SEARCH

SEARCH BY CITATION

Keywords:

  • Polymyalgia rheumatica;
  • Relapse;
  • Recurrence;
  • Interleukin-6;
  • Erythrocyte sedimentation rate;
  • C-reactive protein

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES

Objective

To determine laboratory parameters that may be useful in identifying polymyalgia rheumatica (PMR) patients who require long-term corticosteroid therapy.

Methods

A prospective followup study of 94 consecutive untreated patients with PMR were assessed for relapse/recurrence for a mean of 39 months. This cohort represented all the patients diagnosed over a 4-year period in 2 Italian secondary referral centers. Patients were monitored for clinical signs and symptoms, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and serum interleukin-6 (IL-6). IL-6 levels were also measured in 43 controls matched to the patients for age and sex.

Results

The ESR was elevated in 91.5% of the patients prior to therapy initiation, as were CRP in 98.9% and serum IL-6 in 92.6%. Forty-seven (50.0%) patients had at least 1 relapse/recurrence during the followup period and 24 (25.5%) had at least 2. After 4 weeks of prednisone therapy, ESR was elevated in 13.2% patients, CRP in 41.9%, and serum IL-6 in 37.2%. IL-6 levels remained persistently elevated in 9.9% and CRP in 8.7% of patients during the first year of followup, whereas no patient had persistently elevated ESR. Persistently elevated CRP and IL-6 levels were significantly associated with an increased risk of relapse/recurrence. In particular, patients with persistently elevated levels of IL-6 during the first year of therapy had the highest relative risk.

Conclusion

Despite the control of clinical symptoms, corticosteroids do not adequately control the inflammatory process in a subset of patients with PMR who have persistently elevated levels of CRP and IL-6 and who have a higher risk of relapsing.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES

The course of polymyalgia rheumatica (PMR) is heterogeneous with a variable corticosteroid requirement. A treatment course of 1–2 years is often necessary (1). However, ∼30–50% of the patients have a more chronic, relapsing course and may require low doses of corticosteroids for much longer (2–6). Corticosteroid-related adverse events are common and they are strictly related to cumulative dose. A long-term followup study demonstrated that 65% of patients with PMR had at least 1 serious corticosteroid-related event, such as osteoporotic fractures or diabetes mellitus (7).

To identify the subset of patients who require long-term corticosteroid treatment is essential to guiding treatment strategies. The association from the very beginning of therapy of steroid-sparing agents, such as methotrexate (8), could reduce the serious corticosteroid-related events observed in PMR patients.

Unfortunately, there are no reliable predictors of duration of corticosteroid therapy in PMR. Studies on pretreatment erythrocyte sedimentation rate (ESR) as a prognostic indicator of corticosteroid therapy duration have yielded conflicting results (6, 9–11). Although C-reactive protein (CRP) and interleukin-6 (IL-6) appear to be more sensitive indicators of disease activity, it remains unclear whether they provide any prognostic advantage in clinical practice over ESR (11–16).

The aim of this study was to evaluate and compare the usefulness of ESR, CRP, and IL-6 levels as predictors of relapse or recurrence using detailed followup information from a prospectively studied large cohort of patients with pure PMR.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES

Patients.

Ninety-four consecutive untreated patients with pure PMR, representing all the patients diagnosed over a 4-year period (1994–1997) in 2 Italian rheumatology secondary referral centers (Prato and Reggio Emilia, Italy), were prospectively assessed in these same hospitals. The diagnosis of PMR was accepted if all the following were present: 1) persistent pain (for at least 1 month) involving 2 of the following areas: neck, shoulders, and pelvic girdle; 2) morning stiffness lasting >1 hour; 3) rapid response to prednisone (≤20 mg/day); 4) age >50 years; and 5) absence of other diseases capable of causing the musculoskeletal symptoms (17). Patients with ESR <40 mm/hour who nevertheless satisfied the other criteria were included in the study.

All patients were seronegative for rheumatoid factor. At diagnosis and during the followup period, all patients were evaluated using the American College of Rheumatology (ACR; formerly the American Rheumatism Association) 1987 revised criteria for rheumatoid arthritis (RA) (18). Excluded were those who satisfied these criteria and those with clinical manifestations of giant cell arteritis or who had a positive temporal artery biopsy sample.

The patients were clinically assessed by the same rheumatologist at presentation, once monthly for the first 6 months, then every 3 months during the followup period. A standardized data collection form was used at every visit to record medical information, as described previously (19). Joint radiography was performed in all patients with joint swelling at some point during the course of the illness.

All patients were treated with prednisone at a mean starting dosage of 17.5 mg/day (range 10.0–25 mg/day) as a single morning dose. If the patient did not respond with a reduction of symptoms, the initial dosage was increased by 2.5–5 mg/day before the tapering schedule was started. Prednisone was reduced in all patients according to the same fixed schedule, with a reduction after 1 month if symptoms had resolved. Small monthly decrements of 5–1 mg were successively scheduled until reaching the minimal maintenance dosage of 2.5 mg every other day. This dosage was suspended after 1 month if both clinical and humoral (ESR ≤30 mm/hour and CRP ≤0.5 mg/dl) remission persisted.

Relapses and recurrences were defined as reappearance of clinical symptoms associated with elevated ESR or CRP (ESR >30 mm/first hour or CRP >0.5 mg/dl) in a patient receiving corticosteroids or after discontinuation of treatment. The symptoms were suppressed by resumption of, or increase in, prednisone dosage.

The end of the disease was recorded as the date of permanent discontinuation of therapy without relapse or recurrence. The endpoint of patient followup was the date of the last clinic visit or the date of death.

Over the followup period, no articular deformities or radiologic evidence of erosions were found in the 94 patients. In addition, none of them developed clinical features that satisfied the 1987 modified ACR criteria for RA (18).

Laboratory analysis.

Laboratory tests were performed on blood samples drawn in the morning before daily administration of the scheduled prednisone dose. Serial blood samples were collected concomitantly to clinical evaluation at the start of therapy, at monthly intervals for the first 6 months, then at 3-month intervals until the end of followup.

ESR was determined using the Westergren method. Because most of our patients were women over the age of 50 years, the upper limit of normal considered for ESR was 30 mm/hour (20). CRP was measured by nephelometry (NA latex CRP kit, Behringwerke, Marburg, Germany; upper limit of the normal reference range 0.5 mg/dl). Serum IL-6 concentrations were evaluated at diagnosis and at the 1st, 2nd, 6th, 12th, and 24th months by immunoassay using a commercial enzyme-linked immunosorbent assay kit (R&D Systems, Minneapolis, MN) according to the manufacturer's instructions. The sensitivity of the test is typically >0.70 pg/ml. IL-6 serum levels were also measured in 43 controls matched with the PMR patients for age and sex. Control subjects were randomly recruited from the lists of patients who were under the care of the same public health service family physicians as the cases. Normal IL-6 levels were considered <4 pg/ml (normal mean + 3 SD).

Statistical analysis.

Chi-square, Mann-Whitney, and Wilcoxon paired tests were computed to compare proportions and means. Statistical significance was evaluated at the 0.05 level, 2 tailed.

Cox proportional hazards regression model analyses were used to evaluate the relationship between the occurrence of at least 1 or 2 relapses during the followup and the acute phase reactants, including ESR, CRP, and IL-6 values at diagnosis and at the 1st, 2nd, 6th, 12th, and 24th month. Persistently elevated ESR, CRP, and IL-6 levels at different moments during followup (0–3 months, 0–6 months, 0–12 months, and 0–24 months) were considered additional variables. The models were adjusted for age at diagnosis, sex, systemic manifestations, peripheral arthritis, morning stiffness, and initial prednisone dosage. Statistical analysis was done using Stata statistical software (Release 8.0, 2003; Stata Corporation, College Station, TX).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES

The clinical and demographic characteristics of the patients are shown in Table 1. Table 2 compares PMR patients with 1 or 2 relapses with those without. The duration of prednisone therapy was significantly longer in patients with relapse/recurrence compared with those without.

Table 1. Demographic and clinical findings of the 94 polymyalgia rheumatica patients*
  • *

    Except where otherwise indicated, values are the median (range).

Male/female, %25.5/74.5
Age at onset of disease, years74 (53–86)
Duration of symptoms before diagnosis, months3 (1–29)
Duration of therapy, months18 (6–96)
Duration of followup, months35 (12–96)
Systemic symptoms or signs (fever, anorexia, weight loss) %53
Morning stiffness at diagnosis, hours3 (1–6)
Peripheral arthritis, %14.9
Patients with ≥1 relapse/recurrence, no. (%)47 (50.0)
Patients with ≥2 relapse/recurrence, no. (%)24 (25.5)
Starting prednisone dosage, mg/day17.5 (10.0–25.0)
Table 2. Comparison between the demographic and clinical findings of the polymyalgia rheumatica patients with and without relapse/recurrences*
 Patients without relapse/recurrence (n = 47) Group APatients with ≥1 relapse/recurrence (n = 47) Group BPatients with ≥2 relapse/recurrences (n = 24) Group CA versus B PA versus C P
  • *

    Except where otherwise indicated, values are the median (range). NS = not significant.

Male/female, %25.5/74.525.5/74.529.2/70.8NSNS
Age at onset of disease, years75 (58–86)74 (53–82)75 (65–82)NSNS
Duration of symptoms before diagnosis, months2 (1–10)4 (1–29)4 (1–29)0.0010.0001
Duration of therapy, months13 (6–96)26 (6–89)29 (6–89)0.00010.0001
Duration of followup, months28 (12–83)38 (21–96)38 (28–96)0.0040.002
Systemic symptoms or signs (fever, anorexia, weight loss) no./n (%)20/47 (42.6)24/47 (51.1)14/24 (58.3)NSNS
Morning stiffness at diagnosis, hours3 (1–6)3 (1–5)3 (2–4)NSNS
Peripheral arthritis, no./n (%)8/47 (17.0)6/47 (12.8)4/24 (16.7)NSNS
Starting prednisone dosage, mg/day12.5 (10.0–25.0)20.0 (10.0–25.0)25.0 (10.0–25.0)NSNS

Table 3 shows acute phase protein values at diagnosis and at different times during the followup period. A significant reduction in ESR, CRP, and IL-6 values was observed after the first month of therapy and persisted for the entire 24-month followup. Table 4 shows the number and the percentages of patients with elevated acute-phase reactants at different times during the followup period. ESR, CRP, and IL-6 at baseline were elevated in 91.5%, 98.9%, and 92.6% of PMR patients, respectively. After 1 month of therapy, the percentages fell to 13.2%, 41.9%, and 37.2%, respectively. About one-third of the patients had elevated levels of CRP and IL-6 at the different scheduled times during the first year of followup.

Table 3. Acute phase protein values at diagnosis and at different times during the followup*
 Baseline1st month3rd month6th month12th month24th month
  • *

    Values are the median (range). ESR = erythrocyte sedimentation rate; CRP = C-reactive protein; IL-6 = interleukin-6.

  • P = 0.0001 compared with baseline values.

ESR mm/hour68 (14–128)14 (2–97)13 (2–51)15 (1–61)14 (1–76)15 (1–72)
CRP mg/dl3.5 (0.7–18.2)0.5 (0.2–5.4)0.4 (0.2–4.2)0.4 (0.2–5.6)0.4 (0.2–8.7)0.4 (0.2–8.2)
IL-6 pg/ml22.5 (0.5–159.6)2.8 (0.1–101.3)2.1 (0.1–104.2)2.2 (0.1–94.3)2.0 (0.1–96.9)1.7 (0.1–82.2)
Table 4. Patients with elevated level of acute phase proteins at different times during the followup*
 Baseline1st month3rd month6th month12th month24th month
  • *

    Values are number of patients with elevated levels/total number of patients tested (%). ESR = erythrocyte sedimentation rate; CRP = C-reactive protein; IL-6 = interleukin-6.

  • P = 0.0001 compared with baseline value.

ESR86/94 (91.5)12/91 (13.2)10/88 (11.4)17/93 (18.3)21/92 (22.8)15/82 (18.3)
CRP92/93 (98.9)39/93 (41.9)34/93 (36.5)37/93 (39.8)36/92 (39.1)20/82 (24.4)
IL-687/94 (92.6)35/94 (37.2)24/94 (25.5)30/94 (31.9)25/91 (27.5)11/78 (14.1)

Table 5 shows a subset of patients who presented persistently elevated values of CRP and IL-6 during the followup at the 4 time intervals considered (0–3, 0–6, 0–12, and 0–24 months). No patient had persistent elevation of ESR.

Table 5. Patients with persistently elevated ESR, CRP, and IL-6 levels at different time intervals during the followup*
 0–3 months0–6 months0–12 months0–24 months
  • *

    Values are number of patients with characteristic/total number of patients tested (%). ESR = erythrocyte sedimentation rate; CRP = C-reactive protein; IL-6 = interleukin-6.

ESR----
CRP24/93 (25.8)14/93 (15.1)8/92 (8.7)6/81 (7.4)
IL-618/94 (19.1)12/94 (12.8)9/91 (9.9)5/78 (5.1)

Cox proportional hazards modeling identified the variables related to the acute phase reactants that independently increased the risk of at least 1 or 2 relapses or recurrences (Table 6). Elevated ESRs at diagnosis were correlated to an increased risk of developing ≥1 relapse or recurrence, whereas elevated levels of CRP and IL-6 at diagnosis were not. Given the low number of cases, it was not possible to compute the relative risk of having ≥2 relapses or recurrences for elevated levels of ESR and CRP at diagnosis. We compared the frequencies of relapses or recurrences between quartiles of ESR, CRP, and IL-6 values at diagnosis. However, the frequencies of relapses or recurrences did not differ according to the level of the inflammatory marker.

Table 6. Relative risk of having ≥1 or ≥2 relapses/recurrences among patients with polymyalgia rheumatica for acute phase reactants*
VariableAt least 1 relapse/recurrenceAt least 2 relapses/recurrences
Relative risk*95% CIPRelative risk*95% CIP
  • *

    Relative risk determined by Cox proportional hazards regression analysis and adjusted for age at diagnosis, sex, systemic manifestations, peripheral arthritis, morning stiffness, and initial prednisone dosage. 95% CI = 95% confidence interval; ESR = erythrocyte sedimentation rate; NC = not computed because of the small number of cases; NS = not significant; CRP = C-reactive protein; IL-6 = interleukin-6.

Baseline ESR6.70.9–51.60.07NCNCNC
ESR 1st month0.60.2–1.7NS0.40.05–3.2NS
ESR 3rd month3.31.3–8.40.014.01.2–13.70.03
ESR 6th month5.32.5–11.20.00014.71.7–13.00.003
ESR 12th month2.11.1–4.10.033.71.5–9.50.006
ESR 24th month1.30.5–3.3NS5.41.5–19.70.01
Baseline CRP0.30.03–2.3NSNCNCNC
CRP 1st month0.90.5–1.8NS1.00.4–2.6NS
CRP 3rd month1.30.7–2.8NS1.00.4–2.6NS
CRP 6th month3.11.7–5.80.00011.40.6–3.3NS
CRP 12th month2.61.4–4.90.0035.11.9–13.50.001
CRP 24th month2.51.2–5.00.012.51.0–6.10.05
CRP 0–3 months1.40.7–2.9NS1.10.4–3.1NS
CRP 0–6 months3.71.5–9.20.0052.10.6–7.6NS
CRP 0–12 months4.71.7–12.60.0023.40.8–15.5NS
CRP 0–24 months4.01.3–12.20.025.41.1–26.20.04
Baseline IL-61.20.4–4.4NS1.10.1–9.7NS
IL-6 1st month2.01.0–3.70.042.31.0–5.50.06
IL-6 3rd month1.10.5–2.5NS1.40.5–4.1NS
IL-6 6th month1.60.8–3.1NS4.31.7–11.00.003
IL-6 12th month2.91.5–5.50.0013.81.5–9.40.004
IL-6 24th month1.60.7–3.7NS3.11.2–8.10.02
IL-6 0–3 months2.00.9–4.6NS3.61.3–9.90.01
IL-6 0–6 months1.50.6–3.9NS4.51.4–14.30.01
IL-6 0–12 months1.80.6–5.0NS12.83.4–48.30.0001
IL-6 0–24 months1.40.4–5.1NS7.01.6–31.70.01

Elevated levels of ESR at the 3rd, 6th, and 12th months were significantly associated with an increased risk of developing at least 1 or 2 relapses/recurrences, as were elevated levels of CRP at the 12th and 24th months. Elevated levels of IL-6 at the 12th month were significantly associated with an increased risk of developing at least 1 or 2 relapses or recurrences, whereas elevated levels at the 6th and 24th months were associated with an increased risk of developing at least 2 relapses/recurrences.

Persistently elevated levels of CRP and IL-6 were significantly associated with an increased risk of developing relapses or recurrences. Patients with persistently elevated levels of CRP during the first, 6th, 12th, and 24th months of followup had a risk of developing ≥1 relapse/recurrence that was 2–5 times higher than that of the patients without. Patients with persistently elevated levels of IL-6 had a risk of developing ≥2 relapses/recurrences that was 4–13 times higher than that of the patients without.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES

The aim of our prospective study was to determine whether acute phase reactants could be used to differentiate patients with different disease severity to be able to predict therapy duration. It has been suggested that there are 2 subsets of PMR patients, one of which presents mild self-limiting disease requiring 1–2 years of treatment, and the other is made up of patients with a more chronic, relapsing course of disease who may require low doses of corticosteroids for several years (1–6).

Of all patients with PMR, 65% have at least 1 serious corticosteroid-related event; in particular, the risk of vertebral fractures is 5 times greater among women with PMR (7). A cumulative dose of prednisone of at least 2 gm is one of the variables that independently increased the risk of adverse events. Therefore, efforts should be made to maintain the corticosteroid dosages at the lowest levels possible for the shortest periods of time, particularly in patients with diabetes, osteoporotic fractures, hypertension, glaucoma, and other conditions negatively affected by higher corticosteroid doses.

Recently, a double blind, placebo-controlled study showed a steroid-sparing effect of methotrexate associated to corticosteroids (21). Furthermore, a pilot study showed that infliximab may have a steroid-sparing effect in the treatment of PMR patients who are resistant to corticosteroid therapy and have had corticosteroid-related side effects (22). Therefore, some evidence is accumulating that certain medications could be useful as steroid-sparing agents in PMR, reducing the corticosteroid exposure and thus the adverse effects related to corticosteroids.

The identification of reliable serologic predictors of the duration of corticosteroid therapy will make it possible to define the patients with a relapsing course who require long-term therapy and could benefit from the beginning from the association with steroid-sparing drugs. Few studies have evaluated the predictor role of acute phase markers. Conflicting results have been reported on the usefulness of pretreatment ESR as a prognostic indicator for duration of therapy (6, 9–11). CRP and IL-6 seem to be more sensitive indicators of disease activity (12, 23–25). In particular, IL-6 appears to be a potentially useful biologic marker of disease activity in PMR. IL-6 is the chief stimulator of the production of most acute-phase proteins. Increased production of IL-6 is a characteristic finding in patients with PMR, and corticosteroids rapidly reduce the levels of circulating IL-6 (11, 15, 16, 25). However, it is unclear if CRP and IL-6 have some advantages in clinical practice over ESR (11–16).

Our study confirmed that CRP was more sensitive than ESR for PMR diagnosis. CRP was elevated in 98.9% of the patients at diagnosis, whereas ESR was elevated in 91.5%. Standardized diagnostic and classification criteria are needed for PMR, and CRP levels could be included as diagnostic criteria.

We found that pretreatment ESR was of some prognostic value, whereas pretreatment CRP and IL-6 levels were not. Patients with elevated ESRs at diagnosis had a higher risk of having ≥1 relapse or recurrence. However, the low number of patients with normal ESR values at diagnosis limited the statistical power of this analysis.

Similar to previous studies (16, 25, 26), we observed the presence of a high percentage (around one-third during the first year) of patients with elevated levels of CRP or IL-6 at different times during the followup despite clinical remission. However, the most interesting finding was the evidence of a subgroup of patients who presented persistently elevated values of CRP and IL-6 during the followup period. The final multivariate models revealed that persistently elevated levels of CRP and IL-6 were significantly associated with a higher risk of developing relapses or recurrences.

Our current data support previous observations. Schreiber and Buyse found that patients whose CRP levels persisted to be elevated after 7 days of therapy had a longer treatment course and required higher amounts of corticosteroids (24). Weyand et al showed that persistently elevated levels of IL-6 after 1 month of treatment characterized PMR patients who required higher initial corticosteroid dosages and had to take them for longer (11).

No patient presented persistently elevated ESR values, confirming that this laboratory parameter is less sensitive than CRP and IL-6 in measuring persistent acute phase responses and in identifying patients at higher risk of developing relapses or recurrences. A possible conclusion from the present study is that the corticosteroid dosage should be increased to completely suppress CRP or IL-6, or a steroid-sparing agent should be added to therapy for patients with persistent elevation of CRP or IL-6 levels after 6 months of therapy. The higher risk of disease relapse suggests that these patients are undertreated and that the insufficient suppression of inflammation affects the long-term outcome of these patients.

Potential advantages and limitations of our study warrant discussion. We treated all patients with a fixed schedule of prednisone dose tapering, defined relapses and recurrences in a strict way, and determined all the 3 main acute phase reactants at different fixed times during a long-term followup of at least 2 years. Furthermore, we included an adequate number of patients (n = 94) with pure PMR, eliminating the possible confounding effect of patients with giant cell arteritis. However, the modest sample size of patients with ≥2 relapses or recurrences somewhat limited the statistical power of this analysis.

In conclusion, our study shows that despite its apparent control of clinical symptoms, corticosteroids do not adequately control the inflammatory process in a subset of patients with PMR identifiable by the persistent elevation of CRP and IL-6 levels after initiation of prednisone therapy. Recognizing this subset of patients with chronic, relapsing disease who require long-term corticosteroid treatment is essential to guiding treatment strategies.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES