ARTERITE GIGANTO-CELLULARE, POLIMIALGIA, AORTITI

PROGNOSIS FACTORS AND LONG TERM EVOLUTION OF A COHORT OF 133 PATIENTS WITH GIANT CELL ARTERITIS
E. Hachulla, V. Boivin, U. Pasturel-Michon, J. Bouroz-Joly, M. Perez-Cousin, P.Y. Hatron, B. Devulder
Department of Internal Medicine, 59037 Lille Cedex, France

Objective: To determine outcome and survivorship among patients with a well-defined giant cell arteritis (GCA).

Patients and Methods: During 1977 and 1995, 176 patients have been treated in our department of Internal Medicine for GCA. The patient, his family or his local practitioner were all contact at the date of the study (July-October 1995). Treatment modalities and follow-up were obtained for 133 patients. All patients (except 11) had 3 or more 1990 ACR classification criteria for GCA.

Results: Relapse during withdrawal of steroids was observed in 83 patients (62.4%). No correlation was found with age, sex, initial dose nor type of steroid used (i.e. prednisone or prednisolone). We found only a weakly correlation with initial erythrocyte sedimentation rate (ESR) (p < 0.01, r = 0.23). On 56 patients free of treatment (mean duration of treatment: 40 months), 27 (48%) developped a recurrence of the disease 1 to 25 months later (mean follow-up after the end of the treatment: 52 months). No correlation was found with age, sex, initial dose of steroid, number of relaps during treatment, initial ESR, but a statistical difference was found with the type of steroid used (44 patients had prednisone, 9 were free of steroids since 41 months ± 10, no recurrence; 89 had prednisolone, 47 were free of steroids since 40 months ± 5, 27 recurrence) (p < 0.001). Survival analysis was performed using the Kaplan-Meier method for representation and the Mantel-Menszel for comparison of groups. At the time of the study, 41 patients were dead (30.7%). A significant reduction of survival was found in subgroup men vs women (p = 0.02) (mean age at the diagnosis was not different with sex but life esperancy is smaller in men), presence of initial ocular manifestations vs absent (p = 0.04), prednisone using vs prednisolone (p < 0.01). Prednisone and prednisolone groups did not differ for age, sex, initial ocular manifestations, mean duration of the treatment, dose at 12 months (13 vs 15 mg/d).

Conclusion: Relaps was observed in 62.4%of the patients during withdrawal of steroids (correlated with initial ESR). A recurrence of the disease was observed in 48% of the patients 1 to 25 months after the end of the treatment and was associated with prednisolone using. Long term survival is better in patients without initial ocular manifestations and in patients treated with prednisolone.

CRANIAL ISCHEMIC EVENTS IN GIANT-CELL ARTERITIS PATIENTS (GCA) PRESENTING WITH APPARENTLY ISOLATED POLYMYALGIA RHEUMATICA (PMR)
J. Hernández-Rodríguez, C. Font, B. Coll-Vinent, J. Casademont, A. López-Soto, J.M. Grau, M.C. Cid
Department of Internal Medicine, Hospital Clinic, Barcelona, Spain

Background: Several studies indicate that GCA patients presenting with isolated PMR with no cranial symptoms are at very low risk of developing GCA-related ischemic events. However, this issue remains controversial since the occurrence of irreversible cranial ischemic complications in patients who have been suffering from isolated PMR for several months has been repeatedly reported.

Objective: To assess the risk of developing cranial ischemic events in GCA patients presenting with apparently isolated PMR

Patients and Methods: Forty five GCA patients in whom disease manifested as an apparently isolated PMR for at least 2 months were selected among 220 individuals with biopsy-proven GCA. Clinical manifestations, as well as their chronological appearance before diagnosis were retrospectively recorded.

Results: Thirty four patients presented with PMR for an average of 8 months (range 2 months-5 years) and later developed cranial symptoms for 5 weeks (range 1-48 weeks) which eventually led to diagnosis (group 1). Eleven patients, after presenting a self-limiting brief course of mild cranial symptoms lasting 2 weeks (range 1-8 weeks) developed PMR which was their chief complaint for 4 months (range 2-10 months) and the reason for admission (group 2).
Ten patients suffered permanent cranial ischemic complications at the moment of diagnosis (7 monocular blindness, 2 binocular blindness, 1 cerebral infarct), 1 transient diplopia and 2 amaurosis fugax. Thirteen out of 34 patients in group 1 (38.2%) suffered ischemic complications, either transient or permanent, and no patients in group 2 ever developed ischemic events (p = 0.0195).

Conclusions: Isolated PMR may precede for long time the development of cranial GCA symptoms with all its potential complications. Among patients with GCA presenting with an apparently isolated PMR, the only patients who are at low risk of developing ischemic complications seem to be those in whom a self-burning episode of cranial symptoms can be recorded.

MALIGNANCY ASSOCIATED WITH THE POLYMYALGIC SYNDROME. A PROSPECTIVE STUDY
G. Haugeberg, H. Dovland, V. Johnsen
Department of Rheumatology, Vest-Agder Central Hospital, N-4604 Kristiansand. S, Norway

The aim was to examine and determine diagnosis and clinical findings in patients presenting with proximal muscle pain and/or stiffness (the polymyalgic syndrome) suspected to have polymyalgia rheumatica.

By postal information primary care doctors in the community and central hospital doctors were asked to refer consecutive patients presenting with new symptoms of proximal muscle stiffness or pain suspected to have PMR to the Department of Rheumatology for examination.

During a two-year period (June 95-June 97) a total of 50 patients, 31 females and 19 males were referred and examined. Forty patients were diagnosed to have PMR, two to have temporal arteritis, and three with miscellaneous disorders. Five (10%) had malignant neoplasm (three hematopoietic and two non-hematopoietic). Based on the incidence of cancer in Norway in 1993, a frequency of 1.6% would have been expected in the polymyalgic syndrome group. The difference between observed frequency of malignancy and the age and sex adjusted frequency of cancer in the population was highly significant (Fisher's exact test (two tailed), p = 0.0013).

In conclusion patients presenting with a polymyalgic syndrome turns out to have a variety of diagnoses. Malignant neoplasm seems to be a relative frequent finding in patients presenting with the polymyalgic syndrome. In contrast to other studies we postulate a association between malignancy and the polymyalgic syndrome. Further controlled prospective studies is needed to confirm our findings.

FAVORABLE ROLE OF INTERLEUKIN 10 IN PATIENTS WITH POLYMYALGIA RHEUMATICA
R.H. Straub1, T. Glück1, P. von Landenberg1, J. Georgi3, K. Helmke4, J. Schölmerich1, P. Vaith2, B. Lang1
1 Dept of Int. Med. I, Univ. Med. Center Regensburg; 2 Dept Int. Med., Univ. Med. Center Freiburg; 3 Dept Int. Med. & Rheumatol., Ostseeklinik, Damp; 4 Dept Int. Med. IV, Hospital München-Bogenhausen, Germany

To determine the relationship between the antiinflammatory molecule interleukin (IL)-10 and disease severity, IL-1b, tumor necrosis factor a (TNFa) and IL-1 receptor antagonist (IL-1RA) in patients with polymyalgia rheumatica (PMR). In 102 patients with PMR, we determined the severity of the disease by the presence of typical clinical symptoms (symptom score with a maximum of 10 pts). IL-10, IL-1b, TNFa and IL-1RA were measured in all patients and 31 age-matched healthy subjects by enzyme immunometric assays.

Patients with low serum levels of IL-10 (below the mean + 3 SD of normal subjects: 7.79 pg/ml) as compared to patients with elevated serum levels had more often adynamia (p = 0.045), bilateral muscular pain in shoulders, upper arms or neck (p = 0.045), bilateral muscular pain in the pelvic girdle (p < 0.001), headache (p = 0.014), morning stiffness (p < 0.001), symptoms of depression (p = 0.013), initial weight loss (p = 0.011), and had a higher symptom score (5.5 ± 0.4 vs 3.7 ± 0.3, p < 0.001). The overall symptom score correlated negatively with IL-10 serum levels (RRank = -0.421, p < 0.001). Furthermore, IL-10 correlated negatively with IL-1b (p = 0.013) and TNFa (p = 0.039). The inhibitory impact of IL-10 on IL-1b and TNFa and the stimulatory impact of IL-10 on the ratio of IL-1RA to IL-1b seems to be increased in the presence of corticosteroid treatment.

Elevated serum levels of IL-10 were associated with a more mild form of PMR. This indicates the previously demonstrated antiinflammatory favorable properties of IL-10 in patients with other long-standing chronic inflammatory diseases such as PMR.

MUSCULOSKELETAL MANIFESTATIONS IN GIANT CELL ARTERITIS (GCA)
Carlo Salvarani, Sherine E. Gabriel, Gene G. Hunder
Reggio Emilia, Italy
Mayo Clinic Rochester, MN 55905, USA

Although cranial and ocular symptoms are the most prominent findings in GCA, musculoskeletal manifestations (MSM) have also been reported. To assess the spectrum of MSM in GCA, we investigated these findings in a population-based inception cohort of patients with GCA regarding frequency, temporal relationship to onset of GCA, and outcome.

Over a 42-year period, 128 patients were identified with GCA who satisfied ACR criteria. A standard data collection form was used to record clinical and laboratory findings, and outcome (last follow up or death). Fifty-three (41%) had PMR during the course of illness, 23 before, 17 concurrently with, and 13 after the diagnosis of GCA. Thirty (23%) patients developed one or more peripheral MSM. These included peripheral synovitis in 17, distal extremity swelling with pitting edema in 13, distal swelling without edema in 5, tenosynovitis in 6, carpal tunnel syndrome in 2, and rheumatoid arthritis in 6. Patients frequently had more than one episode of peripheral findings and more than one type of symptom in an episode or in different episodes. Overall, 57 episodes of distal MSM with one or multiple symptoms were observed in the 30 patients. In 14 the initial distal MSM occurred before the diagnosis of GCA, in 4 they were concurrent, and in 12 the initial MSM were after the diagnosis of GCA. Twenty-one of the 30 with peripheral MSM also had PMR, and 16 of the total of 35 peripheral episodes in these PMR patients occurred when PMR was active. Three patients were diagnosed with RA before the diagnosis of GCA and 3 afterwards. Only 2 with RA developed joint erosions during follow up. The onset of PMR and peripheral MSM in the majority was within two years of diagnosis of GCA.

The results indicate that MSM in GCA are common and varied. Most appear linked temporally to the underlying GCA disease processes indicting that the nature of this illness and its clinical expression are broader than often considered. An awareness of the various MSM associated with GCA will help in proper diagnosis and appropriate therapy.

DETECTION OF CHLAMYDIA PNEUMONIAE IN GIANT CELL VASCULITIS AND CORRELATION WITH THE TOPOGRAPHICAL ARRANGEMENT OF TISSUE-INFILTRATING DENDRITIC CELLS
A.D. Wagner, H.C. Gerard1, T. Fresemann, W.A. Schmidt, A. Hudson1, H. Zeidler
Dept. of Rheumatology, Medical School Hanover, Hanover 30625, Germany
1 Dept. of Immunology and Microbiology, Wayne State University, School of Medicine, Detroit, MI 48201, USA

Giant cell arteritis (GCA) is a vasculitis which predominantly affects medium- and large-sized arteries. The recently detected CD4+ T cell activation lead to the conclusion that GCA is an antigen driven disease. With the increasing identification of Chlamydia pneumoniae (Cp) in arterial vessel walls it is of interest to know whether Cp could play a role in the pathogenesis of GCA. Fourteen paraffin-embedded temporal artery biopsy specimens from 9 patients with GCA were examined by immunohistochemistry and PCR for the presence of Cp. In 5 cases tissue specimens were available from the left and the right temporal artery. Three temporal artery tissue specimens without histologic findings were used as negative controls. In tissue samples from 7 patients Cp was detected by immunohistochemistry (polyclonal antibody directed against a fragment of one of the 95/98 kDa proteins) and PCR (against the MOMP gene). Only 2 patients showed no signal for Cp with both techniques. The control samples were all negative for Cp. The results of the immunohistochemical study revealed that Cp predominate in the adventitial layer of the arteries and were found in granulomatous infiltrates.

Additionally the second part of the current study was designed to investigate potent antigen presenting cells. Therefore dendritic cells (DC) were examined by immunohistochemistry for their presence and localisation in consecutive temporal artery specimens. The results showed that there is a strong topographical relation for the presence of Cp and DC in these tissue specimens. The DC predominate in the adventitial layer of the arteries.

In summary in 7 out of 9 patient with GCA Cp was detected in temporal artery biopsies. Therefore these data suggest some pathogenetic relevance of Cp in GCA. Due to the topographical relation, DC may represent the antigen presenting cells.

A PROSPECTIVE, DOUBLE-BLIND, RANDOMIZED, PLACEBO-CONTROLLED TRIAL OF METHOTREXATE (COMBINED WITH CORTICOSTEROIDS) IN THE TREATMENT OF GIANT CELL ARTERITIS (GCA)
R.F. Spiera1, H. Mitnick2, M.J. Kupersmith2, M. Richmond1, H. Spiera3, M. Peterson1, S.A. Paget1
1 Hospital for Special Surgery; 2 NYU Med. Ctr.; 3 Mt. Sinai Med. Ctr, NY, NY, USA

Aim: To determine if methotrexate (MTX) has a corticosteroid (CS)-sparing effect in the treatment of GCA.

Methods: Patients with GCA were enrolled and treated with high dose CS as well as MTX starting at 7.5 mg./week or placebo. CS were tapered as guided by the treating physician, with MTX or placebo dose increased by 2.5 mg./week for disease flare with a maximum allowable dose of 20 mg./week. At remission, MTX or placebo was tapered by 2.5 mg./week to off.

Results: 21 patients were enrolled, 12 randomized to MTX, 9 to placebo. Baseline characteristics (age, height, weight, sedimentation rate, BMD, total CS dose to entry, and quality of life as measured by AIMS and SF36) were comparable between groups. At completion, there was no significant difference between MTX and placebo treated patients with regard to cumulative CS dose (6073 mg. and 5754 mg. respectively, p = 0.7), number of weeks to completion (62 and 51 respectively, p = 0.3), time to <10 mg. Prednisone (59 and 45 respectively, p = 0.2), BMD in lumbar spine (p = 0.2) or hip (p = 0.4) at one year, or quality of life as measured by AIMS and SF36. There was no late vision loss in either group, and only one major relapse in a MTX treated patient (pulseless upper extremity with concomitant rise in sedimentation rate, responsive to resumption of high dose CS). There were few major CS related side effects and these did not significantly differ between groups. Our study was limited by low power (.65) given a greater than expected standard deviations in total CS dose.

Conclusion: With this study design, no CS sparing benefit could be attributable to the addition of MTX to conventional CS therapy of GCA. Both groups did well, with few major CS related side effects, and most patients were safely tapered off CS sooner than reported in many series. The shorter overall duration of steroid treatment in this study probably contributed to the remarkably low frequency of side effects, without increased ischemic risk for the patient.

CLINICAL AND HISTOPATHOLOGICAL ANALYSES OF 174 CASES OF AORTITIS
Francisco Roio, Norman Ratliff, Gary Hoffman
Cleveland Clinic, Cleveland, OH 44195, USA

Objectives: To systematically review 1205 aortic specimens obtained over 20 yrs to determine: 1) frequency of inflammatory lesions, 2) clinical profile, and 3) etiology of aortitis.

Methods: Retrospective. Standardized database. Specimens obtained in the course of surgery or autopsy. Inflammatory foci were found in 174 cases (14.4%). Chart reviews performed for each case with aortitis, and 71 control cases with only atherosclerosis.

Results: Of 1205 specimens, 85% had only atherosclerosis. Inflammatory lesions occurred in 14.4%. In 2% (24) unexpected path features of GCA were present, but only 6 of 24 had headaches or systemic symptoms (SS). Among 9 pts with a history of either SLE (4), sarcoid (1), retroperitoneal fibrosis (2), rheumatic fever (1), or polyarteritis nodosa (1), 6 had SS at time of surgery. Non-specific inflammation, that could not be related to a systemic illness, occurred in 69 pts (5.7%). None of 71 controls had SS.

Conclusions: The spectrum of aortitis is quite broad and includes individuals who lack SS, but who nonetheless have a degree of vascular injury that may contribute to aneurysm formation, aortic regurgitation, or vascular dissection. Whether immunosuppressive therapy for these pts would alter the course of illness is not certain.

PROBABILITY OF GIANT CELL ARTERITIS IN PATIENTS WITH LOW OR NORMAL ERYTHROCYTE SEDIMENTATION RATE
V. Rodríguez-Valverde, M. Delgado-Rodríguez, V.M. Martínez-Taboada, R. Blanco, M. Figueroa1, E. Uriarte1, M.A. Gonzalez-Gay2
H. U. "Marqués de Valdecilla" Facultad de Medicina, Universidad de Cantabria (Santander); 1 H. Na. Sa. de Aránzazu (S. Sebastián); 2 H. Xeral-Calde (Lugo), Spain

Purpose: To estimate the probability of having giant cell arteritis (GCA) when the ESR was: a). Mildly increased, but <50 mm/1 h or b). Normal, defined as less than 15 mm/1 h.

Material and Methods: Our series of 261 patients with biopsy-proven GCA, (11 of them with an ESR < 50, sensitivity [S] = 0.96) and a combined series of 430 patients from 5 studies of the English literature (25 cases with ESR < 50; S = 0.94). The probability of GCA with an ESR < 50, was estimated by the Bayes' theorem: 1 - NPV = 1 - E (1 - p)/[E (1 - p) + p (1 - S)]. As E (specificity) is unknown, 2 values were assumed: 0.5 and 0.95. On the other hand, as none of our 261 cases had an ESR < 15, the probability of having GCA with normal ESR was estimated by the "rule of threes": if no events occurred in x individuals, its probability (with 95% certainty) is = 3/x.

Results: The figures show the probability of having GCA when the ESR is <50, for both values of E. They may help in clinical decision making if the pretest probability of GCA can be approximately known. Our 11 cases with ESR < 50 mm, had cranial manifestations suggestive of GCA (a high pretest probability). Moreover, the probability of GCA with normal ESR was only 1.1%.

Conclusions: If the ESR is <50, a temporal artery biopsy is indicated only when the pretest probability of GCA is high. If the ESR is normal, the probability of GCA is very low (around 1%).

THE CLINICAL UTILITY OF THE HISTOPATHOLOGY "HEALED ARTERITIS" OF THE SUPERFICIAL TEMPORAL ARTERY (STA)
William S. Wilke, Adolfo Guemes, Gregory S. Kosmorsky
The Cleveland Clinic Foundation, Cleveland, OH 44195, USA

We describe a retrospective medical record review of 28 unselected patients with STA biopsies which demonstrated histopathology consistent with "healed arteritis."* This unselected sample represents 35.5% of all STA biopsies with "healed arteritis" encountered during 1985-1992 at this Institution, allowing ³5 year follow-up.

A final clinical diagnosis of Giant Cell Arteritis (GCA) was made in 2 of 28 (7%) patients based on the presence of jaw claudication, atypical response to corticosteroid treatment and Westergren erythrocyte sedimentation rate >50 mm/hr in both patients; polymyalgia rheumatica in 1, and new headaches with scalp tenderness in the other. Other final diagnoses included infection (3), neoplasm (4), Polymyalgia Rheumatica (3), Sjogren's syndrome (1), Rheumatoid Arthritis (1), idiopathic glomerulonephritis (1), and non-inflammatory diseases (11). Five patients experienced improvement of symptoms with corticosteroid treatment, one with headache and one with proximal myalgias. None of the other three experienced new headache, jaw claudication, scalp tenderness or clinical abnormality of the STA.

Comparative information from 28 patients with biopsy-proven GCA and 28 further patients with negative STA biopsies will be available at presentation.

This preliminary report demonstrates that the histopathology "healed arteritis" of the STA is encountered in patients with heterogenous diseases, and rarely identifies patients with GCA.

Reference: Lie JT, Brown AL, Carter ET: Spectrum of aging changes in temporal arteritis. Its significance, in interpretation of biopsy of temporal artery. Arch Path 1970; 9: 278-285.

THE SIGNIFICANCE OF "HEALED ARTERITIS" UPON BIOPSY OF THE TEMPORAL ARTERY FOR SUSPECTED GIANT CELL ARTERITIS
E.T. Weiss, S.C. Lester, S.M. Helfgott, L.B. Klickstein
Harvard University and Brigham & Women's Hospital, Boston, MA 02115, USA

Temporal artery (TA) biopsy is one of five equally weighted ACR criteria for the diagnosis of giant cell arteritis (GCA), however in clinical practice it may be weighted more heavily because it is by far the most specific criterion. Active vasculitis with a predominantly mononuclear cell or granulomatous infiltrate is the positive result. Histopathologic findings compatible with healed arteritis are well-described in reported series but are of uncertain significance. One explanation for the finding of healed arteritis is a delay in the time of biopsy from the time of initiation of steroid therapy (McDonnell et al., Ophthalmology 93: 518, 1986). An alternative explanation is that in some cases healed arteritis represents spontaneous healing of a segment of artery, a "skip lesion" in time, analogous to the skip lesions in space that may occur linearly along an involved artery.

There have been 439 consecutive TA biopsies performed at our hospital from 416 patients during the period between 1987-1997. 70 (15.9%) of these showed active vasculitis and 39 (8.9%) had findings compatible with healed arteritis. Three of the latter group represented re-biopsies of patients treated for GCA for an average of 29 months. Among the remaining 36 patients, we have observed healed arteritis in some individuals who underwent biopsy expeditiously and who had never received steroid therapy previously. Our studies thus far have shown no significant clinical differences between patients with healed or active arteritis. The absence of clinical differences between the two groups supports the concept that healed arteritis in patients with no or minimal steroid treatment represents spontaneous healing of previous vasculitic lesions, a "skip lesion" in time. We suggest that the histopathologic finding of healed arteritis, irrespective of the cause, should be considered a positive biopsy result in the appropriate clinical setting.

VISUAL HALLUCINATIONS - PREDICTORS OF VISUAL LOSS IN TEMPORAL ARTERITIS
G. Nesher, R. Nesher, M. Sonnenblick
Sapir Medical Center, Kfar-Saba and Shaare-Zedek Medical Center, Hebrew University, Jerusalem 91031, Israel

Purpose: Visual hallucinations (VH) in sane people (the Charles Bonnet syndrome) are associated with low vision states. This study evaluated the prevalence, clinical significance and outcome of VH in patients with temporal arteritis (TA).

Methods: All patients diagnosed by the authors as having TA between 1992-1998 were evaluated prospectively. Data on clinical symptoms, including ophthalmic manifestations and questioning the possible occurrence of VH were collected.

Results: TA was diagnosed in 25 patients. 5 presented with acute visual loss (VL) - anterior ischemic optic neuropathy in 4 and central retinal artery occlusion in 1. In 4 of these cases VL was preceded by 1-10 days of recurrent VH, in forms specific for each patient: flowers, a carriage and horses, cats chasing mice, and colorful rays. VH were not the reason for seeking medical advice, and the patients were initially quite reluctant to describe their VH, fearing the labeling of "psychotic disorder". There were 3 women and 1 man, their mean age was 72, and all were living alone. All were biopsy-positive, and had headaches and elevated sedimentation rate. 3 had polymyalgia rheumatica symptoms and one had jaw claudication They had no psychiatric disorder, and were aware of the unreal nature of their vision. VH disappeared in all cases within 10 days of starting steroid therapy. VH did not occur in any patient without VL.

Conclusion: VH are not rare in TA, occuring in 16% of the patients in this study. Their occurrence in cases suspected as TA should alert us to the possible development of VL.

YKL-40 IN THE CIRCULATION AND TEMPORAL ARTERY IN PATIENTS WITHGIANT CELL ARTERITIS AND POLYMYALGIA RHEUMATICA
B. Baslund, J.S. Johansen, C. Garbarsch, M. Stoltenberg, M. Hansen, P. Price
Dpt. of Rheumatology, Hvidovre Hospital, Dpt. of Anatomi A, University of Copenhagen, Denmark
Dpt. of Biology, University of California, San Diego, USA

YKL-40 (also named human cartilage glycoprotein 39 (HCgp-39)), a mammalian member of the chitinase family, and an autoantigen in rheumatoid arthritis, is secreted by synovial cells, chondrocytes, neutrophils and macrophages. The latter may play an important role in the pathogenesis of giant cell arteritis (GCA). The aim of this study was to evaluate serum YKL-40 levels to define if there is a difference between patients with GCA and polymyalgia rheumatica (PMR). Furthermore, the expression of YKL-40 in biopsies of the temporal artery from the patients was studied.

Serum YKL-40 was determined by RIA in patients with clinical symptoms of GCA (N = 19) or PMR (N = 8) followed prospectively during one year treatment with prednisolone. Immunohistochemical expression of YKL-40 in the temporal artery biopsy taken before treatment was studied using affinity purified polyclonal antibodies against YKL-40.

Patients with GCA had at baseline increased serum YKL-40 (median 256 µg/L, range 62-900, p < 0.01) compared to healthy controls (102 µgL) and the level decreased to normal levels during prednisolone treatment (-38% after one month, p < 0.001). Most patients with PMR had normal serum YKL-40 (158 µg/L, range 74-416) and had no changes in serum YKL-40 during treatment. During the one year study period serum YKL-40 did not react as an acute phase reactant at times of relapses when C-reactive protein and ESR raised. Expression of YKL-40 antigen was found in giant and mononuclear cells in the temporal artery biopsy with histological signs of GCA. Whereas no expression of the protein was detected in the biopsies from the patietns with PMR.

In conclussion, YKL40 was found in higher levels in serum from patients with GCA than PMR. The YKL-40 levels in serum did not react as an acute phase reactant. The YKL-40 antigen was detected in the mononuclear cells in the biopsies of the temporal artery.

TREATMENT OF POLYMYALGIA RHEUMATICA DURING THE FIRST SIX MONTHS. A COMPARISON OF TWO STEROID REGIMENS
Thomas Osborn, Trumane Ropos
Cleveland Clinic Florida, Ft. Lauderdale, FL 33309, USA

To compare 2 steroid treatment regimens during the first 6 months of therapy, 100 patients with newly-diagnosed polymyalgia rheumatica (PMR) were randomized into 2 groups of 50. One group received prednisone 20 mg per day for 1 month and a subsequently tapered dose (pred20). The other group received methylprednisilone 80 mg IM and prednisone 10 mg per day for 1 month with a subsequently tapered dose (depo + pred10). Clinical symptoms, physical examination and laboratory were compared at 1 month, 3 months and 6 months.

At 6 months no significant difference was found in proximal joint or muscle pain (8% for pred20 vs 6% for depo + pred10, p = NS), headaches (14% vs 12%, NS), jaw fatigue (0% vs 0%, NS), or joint swelling (6% vs 2%, NS). Physical findings including difficulty raising arms overhead (10% vs 8%, NS), difficulty rising from a chair unassisted (8% vs 10%, NS) and muscle tenderness (6% vs 2%, NS) were not significantly different. Number of patients with an elevated ESR was not significantly different (6% vs 8%, NS). The number of patients considered not responding was 4 in the pred20 and 3 in the depo + pred10 group. Number of patients with relapses within 6 months were identical at 3 in each group. More weight gain, 7 lbs vs -2 lbs (86% vs 18%, p < 0.01) and more bruising (58% vs 28%, p < 0.01) were found in the pred20 group. No visual or neurological complications occurred after treatment with corticosteriods was initiated in either group. There was no significant difference in the mean daily dose of prednisone at 6 months (5.5 mg vs 5.7 mg)

We concluded an initially brief high dose of steriods may allow a sustained earlier use of a lower dose of steriods in treating PMR.

CIRCULATING CD8+ T CELLS IN POLYMYALGIA RHEUMATICA/GIANT CELL ARTERITIS
V.M. Martínez-Taboada, C. Fito, R. Blanco, I. Ayerbe, C. Mata, A. Corrales, R. Gutiérrez-Polo, S. Sánchez-Andrada, J.L. Peña, V. Rodríguez-Valverde
Rheumatology Division, H. U. "Marqués de Valdecilla", Facultad de Medicina, Universidad de Cantabria (39008 Santander), Spain

Objective: During the last few years there have been several studies on T cell subsets in polymyalgia rheumatica (PmR) and giant cell arteritis (GCA) with conflicting results. Whereas some authors have found normal values of CD8+ T cells, others have found a decreased on CD8+ T cells. Furthermore, in some studies, the level of CD8+ cells was found to be related to disease activity, and it has been proposed to use a decrease of CD8+ T cells as a diagnostic criterion for PmR, specially to distinguish from rheumatoid arthritis (RA). The purpose of this study was to clarify the value of the determination of T cell subsets in PmR and GCA.

Methods: T lymphocyte subsets were determined by flow cytometry using a whole blood lysis technique in the following groups: 28 PmR and 6 GCA patients before steroid treatment (mean age, 69.5 ± 6.5 years), 17 RA patients before treatment (58.6 ± 16.3 years), 32 PmR and GCA patients in clinical remission with steroid treatment (73 ± 8.4 years), 55 PmR patients in complete remission without steroid treatment (73.7 ± 7.9 years), and 18 controls with non inflammatory conditions (68.6 ± 8.9 years). PmR was diagnosed according to the criteria proposed by Chuang et al. (Ann Intern Med 1982) and all the patients with GCA had a positive temporal artery biopsy. Total white cell, lymphocyte and platelet counts, haemoglobin, CRP and the ESR were measured by routine techniques. Comparisons were made by means of Student's t test and the Mann-Whitney test.

Results: Compared with controls, CD8+ T cells were not reduced before steroid treatment in patients with PmR or GCA in percentage (26.22% versus 25.09%, p = 0.69) or absolute numbers (p = 0.28). However, these 2 groups differed in hemoglobin levels, platelets, CPR and ESR (p < 0.01). When patients with active PmR or GCA were compared with PmR/GCA patients in remission with steroid therapy, no significant differences were found except for the ESR, hemoglobin levels and platelets (p < 0.01). The same was true when patients with active disease were compared with PmR patients in remission without steroid therapy, the only significant differences were those parameters reflecting the acute phase response: ESR hemoglobin, and platelets (p < 0.03). Compared with RA patients, CD8+ T cells were not reduced in patients with active PmR in percentage (23.2% versus 24.7%, p = 0.55) or absolute numbers (p = 0.21). The only significant differences were age (p = 0.003) and ESR (p = 0.03). The percentage of CD3+ and CD4+ T cells were significantly higher in patients with active PmR/GCA than in non active PmR/GCA patients and controls. However, there was not significant differences in the absolute numbers of CD3+ and CD4+ T cells.

Conclusions: This study does not confirm the previous findings that the percentage or number of circulating CD8+ T cells is reduced in patients with active PmR or GCA. The utility of the determination of CD8+ T cells for diagnostic purpose should be also evaluated in a larger series of patients.

CIRCULATING SOLUBLE ADHESION MOLECULES IN PATIENTS WITH GIANT CELL ARTERITIS. CORRELATION BETWEEN SOLUBLE INTERCELLULAR ADHESION MOLECULE-1 (sICAM-1) LEVELS AND DISEASE ACTIVITY
Blanca Coll-Vinent, Carme Vilardell, Carme Font, Joaquim Oristrell, Josep Hernández-Rodríguez, Alfons López-Soto, Jordi Yagüe, Álvaro Urbano-Márquez, Josep M. Grau, Maria C. Cid
Hospital Clínic, University of Barcelona, 08036 Barcelona, Spain

Objective: To evaluate whether changes in levels of circulating adhesion molecules are related to disease activity in patients with giant cell arteritis (GCA).

Methods: A sandwich ELISA was used to measure soluble intercellular adhesion molecule-1 (sICAM-1), sICAM-3, vascular cell adhesion molecule-1 (sVCAM-1), E-selectin (sE-selectin) and L-selectin (sL-selectin) in sera and plasma from patients with GCA. A cross-sectional study was performed on 64 GCA patients at different activity stages: 36 active, 13 in recent remission (less than 2 years after starting treatment) and 15 in long-term remission (more than 2 years after starting treatment). 12 age and sex-matched healthy donnors served as controls. Thirteen of GCA patients were evaluated longitudinally at the time of diagnosis and serially during follow-up.

Results: At the time of diagnosis, sICAM-1 levels were significantly higher in active GCA patients than in controls (350.91 ± 122.54 ng/mL versus 239.18 ± 52.21 ng/mL, p < 0.01). In contrast, sICAM-3, sVCAM-1, sE-selectin and sL-selectin levels did not differ from those obtained in normal donors. With steroid administration, a decline in sICAM-1 levels was observed, reaching normal values when clinical remission was achieved (263.18 ± 92.7 ng/mL globally, 293.59 ± 108.39 ng/mL in the group of patients in recent remission, and 236.83 ± 70.02 ng/mL in those in long-term remission). In the thirteen patients followed longitudinally, sICAM-1 levels also normalized with clinical remission (225.87 ± 64.25 ng/mL in patients in recent remission, and 256.29 ± 75.15 ng/mL in those in long-term remission).

Conclusions: Circulating sICAM-1 levels clearly correlate with clinically apparent disease activity in GCA patients. Differences with results previously found in patients with other vasculitides may indicate that different pathogenic mechanisms contribute to vascular inflammation in different disorders.

INFLAMMATION-INDUCED ANGIOGENIC RESPONSE AND THE DEVELOPMENT OF CRANIAL ISCHEMIC COMPLICATIONS IN GIANT-CELL ARTERITIS PATIENTS
Maria C. Cid, G.S. Mireia Cebrián, Carme Font, José Hernández-Rodríguez, Blanca Coll-Vinent, Alvaro Urbano-Márquez, Josep M. Grau
Department of Internal Medicine, Hospital Clínic, Barcelona, Spain

Background: Neovascularization is frequently found in vasculitis lesions (Cid et al JCI 1993; 91: 977-85, Cid et al Arthritis Rheum 1994; 37: 1055-1). Angiogenesis may play a dual role in vasculitis. We have previously shown that newly-formed vessels strongly express adhesion molecules for leukocytes (Coll-Vinent et al Arthritis Rheum 1998; 41: 435-44) and, therefore, angiogenesis may amplify the inflammatory reaction. On the other hand, neovascularization may compensate for ischemia by generating new vessels.

Objective: To measure the intensity of neovascularization in temporal arteries from patients with giant-cell arteritis (GCA). To correlate the degree of neovascularization with the intensity of systemic inflammatory response and the development of cranial ischemic events.

Patients and Methods: The study group consisted of 31 patients with biopsy-proven GCA. Five normal temporal arteries from patients with an alternate diagnosis were used as controls. Endothelial cells were immunostained with the lectin Ulex Europaeus. Neovascularization was quantitated as percentage of the artery circumference occupied by neovessels and multiplied by the number of layers in which they were distributed and a 0-6 score was established. Clinical data, particularly the occurrence of cranial ischemic events, and the presence of inflammatory parameters (fever, weight loss, ESR > 85 and Hb < 11 mg/dL) were recorded.

Results: In normal specimens, only scattered vasa vasorum at the adventitia were observed. In GCA lesions, neovascularization occurred at the adventitial layer and at the intima/media junction. Ischemic events occurred in 13 patients. Six had transient events (amaurosis fugax or diplopia) and 7 permanent visual loss. The neovascularization score was lower in patients with ischemic events (1.14 vs 2.15, p = 0.03) and was higher in patients with 3 or more inflammatory parameters (2.55 vs 0.9, p = 0.036).

Conclusion: Inflammation-induced angiogenic response is variable among GCA patients and correlates with the intensity of the systemic inflammatory response. Angiogenesis may compensate for ischemia at distal sites since a poor inflammatory response was associated with a higher frequency of cranial ischemic events.