnormal temporal artery (67%). Our results are consistent with those of Salvarani and Hunder but are at variance with those of Wong et al, who reviewed 22 patients with an ESR < 40 mm/hour and found headache to occur in only 41% of the cases (2). Moreover, our study extends the findings of Salvarani and Hunder by demonstrating heterogeneous acute serologic phase response among patients with low ESR and raised inflammatory proteins in half of the cases. Although active temporal arteritis without any detectable inflammatory response was observed in 3% of the cases in our series, the percentage was not reported in the Salvarani and Hunder series, although the ESR was < 20 mm/hour in 6 patients. Moreover, we found a strong correlation between ESR and C-reactive protein as well as hemoglobin levels and the platelet count, but not the levels of other acute phase reagents. Our results suggest that routine laboratory tests of patients with suspected temporal arteritis should include not only an ESR, but also a C-reactive protein measurement and blood cell counts. However, additional measurement of haptoglobin, orosomucoid, and fibrinogen levels seem of limited value if both ESR and C-reactive protein levels are low.
Wong et al reported the occurrence of visual symptoms in 36% of the patients with temporal arteritis and a low ESR (2). Although no patient in the Salvarani and Hunder series developed permanent blindness, in our series 2 patients with low ESR and no detectable inflammatory response developed permanent blindness and another patient reported transient diplopia. Therefore, the finding of normal ESR and C-reactive protein values does not definitively rule out giant cell arteritis in an elderly patient with visual ischemic symptoms and recent onset headache. Glucorticoid treatment should not be delayed and should be instituted while awaiting the result of temporal artery biopsy.
Hernandez-Rodriguez et al recently reported that a weak inflammatory response is associated with lower corticosteroid requirements and shorter duration of therapy in patients with giant cell arteritis (3). Accordingly, the mean duration of glucocorticoid therapy was only 21.5 months in the patients in the Salvarani and Hunder series, and 21.2 months in our series, providing some evidence that long-term glucocorticoid therapy is seldom required in patients with temporal arteritis and a low ESR.