ORIGINAL RESEARCH published: 01 February 2021 doi: 10.3389/fimmu.2020.625896 Anti-TIF-1g Antibody Detection Using a Commercial Kit vs In-House Immunoblot: Usefulness in Clinical Practice Anaı́s Mariscal 1*, Milena Milán 2, Andrés Baucells 1, Maria Angeles Martı́nez 1, Andrea Garcia Guillen 2, Ernesto Trallero-Araguás 3, Marcelo Alvarado-Cardenas 4, Laura Martı́nez-Martı́nez 1,5, Leticia Alserawan 1, Teresa Franco-Leyva 1, Edited by: Marı́a Teresa Sanz-Martı́nez 6, Laura Viñas-Giménez 6, Hector Corominas 2, Lisa G. Rider, Cándido Juárez 1,5, Iván Castellvı́ 2,5 and Albert Selva-O’Callaghan 4,5 National Institute of Environmental 1 Immunology Department, Hospital de la Santa Creu i Sant Pau, Biomedical Research Institute Sant Pau (IIB Sant Pau), Health Sciences (NIEHS), Barcelona, Spain, 2 Rheumatology Department, Hospital de la Santa Creu i Sant Pau, Biomedical Research Institute Sant United States Pau (IIB Sant Pau), Barcelona, Spain, 3 Rheumatology Unit, Hospital Vall d’Hebrón, Barcelona, Spain, 4 Internal Medicine Reviewed by: Department, Hospital Vall d’Hebrón, Barcelona, Spain, 5 Universitat Autónoma de Barcelona, Barcelona, Spain, Manabu Fujimoto, 6 Immunology Department, Hospital Vall d’Hebrón, Barcelona, Spain Osaka University, Japan Neil John McHugh, University of Bath, United Kingdom Objectives: Anti-TIF-1g autoantibody detection is important for cancer screening in *Correspondence: patients with dermatomyositis. The gold standard for anti-TIF-1g detection, Anaı´s Mariscal immunoprecipitation, is only available from a few specialized laboratories worldwide, so amariscal@santpau.cat commercial ELISA/immunoblot tests have emerged in recent years. To analyze their Specialty section: usefulness in diagnosing cancer-associated dermatomyositis, we compared Euroimmun This article was submitted to Euroline profile with our previously validated in-house immunoblot assay with human Autoimmune and Autoinflammatory Disorders, recombinant TIF-1g. a section of the journal Methods: We included 308 adult patients from Hospital de la Santa Creu I Sant Pau and Frontiers in Immunology Vall Hebró n Hospital (Barcelona, Spain) tested for anti-TIF-1g autoantibodies using the Received: 04 November 2020 Accepted: 14 December 2020 Euroline profile and an in-house immunoblot assay. Published: 01 February 2021 Results: A total of 27 anti-TIF-1g were detected by the Euroline and 12 by the in-house Citation: Mariscal A, Milán M, assay. Fair agreement was observed between Euroline and the in-house immunoblot Baucells A, Martı´nez MA, Cohen’s kappa 0.3163. Expected prevalence of anti-TIF-1g autoantibodies was observed Guillen AG, Trallero-Araguás E, for the two methods for dermatomyositis and undifferentiated connective tissue diseases, Alvarado-Cardenas M, Martı´nez-Martı´nez L, Alserawan L, but unexpectedly high prevalence of anti-TIF-1g autoantibodies was detected by Euroline Franco-Leyva T, Sanz-Martı´nez MT, compared to the in-house immunoblot for other diseases (16.5% Euroline vs 0.8% in- Viñas-Giménez L, Corominas H, Juárez C, Castellvı´ I and house immunoblot, p<0.01). The in-house IB compared to Euroline more reliably detected Selva-O’Callaghan A (2021) cancer in patients with DM with anti-TIF-1g antibodies (p=0.0014 vs p=0.0502 for in- Anti-TIF-1g Antibody Detection house immunoblot vs Euroline). Using a Commercial Kit vs In-House Immunoblot: Conclusion: We recommend using a second validated method to confirm Euroline- Usefulness in Clinical Practice. Front. Immunol. 11:625896. detected anti-TIF-1g antibodies when the dermatomyositis diagnosis is not definitive. doi: 10.3389/fimmu.2020.625896 Furthermore, in the context of definite DM diagnosis with negative anti-TIF-1g antibodies Frontiers in Immunology | www.frontiersin.org 1 February 2021 | Volume 11 | Article 625896 Mariscal et al. Myositis Antibody Testing by Euroline and no other myositis specific antibody, is also recommendable to confirm by a second validated method. Keywords: dermatomyositis, cancer, autoantibody, diagnosis, immunoassay INTRODUCTION PATIENTS AND METHODS Anti-TIF-1g autoantibodies (also anti-TRIM33 or anti-p155/140 To analyze agreement between the methods, we reviewed all autoantibodies) were first described in 2006 when a large cohort Euroline and in-house IB results from 2014 to 2020 held by the of patients with idiopathic inflammatory myopathies was Immunology Department of Hospital de la Santa Creu i Sant Pau examined by immunoprecipitation (IP) (1). In patients with (Barcelona, Spain). We included 266 adult patients, and the dermatomyositis (DM), the prevalence of anti-TIF-1g diagnosis was categorized as follows: DM (21 patients), autoantibodies ranged from 14% (2) to 33% (1). undifferentiated connective tissue diseases (UCTDs) (127 There is a strikingly increased risk of cancer in patients with patients), and other diseases (118 patients). The UCTD group myositis compared to the general population (3), with 32% of was composed mostly of patients with Raynaud phenomenon, dry patients with DM reported to have an associated cancer eyes or mouth or idiopathic inflammatory myopathies other than diagnosis at some point during their illness (4). It is believed DM. The group categorized as other diseases included mostly that around one third of DM cases are paraneoplastic (5). Anti- patients with muscular involvement and/or hyperCKemia. TIF-1g autoantibodies are associated with DM and have been To analyze the prevalence of anti-TIF-1g antibodies as shown to be the best available biomarker for cancer-associated detected by the two methods in a cohort of DM patients and DM (CAM) (1, 6, 7). In fact, a patient with DM who is positive the usefulness of the methods in diagnosing CAM, we included for anti-TIF-1g autoantibodies has 27 times greater odds of an additional 42 patients with DM (total sample 308 patients) having CAM compared to a patient with DM negative for anti- from Vall Hebró n Hospital (Barcelona, Spain). TIF-1g (7). Definitive DM diagnosis was established according to Bohan Gold standard technique for anti-TIF-1g autoantibody and Peter criteria (16, 17) or a score of >90% according to the detection is the visualization of a 155 kDa band by K562 or International Myositis Classification Criteria (18). Patients with DM Hela cells IP. However, the methodology is lengthy and complex were further categorized as having CAM if, 3 years before or after and requires special facilities and trained staff for both technical the diagnosis of DM, they were diagnosed with a subjacent tumor. and interpretation aspects. Protein-based assays—either ELISA or immunoblot (IB)—have the advantage, compared to IP, of Anti-TIF-1g Determination unequivocal results specificity given that the substrate is a known Euroline Autoimmune Inflammatory Myopathies 16 Ag (IgG) protein; they therefore avoid the uncertainty associated with IP Profile (Euroimmun, Lübeck, Germany) was performed semi- in identifying different proteins with similar molecular weights. automatically using EUROBlotMaster from Euroimmun In-house ELISA and IB assays with human recombinant TIF-1g following manufacturer instructions. Immunoblots were graded have been independently developed and validated by two groups as negative or positive by at least two independent experienced (8, 9), while a commercial ELISA for anti-TIF-1g developed by observers on the basis of signal intensity. Medical and Biological Laboratories (MBL; Nagoya, Japan) has The in-house IB with human recombinant TIF-1g was been validated against IP (10). Three commercial IB assays are performed as follows. Purified recombinant protein (4 µg) also available for myositis-specific antibody detection, including encoding the longest isoform of TIF-1g (OriGene, Rockville, anti-TIF-1g determination, namely, D-Tek Myositis 12 IgG Dot Maryland, USA) was run on 4-12% polyacrylamide Criterion™ for BlueDiver (Diagnostic Technology, Belrose, NSW, Australia), Tris-Glycine Extended (TGX) Stain-Free™ precast gels for Euroimmun Euroline Autoimmune Inflammatory Myopathies PAGE (Bio-Rad, Munich, Germany) with Tris-Glycine-SDS (Euroimmun, Lübeck, Germany) and Trinity Biotech running buffer using the Bio-Rad criterion electrophoresis ImmcoStripe Myositis Advanced LIA (Buffalo, New York, system. Western blotting was performed using the Bio-Rad USA). However, although concordance studies exists for the trans-blot turbo transfer system on a nitrocellulose membrane. three assays (11) and for D-Tek and Euroline with IP (12–15), TIF-1g-transferred nitrocellulose was incubated for 1h at room there are no reports of the usefulness of those assays in temperature (RT), in phosphate buffered saline containing 3% diagnosing CAM in a large cohort of patients. non-fat dry milk (blocking buffer) and maintained at -20°C until We analyzed agreement in anti-TIF-1g antibody detection for use. For testing, nitrocellulose was vertically cut into strips and the Euroline profile compared with our in-house IB assay with incubated with human serum samples diluted 1:100 in blocking human recombinant TIF-1g which, in a previous study, showed buffer for 1 h at RT with shaking. After washing, phosphatase very good kappa agreement with IP (8). We also analyzed alkaline-labeled goat anti-human IgG antibody (Invitrogen, prevalence of anti-TIF-1g antibodies as detected by the two USA), at 1:1000 dilution in blocking buffer, was added to each methods in a cohort of DM patients and the usefulness of the strip and incubated 1h at RT with shaking. Color development methods in diagnosing CAM. was with phosphatase reagent (BCIP/NBT, Sigma-Aldrich, St Frontiers in Immunology | www.frontiersin.org 2 February 2021 | Volume 11 | Article 625896 Mariscal et al. Myositis Antibody Testing Louis, Missouri, USA). Positive and negative controls were TABLE 2 | Anti-TIF-1g antibody detection by Euroline and in-house immunoblot by diagnostic group. included in all batches. Immunoblots were graded, by at least two independent experienced observers, as negative or positive Diagnosis (patients/sample) Euroline In-house immunoblot p based on signal intensity at 155 kDa molecular weight. DM (63/308) 16 (25.4%) 22 (34.9%) 0.3318 Determinations of anti-TIF-1g were made as routine testing at UCTDs (127/308) 3 (2.4%) 6 (4.7%) 0.4999 the time of request. Samples were tested by the two methods at the Other (118/308) 17 (14.4%) 1 (0.8%) 0.0001 same time point, avoiding freezing cycles. Results of each assay type were interpreted without knowledge of the results of the other assay. (21%) had anti-TIF-1g antibodies detected by Euroline, and in- house IB, respectively. Statistical Analysis In diagnosing CAM, Euroline sensitivity and specificity values Cohen’s kappa coefficient was used to measure agreement were 50% and 83%, respectively, and in-house IB sensitivity and between the Euroline and in-house IB results. Fisher’s exact specificity values were 75% and 79%, respectively. The AUC test was used to compare sensitivity, specificity and positive and values were 0.665 (95% CI: 0.5–0.83) for Euroline and 0.769 negative predictive values (PPV and NPV, respectively) for the (95% CI: 0.628–0.91) for the in-house IB (Figure 1). The in- tests. The Delong et al. (1988) method was used to calculate the house IB compared to Euroline more reliably detected cancer in standard error for the area under the receiver operating patients with DM with anti-TIF-1g antibodies (p=0.0014 vs characteristic (ROC) curve (AUC) and the difference between p=0.0502 for in-house IB vs Euroline). two AUCs. All tests were two-sided, and probability values of Table 3 summarizes sensitivity and specificity values, PPVs, p<0.05 were considered statistically significant. Analyses were and NPVs for CAM diagnosed based on anti-TIF-1g antibody performed using SPSS V.21.0 for Cohen’s kappa coefficient, detection by Euroline and the in-house IB. Graphpad Prism 7 for Fisher’s exact test, and MedCalc statistical software for the Delong et al. method. DISCUSSION RESULTS We report results for a comparison between a commercial anti- Anti-TIF-1g Detection by Euroline vs TIF-1g assay (Euroline) and an in-house anti-TIF-1g assay (in- In-House IB house IB)—previously validated using a gold standard method, A total of 266 patients were included in this part of the study. namely IP (8)—based on determining sensitivity, specificity, Agreement between Euroline and in-house IB assays is PPV, and NPV for a large cohort of patients. summarized in Table 1. The two methods agreed in 90.6% (241/ Agreement between the Euroline and the in-house IB 266) of cases, yielding fair agreement according to Cohen’s kappa methods was, on the whole, fair (k=0.3163; 95% CI: 0.1199– [k=0.3163; 95% confidence interval (CI): 0.1199-0.5126, p<0.0001]. 0.5126, p<0.0001). The main disagreement was in relation to 20 anti-TIF-1g positive results detected by Euroline but not by the Diagnostic Classification of Anti-TIF-1g in-house IB, with 16 (80%) of those results for patients in the Positive Patients by Euroline vs other-diseases group. While agreement of k=0.7 between In-House IB Euroline and IP has previously been established for anti-TIF-1g Of the 308 patients included for this part of the study, 127 had UCTDs, 118 had other diseases and 63 had DM. Euroline and in- house IB assay results according to diagnosis are summarized in Table 2. Relationship Between CAM and Anti-TIF-1g Detected by Euroline vs In-House IB Of the 63 patients diagnosed with DM, 16 had CAM, of whom 8 (50%) and 12 (75%) had anti-TIF-1g antibodies detected by Euroline and in-house IB, respectively. There were 4 patients who had anti-TIF-1g antibodies detected by the in-house IB but not by Euroline. Of the 47 patients without CAM, 8 (17%) and 10 TABLE 1 | Anti-TIF-1g antibody results for Euroline and in-house immunoblot. Patients/sample Euroline In-house immunoblot 234/266 Negative Negative 20/266 Positive Negative 5/266 Negative Positive FIGURE 1 | Receiver operating characteristic (ROC) curve. 7/266 Positive Positive Frontiers in Immunology | www.frontiersin.org 3 February 2021 | Volume 11 | Article 625896 Mariscal et al. Myositis Antibody Testing TABLE 3 | Sensitivity, specificity and positive and negative predictive values 66.7%, while NPV ranged from 92% to 95% (6, 7). Both methods (PPV and NPV, respectively) for anti-TIF-1g antibody detection of cancer in compared in this study have slightly lower PPV and NPV than patients with cancer-associated dermatomyositis. those established for anti-TIF-1g antibodies detected by IP. Euroline In-house IB p However, it would have been advisable to establish PPV and NPV in our cohort by IP. Sensitivity (%) 50 75 0.2734 Specificity (%) 83 79 0.7939 Our data indicate that anti-TIF-1g antibodies as detected by PPV (%) 50 54.5 >0.9999 Euroline and the in-house IB have sensibility and specificity NPV (%) 83 90 0.3675 values, PPVs and NPVs similar to those reported in the literature for CAM diagnosis. However, the high number of patients antibody detection in two studies (13, 14), those studies referred positive for anti-TIF-1g antibodies according to Euroline in the to patients with myositis, while in our study the main context of other diseases could mislead the initial diagnosis disagreement reflected other diseases. towards DM and, outside the context of a DM, could lead The usefulness of anti-TIF-1g antibodies as a cancer clinicians to look for cancer. No studies exist to date that biomarker has been conclusively proven only for patients with associate anti-TIF-1g antibody with cancer outside a DM DM. However, immunology and clinical analysis laboratories diagnosis. Therefore, the high proportion of positive results in receive myositis antibody (anti-TIF-1g antibodies included) patients without DM could lead to cancer screening of patients screening requests with no definitive diagnosis of DM. Other for whom it is not known whether they are at greater risk of authors have found unacceptably high false positive rates (FPRs) cancer. A limitation of this study is the relatively small for Euroline-detected myositis-specific and myositis-associated representation of anti-TIF-1g positive results. antibodies in healthy controls (19), so the high FPRs may to While multiple testing increases efficiency, it also has the some extent be inherent to the technique. Unexpectedly high disadvantage of increasing the likelihood of false positive results; percentages of anti-TIF-1g antibodies detected by Euroline in therefore, if anti-TIF-1g antibodies are detected by Euroline, we patients without DM dilutes the usefulness of anti-TIF-1g as a strongly recommend confirming the results using a second myositis-specific antibody (1, 2, 6, 20, 21) and so reinforces the validated method, most especially in the absence of a definitive clinician’s role in correct DM diagnosis. This means that pre-test DM diagnosis. We also recommend, to avoid false-negative probability in any given patient is an important aspect of using results, confirming anti-TIF-1g antibodies using a second Euroline for anti-TIF1g detection. If DM diagnosis was right, validated method if a DM diagnosis is definitive, the Euroline there was a good agreement between Euroline and the in-house result is negative for anti-TIF-1g antibodies and other myositis- IB, agreement was poor in the other scenarios. specific antibodies. The presence of anti-TIF-1g antibodies in DM by IP ranges from 13% to 33% (1, 2, 6, 20). Both Euroline and the in-house IB assays have equivalent sensitivity values that also concur with CONCLUSION previously described anti-TIF-1g antibody detection rates in the DM context. Although differences in results between a commercial anti-TIF1g Of the 16 patients with CAM in our study, four had anti-TIF- assay (Euroline) and an in-house IB previously validated against 1g antibodies detected by the in-house IB but not by Euroline IP are not significant for patients with DM and CAM, agreement (50% of false negative results). A similar false negative rate has in the absence of a definitive DM diagnosis is poor. Our results been reported for Euroline comparisons with IP (12). The non- underpin the importance of a reliable DM diagnosis by the detection of anti-TIF-1g antibodies potentially results in less clinician (pre-test probability) and suggest that the usefulness of rigorous malignancy screening and later detection and treatment. anti-TIF1g antibody detection depends not only on the method Previous studies have reported anti-TIF-1g antibody used but also on an accurate DM diagnosis. frequency by IP of 50%–71.4% for patients with CAM and of 4.5%–14.3% for patients without CAM (2, 6); those rates are quite similar to those obtained by both methods in our study. In diagnosing CAM, for anti-TIF-1g antibodies detected by DATA AVAILABILITY STATEMENT IP, sensitivity and specificity have been reported as 50%–100% The raw data supporting the conclusions of this article will be and 79%–100%, respectively, while pooled estimated sensitivity made available by the authors, without undue reservation. and specificity values in a meta-analysis were 78% and 89%, respectively (7); the sensitivity and specificity values in our study for both methods therefore lie within the ranges defined. Although the in-house IB was more accurate in diagnosing ETHICS STATEMENT CAM than Euroline, both methods were less accurate than IP (the gold standard), as the meta-analysis reported a pooled The studies involving human participants were reviewed and calculated AUC for anti-TIF-1g detected by IP of 0.91 (95% approved by Institut de Recerca de l’Hospital de la Santa Creu I CI: 0.88–0.93) (7). Sant Pau (IIB Sant Pau). Written informed consent for The PPV and NPV for anti-TIF-1g antibodies detected by IP participation was not required for this study in accordance to predict CAM have been established: PPV ranged from 58% to with the national legislation and the institutional requirements. Frontiers in Immunology | www.frontiersin.org 4 February 2021 | Volume 11 | Article 625896 Mariscal et al. Myositis Antibody Testing AUTHOR CONTRIBUTIONS MS, LV-G, LA, and TF-L contributed with anti-TIF-1g antibody detection. AM, CJ, and LM-M contributed to data analysis and AM and MAM established the study design. AS-O’C, IC, MM, AG, manuscript writing. All authors contributed to the article and ET-A, MA-C, and HC contributed with diagnoses of patients. AB, approved the submitted version. 13. Mahler M, Betteridge Z, Bentow C, Richards M, Seaman A, Chinoy H, et al. REFERENCES Comparison of three immunoassays for the detection of myositis specific 1. Targoff IN, Mamyrova G, Trieu EP, Perurena O, Koneru B, O’Hanlon TP, antibodies. Front Immunol (2019) 10:848(APR). doi: 10.3389/ et al. A novel autoantibody to a 155-kd protein is associated with fimmu.2019.00848 dermatomyositis. Arthritis Rheum (2006) 54(11):3682–9. doi: 10.1002/ 14. Cavazzana I, Fredi M, Ceribelli A, Mordenti C, Ferrari F, Carabellese N, et al. art.22164 Testing for myositis specific autoantibodies: Comparison between line blot 2. Kaji K, Fujimoto M, Hasegawa M, Kondo M, Saito Y, Komura K, et al. and immunoprecipitation assays in 57 myositis sera. J Immunol Methods Identification of a novel autoantibody reactive with 155 and 140 kDa nuclear (2016) 433:1–5. doi: 10.1016/j.jim.2016.02.017 proteins in patients with dermatomyositis: An association with malignancy. 15. Fiorentino DF, Gutierrez-Alamillo L, Hines D, Yang Q, Casciola-Rosen L. Rheumatology (2007) 46(1):25–8. doi: 10.1093/rheumatology/kel161 Distinct dermatomyositis populations are detected with different 3. Tiniakou E, Mammen AL. Idiopathic Inflammatory Myopathies and autoantibody assay platforms. Clin Exp Rheumatol (2019) 37(6):1048–51. Malignancy: a Comprehensive Review. Clin Rev Allergy Immunol (2017) 52 16. Hohlfeld R, Engel AG. Polymyositis and Dermatomyositis. Autoimmune Dis (1):20–33. doi: 10.1007/s12016-015-8511-x (2006), 13:453–63. doi: 10.1016/B978-012595961-2/50037-8 4. Hill CL, Zhang Y, Sigurgeirsson B, Pukkala E, Mellemkjaer L, Airio A, et al. 17. Sheagren JN. The New England Journal of Medicine Downloaded from Frequency of specific cancer types in dermatomyositis and polymyositis: A nejm.org at University of Arizona on September 29, 2014. For personal use population-based study. Lancet (2001) 357(9250):96–100. doi: 10.1016/ only. No other uses without permission. From the NEJM Archive. Copyright S0140-6736(00)03540-6 © 2010 Massachusetts Medical Society. All rights reserve. N Engl J Med (1984) 5. Pinal-Fernandez I, Ferrer-Fabregas B, Trallero-Araguas E, Balada E, Martı́nez 310(21):1368–73. doi: 10.1056/NEJM198405243102107 MA, Milisenda JC, et al. Tumour TIF1 mutations and loss of heterozygosity 18. Bottai M, Tjärnlund A, Santoni G, Werth VP, Pilkington C, De Visser M, et al. related to cancer-associated myositis. Rheumatol (United Kingdom) (2018) 57 EULAR/ACR classification criteria for adult and juvenile idiopathic (2):388–96. doi: 10.1093/rheumatology/kex413 inflammatory myopathies and their major subgroups: A methodology 6. Trallero-Araguá s E, Labrador-Horrillo M, Selva-O’Callaghan A, Martı́nez MA, report. RMD Open (2017) 3(2):1–10. doi: 10.1136/rmdopen-2017-000507 Martı́nez-Gó mez X, Palou E, et al. Cancer-associated myositis and anti-p155 19. Bundell C, Rojana-udomsart A, Mastaglia F, Hollingsworth P, McLean-Tooke autoantibody in a series of 85 patients with idiopathic inflammatory myopathy. A. Diagnostic performance of a commercial immunoblot assay for myositis Med (Baltimore) (2010) 89(1):47–52. doi: 10.1097/MD.0b013e3181ca14ff antibody testing. Pathology (2016) 48(4):363–6. doi: 10.1016/j.pathol. 7. Trallero-Araguá s E, Rodrigo-Pendá s JÁ , Selva-O’Callaghan A, Martı́nez- 2016.03.012 Gõmez X, Bosch X, Labrador-Horrillo M, et al. Usefulness of anti-p155 20. Fujikawa K, Kawakami A, Kaji K, Fujimoto M, Kawashiri S, Iwamoto N, et al. autoantibody for diagnosing cancer-associated dermatomyositis: A systematic Association of distinct clinical subsets with myositisspecific autoantibodies review and meta-analysis. Arthritis Rheum (2012) 64(2):523–32. doi: 10.1002/ towards anti155/140kDa polypeptides, anti-140-kDa polypeptides, and art.33379 antiaminoacyl tRNA synthetases in Japanese patients with dermatomyositis: 8. Labrador-Horrillo M, Martı́nez MA, Selva-O’Callaghan A, Trallero-Araguá s A singlecentre, crosssectional stud. Scand J Rheumatol (2009) 38(4):263–7. E, Balada E, Vilardell-Tarré s M, et al. Anti-TIF1ã antibodies (anti-p155) in doi: 10.1080/03009740802687455 adult patients with dermatomyositis: Comparison of different diagnostic 21. Shimizu K, Kobayashi T, Kano M, Hamaguchi Y, Takehara K, Matsushita T. assays. Ann Rheum Dis (2012) 71(6):993–6. doi: 10.1002/art.33379 Anti-transcriptional intermediary factor 1-g antibody as a biomarker in 9. Aggarwal R, Oddis CV, Goudeau D, Fertig N, Metes I, Stephens C, et al. Anti- patients with dermatomyositis. J Dermatol (2020) 47(1):64–8. doi: 10.1111/ transcription intermediary factor 1-gamma autoantibody ELISA development 1346-8138.15128 and validation. Rheumatol (United Kingdom) (2014) 53(3):433–7. doi: 10.1093/rheumatology/ket383 Conflict of Interest: The authors declare that the research was conducted in the 10. Fujimoto M, Murakami A, Kurei S, Okiyama N, Kawakami A, Mishima M, absence of any commercial or financial relationships that could be construed as a et al. Enzyme-linked immunosorbent assays for detection of anti- potential conflict of interest. transcriptional intermediary factor-1 gamma and anti-Mi-2 autoantibodies in dermatomyositis. J Dermatol Sci [Internet] (2016) 84(3):272–81. Copyright © 2021 Mariscal, Milán, Baucells, Martı ́nez, Guillen, Trallero-Araguás, doi: 10.1016/j.jdermsci.2016.09.013 Alvarado-Cardenas, Martı ́nez-Martı ́nez, Alserawan, Franco-Leyva, Sanz-Martı ́nez, 11. Vulsteke JB, De Langhe E, Claeys KG, Dillaerts D, Poesen K, Lenaerts J, et al. Viñas-Giménez, Corominas, Juárez, Castellvı ́ and Selva-O’Callaghan. This is an Detection of myositis-specific antibodies. Ann Rheum Dis (2019) 78(1):e7. open-access article distributed under the terms of the Creative Commons Attribution doi: 10.1136/annrheumdis-2017-212915 License (CC BY). The use, distribution or reproduction in other forums is permitted, 12. Tansley SL, Li D, Betteridge ZE, McHugh NJ. The reliability of immunoassays provided the original author(s) and the copyright owner(s) are credited and that the to detect autoantibodies in patients with myositis is dependent on original publication in this journal is cited, in accordance with accepted academic autoantibody specificity. Rheumatology (2020) 1:59(8):1–6. doi: 10.1093/ practice. No use, distribution or reproduction is permitted which does not comply with rheumatology/keaa021 these terms. Frontiers in Immunology | www.frontiersin.org 5 February 2021 | Volume 11 | Article 625896