Trallero-Araguás, Ernesto MD; Labrador-Horrillo, Moisés MD, PhD; Selva-O'Callaghan, Albert MD, PhD; Martínez, Maria Angeles PhD; Martínez-Gómez, Xavier MD; Palou, Eduard MD, PhD; Rodriguez-Sanchez, Jose Luis MD, PhD; Vilardell-Tarrés, Miquel MD, PhD
The association between cancer and idiopathic inflammatory myopathies has been extensively reported in the medical literature.16 Since historical population-based cohort studies confirmed an increased risk of cancer in patients with dermatomyositis (DM) and polymyositis (PM),7,9,73,21 diagnosing occult cancer in these patients has become a challenge for clinicians. Cancer can develop before, concurrently, or subsequent to the onset of idiopathic inflammatory myopathy. The overall cancer risk is reported to be higher in DM than in PM, particularly during the 3 years before or after the diagnosis of the idiopathic inflammatory myopathy. The incidence of cancer in published series of patients with idiopathic inflammatory myopathy ranges from 9% to 42%,7,9,13,21,27 and malignant disease is one of the main causes of mortality in this population.1,25 Cancer screening is now mandatory routine practice in patients with a diagnosis of myositis, but there is no consensus as to the method or frequency with which patients with an idiopathic inflammatory myopathy should be tested to rule out neoplasm during the follow-up.
Several authors have attempted to find positive and negative predictive markers of malignancy to identify subgroups of the idiopathic inflammatory myopathies with the highest risk of developing cancer. Certain characteristic clinical, analytical, and treatment response features have been proposed, with uncertain importance in clinical practice.2-4,10 In separate efforts using immunoprecipitation techniques, Targoff et al24 and Kaji et al15 described 2 new and similar myositis-specific autoantibodies-reactive with a 155-kDa nuclear protein and with 2 155/140-kDa nuclear proteins, respectively-that showed a significant association with CAM. This association has been confirmed by Chinoy et al8 and Gunawardena et al,12 who found the same anti-p155/140-kDa autoantibody previously described by Kaji et al. Although it has not been proven that anti-p155-kDa and anti-p155/140-kDa autoantibodies are, in fact, the same, it is broadly assumed that they are identical.11,17
Because of the importance of CAM in patients with idiopathic inflammatory myopathy and the potential significance of this new antibody in the management of this condition, we undertook the current study in our historical cohort of patients with idiopathic inflammatory myopathy to determine the incidence and clinical characteristics of CAM and the frequency of anti-p155, and to analyze the association between this autoantibody and cancer.
PATIENTS AND METHODS
This study was performed in adult patients with PM and DM from a historical cohort of patients diagnosed with idiopathic inflammatory myopathy at Vall d'Hebron General Hospital in Barcelona, Spain, between 1983 and 2007. Our center is a single teaching hospital with approximately 700 acute-care beds, attending a population of nearly 450,000 inhabitants. All myositis patients in this population are referred to Vall d'Hebron for diagnosis and therapy, regardless of the severity of the disease. Patients included in the study gave informed consent to the use of their serum for research purposes. Serum samples are routinely collected from these patients mainly at the diagnosis and also during follow-up in our outpatient clinic, and stored at −80°C. The study was approved by the institutional review board of our hospital.
The diagnosis of DM and PM was based on the criteria of Bohan and Peter,5,6 which include symmetrical proximal muscle weakness, increased serum muscle enzymes, electromyography abnormalities, typical histologic findings on muscle biopsy, and characteristic dermatologic manifestations (heliotrope rash and Gottron papules). Only patients with definite or probable disease were included. The Sontheimer criteria were used to diagnosis amyopathic DM.22 Patients who met the criteria for myositis overlap syndrome,26 sporadic inclusion body myositis,19 or juvenile dermatomyositis5,6 - diseases that have shown no clear, accepted relationship with cancer-were excluded.
Clinical and Laboratory Data
Clinical and laboratory data were obtained by taking a standardized clinical history and conducting a physical examination, and by reviewing the patient's medical records. The information was recorded in a database that included demographic data, signs and symptoms, laboratory findings, presence of associated conditions, systemic and organ involvement, clinical course, number and nature of treatments, and treatment responses. Creatine kinase peak was codified as low (<200 U/L), medium (200-1500 U/L), and high (>1500 U/L). Treatment response was defined as complete (patient totally recovered without evidence of active disease), partial (evidence of clinical improvement short of a complete clinical response), or none (no evidence of clinical improvement).
Cancer-associated myositis (CAM) was defined according to the modified Bohan and Peter classification26 as cancer occurring within 3 years of the myositis diagnosis, as well as the fact that if cancer was cured, myositis was also cured.
Stored serum samples from each patient were prospectively screened by immunoprecipitation assays using extracts of a cell line derived from cervical cancer (HeLa cells) labeled with 35S-protein, as previously described.14 Briefly, HeLa cells were allowed to incorporate 35S-methionine-cysteine (5 μCi/mL; Amersham, Arlington Heights, IL) in a methionine-cysteine-free minimal essential medium (Sigma, St. Louis, MO) supplemented with 1% glutamine and antibiotics for 16-18 hours. Labeled cell extracts were then obtained by ultrasound exposure. To detect autoantibodies, 10-15 μL of serum sample from each patient was incubated with 2 mg of protein-A-Sepharose CL-4B (Amersham Pharmacia Biotech, Uppsala, Sweden) in 500 μL of immunoprecipitation buffer (10 mM Tris HCl, pH 8.0, 500 mM NaCl, 0.1% Nonidet P40) for 2 hours at 4°C, and then washed 5 times with immunoprecipitation buffer. Subsequently, 100 μL of 35S-methionine-cysteine-labeled HeLa cell extracts were mixed with antibody-coated Sepharose beads, rotated at 4°C for 2 hours, and washed. Immunoprecipitated proteins were separated by 10% sodium dodecyl sulphate-polyacrylamide gel electrophoresis and localized in the gel by autoradiography. The prototype anti-p155-kDa sera were provided by the Oklahoma Medical Research Foundation (Oklahoma City, OK).
Data on the presence of other myositis-specific autoantibodies and myositis-associated autoantibodies in these patients have been described previously.20 Antibodies directed against extractable nuclear antigens (Ro, La, RNP, Sm) and anti-histidyl-tRNA synthetase (anti-Jo-1) were screened by enzyme-linked immunosorbent assay (ELISA). Sera were tested by protein and RNA immunoprecipitation, which enabled detection of other synthetases, myositis-specific autoantibodies, and myositis-associated autoantibodies (anti-Mi-2, anti-SRP, anti-Ro 52, anti-Ro 60, anti-La, anti-PM-Scl, and anti-RNP) that may have been overlooked by ELISA.
In 59 of the myositis patients, genomic DNA was extracted, and HLA-DRB1 and DQA1 class II allele typing was performed by polymerase chain reaction-reverse sequence specific oligonucleotide (PCR-SSOr) using commercial kits (Tepnel Lifecodes, Stamford, CT) and Luminex xMAP technology.
A descriptive analysis was performed on the demographic data and some clinical features. Percentages (with 95% confidence intervals [CIs] for the main results) were used to describe the qualitative data (sex, myositis treatment response, creatine kinase peak, presence of CAM, anti-p155 autoantibody, dysphagia, interstitial lung disease, skin necrosis, Gottron papules, V sign, shawl sign, and heliotrope rash), and the median with the interquartile range was used to describe quantitative data (age at myositis onset). Possible factors associated with the presence of CAM and anti-p155 autoantibody were investigated. The Fisher exact test or the chi-square test, when appropriate, were used to assess the relationships between the presence of CAM or anti-p155 autoantibody and the qualitative variables, and the Mann-Whitney U test was performed to determine relationships with the quantitative variables. The Holm procedure was used to adjust p values to minimize the effect of multiple comparisons. The strength of the associations between variables was measured using odds ratios (ORs) with 95% CIs. The positive and negative predictive values of anti-p155 autoantibody for CAM were calculated. Statistical significance was set at p < 0.05, and all analyses were performed with SPSS, version 15.0 (SPSS, Chicago, IL).
Eighty-five patients, 20 PM and 65 DM, were included in the study: 78 patients met the criteria for definite DM/PM, 5 for probable DM/PM, and 2 for amyopathic DM according to the Bohan and Peter, and Sontheimer criteria.5,6,22 The characteristics of the series are shown in Table 1. A distinct 155-kDa protein was detected by immunoprecipitation assay in some patients, occasionally accompanied by a weaker, 140-kDa protein band (Figure 1). The characteristics of CAM patients and the immunoprecipitation results are described below.
Cancer and Myositis
Cancer was diagnosed in 24 patients (21 DM and 3 PM), with 16 cases classified as CAM (14 DM and 2 PM). The characteristics of patients with cancer are shown in Table 2. Cancer site distribution in patients with CAM was as follows: breast (n = 4), ovarian (n = 3), lung (n = 2), colon (n = 2), cholangiocarcinoma (n = 1), cervix (n = 1), gastric (n = 1), lymphoma (n = 1), and pancreas (n = 1). Adenocarcinoma was the most common type of cancer recorded in CAM (87.5%). In the CAM group, 6 cancers were diagnosed at the same time as myositis, 6 were diagnosed during the first year before or after myositis onset, 2 between the first and second year, and 2 between the second and third year. Cancer preceded myositis in only 2 patients with CAM, by 8 months in 1 (Patient 7) and 34 months in the other (Patient 10).
The clinical and laboratory characteristics of patients with CAM were assessed comparing this group with no-CAM patients. No significant difference in sex distribution was found between the CAM group (women: 75%) and no-CAM group (women: 75.3%) (p = 1). Median age at time of myositis onset was higher in the CAM group (63 yr) than in the no-CAM group (51 yr) when both PM and DM patients were analyzed together, but the difference was not statistically significant (p = 0.18). Similarly, no statistical association was found when the analysis was performed in the DM group alone (60.4 vs. 53.2 yr, p = 1). In DM patients, the shawl sign was the only clinical characteristic seen more often in patients with CAM than in those without (p < 0.01). There were no statistical associations in either the overall group or in the separate idiopathic inflammatory myopathies (DM and PM) between the CAM and no-CAM groups when other signs and symptoms were analyzed, including peak creatine kinase, dysphagia, interstitial lung disease, and treatment response, or when dermatologic features were analyzed in DM (skin necrosis, heliotrope rash, Gottron sign, and V sign).
Anti-p155 Autoantibody Detection and Clinical Characteristics
Anti-p155 autoantibody was detected in 16 of the 85 patients with idiopathic inflammatory myopathy (19%; 95% CI, 3.4%-39.6%): 1 with PM and 15 with DM. Based on this finding, the clinical and laboratory features of patients with anti-p155 were analyzed exclusively in the DM group (Table 3). No statistical association was found between anti-p155 autoantibody and sex or age at myositis diagnosis. In contrast to anti-p155-negative DM patients, anti-p155-positive patients had the shawl sign and V sign more frequently, but differences were not significant (71% vs. 22% and 80% vs. 46%, respectively). Nonetheless, anti-p155-positive patients had a significantly lower incidence of interstitial lung disease (0% vs. 42%, p = 0.01).
Relationship Between Anti-p155 Antibody and CAM
Among the 85 patients with idiopathic inflammatory myopathy, anti-p155 autoantibody was found in 10 of the 16 patients classified as having CAM (62.5%; 95% CI, 35.4%-84.8%) but in only 6 of the 69 no-CAM patients (8.7%; 95% CI, 3.3%-18%) (p < 0.01). Neither of the 2 patients with PM and CAM had anti-p155 autoantibody. Anti-p155 was present in 10 of the 14 DM patients with CAM (71.4%) and in 5 of the 51 DM patients without CAM (9.8%) (OR, 23; 95% CI, 5.2-101.2; p < 0.001).
Among the 16 anti-p155-positive patients, cancer was diagnosed in 11 patients, with 10 cases classified as CAM (62.5%). In DM, the negative and positive predictive value of presence of the anti-p155 autoantibody for a diagnosis of CAM was 92% and 66.7%, respectively (Table 4).
Myositis-Specific and Myositis-Associated Autoantibodies
Myositis-specific autoantibodies, including any anti-synthetase autoantibodies (Jo-1, PL-7, PL-12, EJ, OJ, and KS), anti-Mi2, or anti-SRP antibodies, were present in 31 of 85 patients with idiopathic inflammatory myopathy (36.5%; 95% CI, 26.3%-47.6%), and in 22 of 65 patients with DM (33.8%; 95% CI, 22.6%-46.7%). No myositis-specific autoantibodies were detected in patients positive for anti-p155 autoantibody. In contrast, myositis-associated autoantibodies were identified in some anti-p155-positive patients (see Table 2). Among patients with CAM and negative anti-p155 autoantibody, the myositis-specific autoantibodies anti-Jo-1 and anti-Mi2 were detected separately in 2 patients (Patients 13 and 15, respectively) (see Table 2). The diagnostic accuracy of myositis-specific autoantibodies and myositis-associated autoantibodies (including anti-U-RNP, anti-Ku, anti-PM-Scl and anti-Ro) was assessed by calculating their negative and positive predictive values. In our cohort, a result negative for myositis-specific autoantibodies/myositis-associated autoantibodies had a positive predictive value of 27.8% and a negative predictive value of 89.6% for the diagnosis of CAM (31% and 86.1%, respectively, in the DM group). Use of a combined strategy in DM patients, consisting of an absence of all myositis-specific autoantibodies/myositis-associated autoantibodies analyzed or positive status to anti-p155 autoantibody, yielded a positive predictive value of 34.5% and a negative predictive value of 91% (see Table 4).
None of the anti-p155 autoantibody-positive patients presented the DQA1*0102 allele, compared with 38% of the anti-p155-negative patients (p = 0.01; OR, 15.7). We did not find the reported association24 between DQA1*0301 and presence of anti-p155 in our population, although a similar trend was observed: 16.7% of anti-p155 autoantibody-positive patients presented the DQA1*0301 allele compared with 8.3% of anti-p155-negative patients.
In this report we analyze the presence of cancer and anti-p155 autoantibody in a large cohort of patients with inflammatory idiopathic myositis and describe a number of clinical, serologic, and laboratory features of CAM and p155-positive patients.
The incidence of CAM in the current series was 19% in the complete cohort of patients with idiopathic inflammatory myopathy and 22% in the DM group. As has been reported, adenocarcinoma was the most frequent type of idiopathic inflammatory myopathy-related cancer,3,13 distributed in a broad spectrum of locations. In contrast to other studies,4,13 no association was found between age or sex and CAM, either in the overall cohort (PM/DM) or in the DM group alone. The small number of cases of CAM in PM patients precluded a separate statistical analysis in this group. Attending to the clinical and laboratory features examined, only the shawl sign showed a significant association with CAM in DM (p < 0.01).
The present study provides further evidence of the clear association between CAM and autoantibody against a 155-kDa protein in DM patients, as was previously described by Targoff et al.24 The increased risk of developing CAM in anti-p155-positive patients compared to the anti-p155-negative group was significant, with an OR of 23 (95% CI, 5.23-101.2; p < 0.001). The utility of the anti-p155 autoantibody was mainly supported by its high negative predictive value for CAM, which was 92% in DM patients. Similar results have been obtained by other authors (Table 5).
A 140-kDa protein was identified in 9 of the 16 patients with anti-p155 but, in all cases, it appeared as a weaker and apparently irrelevant protein band (data not shown). Despite their differing immunoprecipitation patterns, there is general agreement in considering anti-p155-kDa and anti-p155/140-kDa as largely identical autoantibodies because of their similar association with CAM and with certain clinical features.11,17 Both have been related to some dermatologic signs and to the absence of interstitial lung disease.15,24 Our study supports these reported associations, although the higher incidence of shawl sign and V sign in anti-p155-positive patients was not statistically significant when adjusted by the Holm procedure. We believe that the differences in protein patterns may be explained in part by the use of a different antigen source in the immunoprecipitation assays and by technical differences. Whereas Kaji, Chinoy, and Gunawardena,8,11,15 who described anti-p155/140 but not anti-p155 antibody, all used extracts from the K562 leukemia cell line as the antigen source in their immunoprecipitation experiments, both Targoff and our group used extracts from HeLa cells, a line derived from cervical cancer. The possibility that the antigen source can influence protein patterns in immunoprecipitation has been reported previously.18
In contrast to the findings of other authors,24 there was no significant association between the DQA1*0301 HLA alleles and presence of anti-p155 autoantibody in the current study, although we did detect a nonsignificant trend to a greater frequency of the DQA1*0301 allele in anti-p155-positive patients. The lack of a significant association in our population cannot be attributed to differences in population allelic frequencies, since the DQA1*0301 allele is found in about 10% of the white population of both the United States and Spain. We also documented an absence of the DQA1*0102 allele in all anti-p155-positive patients, which may indicate that this particular allele is a protective factor for development of the anti-p155 autoantibody.
To our knowledge, this is the first study in which a positive 155-kDa protein band has been described in a patient with PM, and in this case, unrelated with CAM. Anti-p155 autoantibody has been described in adult and juvenile DM, in some adult and juvenile connective tissue disease-associated myositis, and in 1 case of systemic lupus erythematosus, but not in PM patients.24 Despite evidence that anti-p155 autoantibody is not confined to DM patients alone, we believe that, in practice, anti-p155 should be considered a DM-specific autoantibody. Thus, the predictive value of this antibody in CAM should be interpreted only in adult DM patients. Its significance in juvenile DM remains uncertain.23
One limitation of the current study is related to the fact that there are no definitive criteria to establish the diagnosis of CAM. The concept of CAM is based on the idea that myositis is a paraneoplastic phenomenon. Currently, CAM is a consensus diagnosis defined according to data from historical population-based cohort studies that quantified the increased risk of cancer in idiopathic inflammatory myopathy over the time before and after myositis onset. The limit of 3 years around the diagnosis of myositis was established by consensus, but it cannot guarantee a real association between the 2 conditions. Some studies have shown an increased risk of cancer even 4 or 5 years after myositis onset.27 This issue could constitute a potential bias when analyzing the true value of anti-p155 autoantibody as a marker of paraneoplastic myositis. In addition, it should be mentioned that the incidence of CAM may have been underestimated in the current study. Four patients with PM and 15 with DM had a follow-up of less than 3 years, including 2 patients with DM and anti-p155 autoantibody who had a follow-up of 1.96 and 1.95 yr, respectively. All these patients were included in the non-CAM group, although the development of CAM could not be strictly ruled out. The median follow-up (interquartile range) of these 19 patients was 1.9 years (range, 0.6-2.3 yr). To minimize the potential bias of this factor, we repeated the statistical analysis excluding these 19 patients (data not shown), and no differences were found with regard to the previous results. Although it was not mentioned in any of the earlier reports, other authors may have had a similar problem. As described in the study, 10% of DM patients in the study by Chinoy et al8 had a follow-up shorter than 3 years, and only 58% of the 52 DM patients included in the study by Kaji et al15 had been followed for longer than 2 years.
Chinoy et al8 suggested that a serologic strategy combining positive status to anti-p155 or an absence of other myositis-specific/myositis-associated autoantibodies was a better approach to identify DM patients with CAM than anti-p155 alone. Their analysis included only myositis-specific/myositis-associated autoantibodies routinely analyzed in the local immunology laboratory: anti-Jo-1, anti-U1-RNP, anti-U3-RNP, anti-Ku, and anti-PM-Scl. We performed a similar analysis in our cohort using a larger group of myositis-specific/myositis-associated autoantibodies that included other anti-synthetase autoantibodies (PL-7, PL-12, EJ, OJ, and KS), as well as anti-Mi2, anti-SRP, and anti-Ro, in addition to those mentioned above. We found that neither this approach nor a second analysis that included only the myositis-specific/myositis-associated autoantibodies determined by Chinoy (data not shown) resulted in higher predictive values than those obtained with assessment of anti-p155 autoantibody alone. Rather than attributing this discrepancy to the differing myositis-specific/myositis-associated autoantibodies analyzed in the 2 studies, we believe it is the scant number of cases of CAM in the 2 series, which limits the power of the statistical analyses, that likely explains these differences.
Our observations provide further evidence that anti-p155 autoantibody is the best available serologic marker for CAM. Negative status to this antibody helps to rule out cancer because of its high negative predictive value, and alternatively, the presence of this autoantibody would be useful for selecting a subgroup of patients with a higher risk of CAM, in whom follow-up and complementary examinations should be more exhaustive and are likely to be more fruitful. Efforts directed toward better management of CAM should include prospective studies investigating this new serologic autoantibody.
We thank Professor Ira N. Targoff (University of Oklahoma, OK) for kindly providing the anti-p155 reference sera.
1. Airio A, Kautiainen H, Hakala M. Prognosis and mortality of polymyositis and dermatomyositis patients. Clin Rheumatol
2. Amoura Z, Duhaut P, Huong DL, et al. Tumor antigen markers for the detection of solid cancers in inflammatory myopathies. Cancer Epidemiol Biomarkers Prev
3. Andras C, Ponyi A, Constantin T, et al. Dermatomyositis and polymyositis associated with malignancy: a 21-year retrospective study. J Rheumatol
4. Basset-Seguin N, Roujeau JC, Gherardi R, Guillaume JC, Revuz J, Touraine R. Prognostic factors and predictive signs of malignancy in adult dermatomyositis. A study of 32 cases. Arch Dermatol
5. Bohan A, Peter JB. Polymyositis and dermatomyositis (first of two parts). N Engl J Med
6. Bohan A, Peter JB. Polymyositis and dermatomyositis (second of two parts). N Engl J Med
7. Buchbinder R, Forbes A, Hall S, Dennett X, Giles G. Incidence of malignant disease in biopsy-proven inflammatory myopathy. A population-based cohort study. Ann Intern Med
8. Chinoy H, Fertig N, Oddis CV, Ollier WE, Cooper RG. The diagnostic utility of myositis autoantibody testing for predicting the risk of cancer-associated myositis. Ann Rheum Dis
9. Chow WH, Gridley G, Mellemkjaer L, McLaughlin JK, Olsen JH, Fraumeni JF Jr. Cancer risk following polymyositis and dermatomyositis: a nationwide cohort study in Denmark. Cancer Causes Control
10. Fardet L, Dupuy A, Gain M, et al. Factors associated with underlying malignancy in a retrospective cohort of 121 patients with dermatomyositis. Medicine (Baltimore)
11. Gunawardena H, Betteridge ZE, McHugh NJ. Newly identified autoantibodies: relationship to idiopathic inflammatory myopathy subsets and pathogenesis. Curr Opin Rheumatol
12. Gunawardena H, Wedderburn LR, North J, et al. Clinical associations of autoantibodies to a p155/140 kDa doublet protein in juvenile dermatomyositis. Rheumatology (Oxford)
13. Hill CL, Zhang Y, Sigurgeirsson B, et al. Frequency of specific cancer types in dermatomyositis and polymyositis: a population-based study. Lancet
14. Hirakata M, Suwa A, Nagai S, et al. Anti-KS: identification of autoantibodies to asparaginyl-transfer RNA synthetase associated with interstitial lung disease. J Immunol
15. Kaji K, Fujimoto M, Hasegawa M, et al. Identification of a novel autoantibody reactive with 155 and 140 kDa nuclear proteins in patients with dermatomyositis: an association with malignancy. Rheumatology (Oxford)
16. Levine SM. Cancer and myositis: new insights into an old association. Curr Opin Rheumatol
17. Madan V, Chinoy H, Griffiths CE, Cooper RG. Defining cancer risk in dermatomyositis. Part II. Assessing diagnostic usefulness of myositis serology. Clin Exp Dermatol
18. Mimori T, Imura Y, Nakashima R, Yoshifuji H. Autoantibodies in idiopathic inflammatory myopathy: an update on clinical and pathophysiological significance. Curr Opin Rheumatol
19. Needham M, Mastaglia FL. Inclusion body myositis: current pathogenetic concepts and diagnostic and therapeutic approaches. Lancet Neurol
20. Selva-O'Callaghan A, Labrador-Horrillo M, Solans-Laque R, Simeon-Aznar CP, Martinez-Gomez X, Vilardell-Tarrés M. Myositis-specific and myositis-associated antibodies in a series of eighty-eight Mediterranean patients with idiopathic inflammatory myopathy. Arthritis Rheum
21. Sigurgeirsson B, Lindelof B, Edhag O, Allander E. Risk of cancer in patients with dermatomyositis or polymyositis. A population-based study. N Engl J Med
22. Sontheimer RD. Cutaneous features of classic dermatomyositis and amyopathic dermatomyositis. Curr Opin Rheumatol
23. Targoff IN. Autoantibodies and their significance in myositis. Curr Rheumatol Rep
24. Targoff IN, Mamyrova G, Trieu EP, et al. A novel autoantibody to a 155-kd protein is associated with dermatomyositis. Arthritis Rheum
25. Torres C, Belmonte R, Carmona L, et al. Survival, mortality and causes of death in inflammatory myopathies. Autoimmunity
26. Troyanov Y, Targoff IN, Tremblay JL, Goulet JR, Raymond Y, Senecal JL. Novel classification of idiopathic inflammatory myopathies based on overlap syndrome features and autoantibodies: analysis of 100 French Canadian patients. Medicine (Baltimore)
27. Zantos D, Zhang Y, Felson D. The overall and temporal association of cancer with polymyositis and dermatomyositis. J Rheumatol
© 2010 Lippincott Williams & Wilkins, Inc.