Polymyositis (PM), dermatomyositis (DM), and sporadic inclusion body myositis (IBM) are 3 distinct idiopathic inflammatory myopathies that are characterized by a common histologic endomysial inflammation and distinct immune-mediated mechanisms (1). In PM and sporadic IBM, sensitized CD8+ cytotoxic T cells recognize as-yet-unidentified muscle antigens. The cytotoxic cells surround, invade, and destroy nonnecrotic muscle fibers that express class I major histocompatibility complex molecules (2). Muscle fiber necrosis occurs by a mechanism represented by the exocytosis granule model, in which toxic proteins, especially perforin and granzyme B, are released (3). DM differs from the other 2 diseases clinically, because of the characteristic rash that accompanies or often precedes the muscle weakness, and also differs immunopathologically. DM is characterized by an intramuscular microangiopathy, which is mediated by the complement C5b–C9 membranolytic attack complex, and which leads to the destruction of endothelial cells, loss of capillaries, muscle ischemia, muscle fiber necrosis, and perifascicular atrophy (1, 4, 5).
Despite their distinct characteristics, the treatment of PM and DM is similar (6). Numerous uncontrolled studies have used combinations of prednisone, methotrexate, azathioprine, cyclophosphamide, cyclosporine, and plasma exchange, but in 30–50% of patients, the patients' conditions resisted all therapies and the patients remained physically disabled (6). In sporadic IBM, no treatments have demonstrated significant beneficial effects (7).
High-dose intravenous immunoglobulin (IVIG) has been shown to be beneficial in a few open, prospective studies on adult and/or juvenile DM with short followup periods (8–22), and in 1 controlled crossover study (23). In DM, IVIG prevents activated complement from damaging cutaneous and muscular lesions, inhibits C3 uptake in the sera, reduces serum levels of the SC5b–9 complex, depletes C3b NEO (a neoantigen that is expressed on the surface of activated C3 component upon incorporation into immune complexes) in muscles, prevents membrane attack complex deposits from entering the endomysial capillaries, and restores the capillary network (23–25). However, in PM, no controlled studies or significant cohort studies have been published. We carried out an open, prospective study on the use of IVIG in the treatment of severe, active PM refractory to traditional treatments, and established the long-term efficacy and safety of the use of IVIG in this inflammatory myopathy.
PATIENTS AND METHODS
- Top of page
- PATIENTS AND METHODS
Patients. Our study included 180 patients with idiopathic inflammatory myopathies, of which 92 (47 with PM, 30 with IBM, and 15 with DM) were treated with IVIG as a first-, second-, or third-line therapy. The study comprised 35 adult white patients (20 women, 15 men) who were referred to us because of therapy-resistant PM and were treated with IVIG as a third-line therapy with a followup of over 3 years. All patients fulfilled the usual diagnostic criteria for PM (26). Muscle biopsy samples were systematically reviewed to exclude sporadic IBM. None of the patients had a history of malignant disease. Two patients had myositis associated with connective tissue disease (1 with sicca syndrome, 1 with thyroiditis). Patients with severe heart or renal failure, IgA deficiency (27), and pregnant women were excluded from the study. The mean age of the subjects was 43.5 years (SD 16.8). Eleven patients had an esophageal disorder (proximal dysphagia or uncoordinated peristalsis from striated muscle dysfunction). Raynaud's phenomenon was noted in 3 patients. Visceral disorders included 2 cases of pulmonary fibrosis and 2 cases of cardiomyopathy (1 dilated cardiomyopathy and 1 tachyarrhythmia).
The average duration of PM before IVIG was 3.0 years (SD 2.7 years, with a range of 6 months to 14 years). All 35 patients had not responded to the traditional therapies or had experienced side effects. These therapies were steroids (n = 35), methotrexate (n = 24), azathioprine (n = 13), cyclophosphamide (n = 4), cyclosporine (n = 7), chlorambucil (n = 1), plasmapheresis (n = 8), lymphopheresis (n = 1), and total body irradiation (n = 1). Steroids or immunosuppressant agents were given for at least 4–6 months. All patients received at least a combination of steroids and immunosuppressive drugs, including methotrexate or azathioprine, before the study.
The patients' treatment was not changed in the 2 months before the introduction of IVIG and the doses were not increased during gammaglobulin therapy. The mean length of time between the last time that the treatment was modified and the beginning of IVIG treatment was 5.1 months (SD 1.3). Despite these treatments, the disease was still clinically active: muscle strength, the muscle disability scale (MDS) score, and serum muscle enzyme levels were stable or worsened prior to IVIG treatment.
Treatment protocols. We used preparations of commercial, polyvalent human IVIG with an elevated proportion of intact plasma IgG. IVIG was administered slowly (<200 ml/hour) at 1 gm/kg body weight for 2 consecutive days per month, as described previously (9). The doses and immunosuppressive medication remained constant during the study period, as did routine daily activities. In the absence of clinical response, IVIG therapy was stopped after the third dose (8, 9). The patients who were considered responders received at least 6 doses of IVIG.
Clinical assessment. Clinical response was assessed before each IVIG infusion, according to 3 criteria.
Muscle power was gauged using a simple modification of the British Medical Research Council grading system (28), which assigns the numbers 0–5 to indicate the level of muscle power. To increase precision, each number was subdivided, resulting in a possible score of between 0 and 11, as previously described (9). The intensity of the muscular deficit was determined from the total quotation of 8 muscles: neck flexors, trapezius, deltoid, biceps, psoas, maximus and medius gluteus, and quadriceps. Since muscle weakness is symmetric in PM, only 1 side was tested. The theoretical maximum score was 88 points (normal muscle power). All patients were tested by the same independent examiner, who is a specialist in neuromuscular diseases. As described in our previous studies (8, 9, 29), IVIG treatment was considered to be successful if the muscle power score increased by ≥18 points compared with the initial score.
An MDS was constructed to explore both proximal strength (proximal upper limbs, proximal lower limbs) and axial and pharyngeal muscle disability in PM and DM. Our MDS included 18 items, as follows: limb musculature, with 4 items for the arms and 7 for the hips and legs, and axial and pharyngeal musculature, with 7 items. Each of the 18 items was scored on a 5-point Likert scale: for the limbs, 4 ratings ranged from 3 = impossible to 0 = no difficulty, whereas the ratings for axial muscles ranged from 6 to 0 to give these ratings double weight. This scale, which was valid and reliable (30) and could easily be completed within 5–10 minutes, was performed before each IVIG infusion by the same independent physician. Total scores (sum of all items) ranged from 75 (maximum disability) to 0 (no disability). IVIG treatment was considered to be successful if the MDS score decreased by ≥8 points compared with the initial MDS score.
Dysphagia and pharyngeal musculature were evaluated by use of specific clinical evaluation and esophageal manometric investigations.
Biochemical assessment. The biochemical assessment included the measurement of serum levels of muscle enzymes (creatine kinase [CK]; normal values <110 units/liter), complete blood cell counts, and the measurement of serum immunoglobulin levels before each IVIG infusion. All samples were obtained by the same laboratory with the use of standard techniques.
Statistical analysis. The mean muscle strength and the MDS scores of the 35 patients were assessed before and after each IVIG infusion. The Student's t-test was used to compare the muscle strength, CK values, and the mean reduction in steroid dose before and after IVIG treatment. The nonparametric Wilcoxon test was used to examine the MDS scores. The mean CK values obtained after each IVIG infusion, and the mean steroid dose reduction before and after IVIG therapy, were compared using Student's paired t-tests. Significance was assessed at P values equal to 0.05.
- Top of page
- PATIENTS AND METHODS
Clinical analysis.Muscle testing. Clinical improvement was observed in 25 of the 35 patients treated with IVIG (71.4%). The improvement in muscle power was between 18 points and 25 points for 15 patients, 26–35 points for 5 patients, and ≥36 points for 5 patients, compared with the initial muscle testing scores. The clinical assessment as gauged by the mean muscle power score showed a significant improvement after 3 IVIG infusions compared with before the initiation of the therapy (for the 35 PM patients, mean initial muscle power score 47.8 points [SD 11.1], mean muscle power score before the fourth IVIG infusion 67.4 points [SD 12.7]; P < 0.01) (see Figure 1A). Clinical improvement was usually noted within 3 IVIG infusions. The muscle power in 10 of the 25 patients whose condition improved almost returned to normal, and 15 had clinical improvement without normalization. A worsening of clinical status was observed in 1 patient during the IVIG infusions. No clinical improvement was noted in 10 patients, despite 3 IVIG infusions. The infusions were subsequently discontinued in these patients.
Figure 1. Clinical (A) and biologic (B) evolution of 35 patients with polymyositis receiving intravenous immunoglobulin (IVIG) therapy. Bars show the mean ± SD. CK = creatine kinase.
Download figure to PowerPoint
MDS. The MDS score was improved in 28 patients, including the 25 patients who also benefited from significant improvements in their muscle power. Proximal strength (proximal upper limbs, proximal lower limbs) as well as axial and pharyngeal muscle functions were improved by IVIG therapy. The mean MDS score for the 35 patients was significantly improved after 3 IVIG infusions compared with before the initiation of treatment (for the 35 PM patients, mean MDS score 22.3 points [SD 8.0], after 3 IVIG infusions 10.9 points [SD 6.1]; P < 0.01). Three patients showed improvement on the MDS (decrease in MDS score of >8 points), but their muscle power score did not improve significantly (clinical improvement <18 points compared with initial power).
Esophageal disorders. Esophageal disorders disappeared in 8 of the 11 patients, which was defined by the complete disappearance of swallowing disorders, nasal regurgitation, coughing while eating, pain and/or preprandial discomfort in the sternal area, and gastroesophageal reflux or aspiration infectious pneumonia secondary to reflux, and this was confirmed by control esophageal manometry. The condition of 1 patient was significantly improved, but the esophageal disorders did not disappear (persistance of gastroesophageal reflux).
Spare steroid doses. In those patients who demonstrated a response to IVIG therapy, steroid doses were progressively reduced, following a definite schedule (10% of the preceding dose every 10–15 days). This improvement meant that steroid doses could be significantly reduced after the third or the fourth IVIG infusion in all of the 22 patients who received steroids (mean initial steroid doses before IVIG therapy 32.7 mg/day [SD 17.2], mean steroid doses after 4 IVIG infusions 21.9 mg/day [SD 11.2]; P < 0.05).
Biochemical analysis. Serum CK levels were initially normal in 2 patients and remained unchanged during IVIG infusions. All of the 33 patients with initially elevated CK levels showed biochemical improvement. The mean CK levels for these patients decreased significantly, from 1,880 ± 620 units/liter before IVIG infusion to 520 ± 240 units/liter before the fourth IVIG infusion for the 33 patients (P < 0.01) (Figure 1B). CK levels usually dropped within 2 months of the first IVIG infusion. The serum CK levels of 19 patients returned to normal, and the other 14 patients showed a biologic improvement without normalization.
All patients with clinical improvement showed biochemical improvements or had normal initial CK serum levels. For these patients, an initial drop in the CK serum levels preceded improvements in muscular deficit. Serum immunoglobulin levels, which were determined before each IVIG infusion, remained stable in all patients.
Comparison of responders and nonresponders. There were no significant differences in the age, sex, clinical and biochemical status, prior therapy, or associated diseases between the group of patients who responded to the treatment and the group of those who did not respond (Table 1), except for the disease duration and the initial CK levels. The patients with the most major clinical and biochemical improvements had a shorter duration of disease. The mean (±SD) disease duration before IVIG therapy for the patients who responded to IVIG treatment was 26 ± 12 months. The patients who did not respond to IVIG therapy had a mean disease duration of 43 ± 16 months (P < 0.05). However, 2 patients, with a 4- and 6-year history of PM, showed dramatic improvement after IVIG infusions. We had no information about myositis-specific autoantibodies, and therefore no relationship between autoantibody status and responses to therapy could be determined. Mean serum CK levels in the responder group were much higher than those in the nonresponder group (Table 1).
Table 1. Comparison between the group of IVIG responders and IVIG nonresponders in 35 patients with polymyositis*
| ||Number of patients||Mean ± SD duration of disease, months||BMRC muscle testing score||MDS score||CK, units/liter||Mean dose of steroids, mg/day|
|Before IVIG||After IVIG||Before IVIG||After IVIG||Before IVIG||After IVIG||Before IVIG||After IVIG|
|IVIG responders||25||26 ± 12||45.3||69.8†||21.1||8.4†||2,010||420†||32.7||21.9†|
|IVIG nonresponders||10||43 ± 16†||42.9||56.6||24.2||16.7||1,650||710||29.8||25.8|
Tolerance. IVIG displays rare and benign side effects (28). In our patients, tolerance was generally excellent in 29 subjects and side effects were observed in the 6 other patients. These side effects included 4 cases of mild headache during or after infusions and 3 cases of fever with shivering and sweating (1 patient had first developed headaches). These adverse reactions disappeared spontaneously after the infusions were discontinued and did not recur during further IVIG infusions.
Followup (Figure2). The mean (±SD) followup period for the 25 patients with PM who responded well to IVIG treatment was 51.4 ± 13.1 months. Of these 25 patients, 12 remained in full remission (absence of myositis activity following discontinuation of all drug therapy) or showed a complete clinical response (absence of myositis activity with remaining treatment) following their initial course of IVIG; 5 of these patients were taking no medication during remission, while 7 patients were receiving low doses (<7 mg/day) of steroids with a mean (±SD) followup period of 39 ± 14.3 months after the discontinuation of IVIG therapy. These patients returned to normal muscle strength, normal muscle function, and normal serum muscle enzyme levels (the 5 patients in full remission) or experienced only minor clinical sequelae (mean ± SD muscle power score 82 ± 6 points) with normal serum muscle enzyme levels in the others.
Figure 2. Clinical evolution of the 25 intravenous immunoglobulin (IVIG) responders as gauged by the muscle strength score (A) and muscle disability scale (75 points = maximum disability, 0 points = no disability) (B), and effects of IVIG on the dose of spare steroids (n = 22). Bars show the mean ± SD.
Download figure to PowerPoint
Six patients remained dependent on IVIG infusions but did not require additional medication. IVIG dependence was defined as the deterioration of clinical and/or biochemical status during the weeks following the previous course of 6 IVIG infusions. To reduce the high costs of immunoglobulin maintenance therapy, the following IVIG treatment schedule was used: 6 monthly treatments with 2 gm/kg body weight IVIG, followed by monthly treatments with 1 gm/kg. The clinical and biochemical status of all patients receiving half doses remained stable, allowing long-term, persistent doses of spare steroids (Figure 2). These 6 patients received an average of 21 ± 7 doses of IVIG (6 full doses, then 15 ± 7 half doses of IVIG), with no tolerance problems resulting from long-term use.
Seven of the 25 patients who responded well initially relapsed after an average of 17.1 months (range 4–23 months) after the discontinuation of 6 IVIG infusions. These patients did not initially benefit from a degressive schedule of IVIG therapy with half doses. Four of them responded well to a new course of 6 IVIG infusions. The remaining 3 patients were prescribed additional immunosuppressive drugs, and subsequent IVIG infusions were not given.
- Top of page
- PATIENTS AND METHODS
Since the first studies about the use of IVIG for the treatment of idiopathic inflammatory myopathies (8, 9), few articles have been published on IVIG and myositis. The majority of these published reports were case reports or small, open studies about adult and/or juvenile DM (10–22), and only 1 was a controlled crossover study in adult DM (23).
In PM, no controlled studies or long-term studies with sufficiently large cohorts have been published. Our study showed that the administration of IVIG in steroid-resistant patients resulted in a significant and persistent improvement in their chronic PM, as defined by clinical and biochemical assessments. This benefit allowed the patients' prednisone dose to be reduced by >50% compared with their initial dose, after 6 months of IVIG infusions (Figure 2). Nearly half of the improved patients remained in full remission following their initial course of IVIG and were taking no medication or low doses (<7 mg/day) of steroids over the long-term followup period.
Many lines of evidence suggest that cellular autoimmune mechanisms are involved in the pathogenesis of PM (1–3). The muscle infiltrate contains mainly monocytes and lymphocytes, as well as perinecrotically distributed, activated cytotoxic T cells, natural killer cells, and macrophages (2). As previously noted, sensitized CD8+ cytotoxic T cells recognize muscle antigens that are yet to be identified. The cytotoxic cells surround the nonnecrotic muscle fibers and occasionally invade and destroy them, which leads to phagocytosis and fiber necrosis (2). In the exocytosis granule model, muscle fiber necrosis occurs via release of toxic proteins, especially perforin and granzyme B (3).
The modes of action of IVIG in myositis probably vary according to the type of myositis. In DM, IVIG acts by preventing activated complement from damaging cutaneous and muscular lesions, inhibiting C3 uptake in the sera, reducing serum levels of SC5b–9 complex, depleting muscle from C3b NEO and membrane attack complex deposits from the endomysial capillaries, and restoring the capillary network (23–25). IVIG treatment led to a decrease in class I major histocompatibility complex and intercellular adhesion molecule 1 staining, which suggests that IVIG binds to Fc receptors on macrophages and leads to a decreased production of pathologically important cytokines (25). Moreover, IVIG down-regulates transforming growth factor β1, which is involved in the pathogenesis of chronic inflammation, fibrosis, and the accumulation of extracellular matrix proteins in DM, but not in IBM (22). Some authors believe that soluble interleukin-2 receptor serum levels may be useful predictors of IVIG-induced treatment response and DM activity (21).
Recently, a unifying theory on the action of IVIG in autoantibody-mediated diseases has been reported (31). A new protein transport receptor for IgG, called FcRn, protects pinocytosed IgG from catabolism. IVIG treatment leads to a transient hypergammaglobulinemia and saturates FcRn. In this state, a higher proportion of endogenous, pathologic autoantibodies is available to be catabolized, leading to a reduction in autoantibodies and disease activity. Given the high frequency of autoantibodies seen in myositis (32), this mechanism may apply to PM. Interestingly, corticosteroids down-regulate the expression of FcRn messenger RNA, and may act by a similar mechanism (31).
However, in PM, the exact mode of action of IVIG has not been elucidated. Intravenous immune serum globulin is derived from large pools of serum from thousands of donors, providing a large range of antibodies. In addition to the recent theory described above, several mechanisms may be proposed and associated: 1) the blockade of IgG Fc receptors on phagocytic cells by the infused IVIG, thereby preventing the removal of cells, sensitized with autoantibodies of IgG isotype, from the circulation; 2) the presence of antiidiopathic antibodies capable of suppressing autoantibody secretion in the pool of plasma donors used for IVIG production; 3) a direct effect on B and T lymphocytes; 4) the inhibition of adhesion molecules and cytokine production; and/or 5) the presence of soluble CD4, CD8, and HLA in IVIG, which modify antigen recognition by target cells. These 2 last theories are probably important, since IVIG might also provide a source of antiidiotypic antibodies directed against chemokines, HLA antigens, adhesion molecules, and especially the selectin family, and antiapoptotic molecules and metalloproteases, which are known to be involved in the pathogenesis of myositis (22, 33–40).
In our study, PM was a chronic and refractory disease in all patients, despite specific combination therapy. There were no changes in the treatment in the 2 months before the introduction of IVIG and no increases in dose during gammaglobulin therapy. The mean time between the last dose modification of traditional therapy and IVIG infusions was 5.1 months (SD 1.3). Despite these therapies, PM remained clinically active, justifying the introduction of a new therapy.
Our study confirms the efficacy of IVIG in PM. Seventy percent of patients significantly improved clinically following immunoglobulin therapy. The MDS score in 3 of the 10 remaining patients partially improved, but their muscle power did not significantly improve. The benefit of IVIG in PM is similar to that observed in DM (23).
The relationship between serum autoantibody status and clinical response to IVIG is not known. Unfortunately, especially for the first patients included in the study, we had no information about myositis-specific autoantibodies, and no relationship between autoantibody status and responses to therapy can be established. We found 2 distinctions between the group of responders and the others. First, there is probably a relationship between disease duration and response to IVIG therapy. The patients with the most major clinical and biochemical improvements had had the disease for shorter periods of time (mean ± SD 26 ± 12 months versus 43 ± 16 months; P < 0.05). Similar to sporadic IBM, immunomodulatory agents will probably not improve strength or function in longstanding myositis that is characterized by major atrophy and muscle fat replacement. Probably for this same reason, the mean serum CK levels in the responder group were much higher than those in the nonresponder group (Table 1).
In our study, 28% of IVIG responder patients (7 of 25) relapsed, with a mean followup over 4 years. These results are significantly different from the rates of relapses reported in the literature (41, 42). Relapses occurred in 77% of PM patients during steroid withdrawal or periods of stable maintenance therapy (41). The frequency of relapses seems to be higher in PM patients than in those with DM or overlap syndromes (41).
Our data suggest that intravenous immunoglobulins are effective in the treatment of chronic, refractory PM. Immunoglobulins are safer and better tolerated than corticosteroids and other immunosuppressive drugs and, despite the cost, may be considered as a new steroid-sparing agent. The cost can be reduced by using half doses (1 gm/kg monthly) for effective long-term maintenance therapy. The appropriate therapeutic approach to treatment with IVIG, such as dose, duration of treatment, number of infusions, and modes of action, will be clarified in future multicenter studies.