Clinicopathological and serological spectrum of idiopathic inflammatory myopathies: Experience from a neuromuscular referral center in Cyprus

Abstract

Background

The classification of idiopathic inflammatory myopathies (IIMs) has evolved substantially with the identification of myositis-specific autoantibodies (MSAs). However, how contemporary classification frameworks are applied in routine clinical practice remains variable. This study provides a clinicopathological and serological characterization of IIMs from a national neuromuscular referral center in Cyprus.

Methods

We conducted a retrospective, single-center cohort study of adult patients diagnosed with IIM between 2007 and 2025. Patients were classified using a clinicopathological and serological approach into dermatomyositis (DM), immune-mediated necrotizing myopathy (IMNM), antisynthetase syndrome (ASyS), overlap myositis (OM), inclusion body myositis (IBM), and polymyositis (PM), with the 2017 EULAR/ACR criteria used as a reference framework and historical diagnoses re-evaluated.

Results

36 patients were included. Final subtype distribution: IMNM in 14 patients (39%), DM in 7 (19%), IBM in 4 (11%), OM in 5 (14%), ASyS in 3 (8%), PM in 2 (6%). Only one patient remained non-classifiable. Autoantibody testing was performed in 26/36 patients; MSAs were detected in 19/26 tested (73%; 53% of the total cohort), most commonly anti-HMGCR, while anti-NT5C1A antibodies were detected in three additional patients. Muscle biopsy was performed in the majority of patients, reflecting historical diagnostic practices.

Conclusion

This study demonstrates real-world application of contemporary IIM classification, enabling accurate disease subtyping with direct implications for malignancy surveillance and interstitial lung disease monitoring, as well as reclassification of cases, most notably those previously labelled as PM. The findings further underscore the role of muscle biopsy, particularly when serology is negative, discordant, or unavailable.

Introduction

Idiopathic inflammatory myopathies (IIMs) are a heterogeneous group of autoimmune muscle diseases that include dermatomyositis (DM), polymyositis (PM), immune-mediated necrotizing myopathy (IMNM), inclusion body myositis (IBM), anti-synthetase syndrome (ASyS), and overlap myositis (OM). Each subtype is defined by a characteristic combination of clinical manifestations, autoantibody profiles, and pathological findings.^1–6

The discovery of myositis-specific and myositis-associated autoantibodies (MSA/MAA) has substantially refined disease classification,^7–10 while muscle biopsy remains highly important, particularly in seronegative or atypical presentations.^3,6,11–13 International studies have highlighted variations in the distribution of IIM subtypes between regions and clinical settings, reflecting both epidemiological and referral differences. ¹⁴

Despite substantial advances in the serological and pathological classification of IIMs, the translation of contemporary classification frameworks into routine clinical practice remains variably documented across different healthcare settings. We therefore conducted a retrospective study at a national neuromuscular referral center in Cyprus, where IIMs have not previously been systematically studied, to provide an integrated characterization of the clinical, serological, and pathological spectrum of IIMs as encountered in routine practice. This approach incorporated both initially well-classified cases and systematic re-evaluation of historical diagnoses, including cases previously classified as PM. We further aimed to describe subtype distribution and the complementary roles of MSA testing and muscle biopsy, with implications for malignancy surveillance and interstitial lung disease (ILD) monitoring.

Methods

Study design and setting

We conducted a retrospective cohort study at the Cyprus Institute of Neurology and Genetics (CING), a national referral center for neuromuscular disorders.

Inclusion criteria

All adult patients (≥18 years) diagnosed with IIM between January 2007 and September 2025 were eligible. Inclusion required: (i) a clinical diagnosis of IIM established by a neuromuscular specialist, and (ii) availability of sufficient clinical, serological, and/or pathological data to allow confident classification.^1,6,15 Patients with insufficient documentation were excluded, as well as those with inherited, metabolic, drug-induced, or other non-inflammatory myopathies.

Data collection

Demographic, clinical and laboratory data were extracted from medical records. Variables included age at onset, sex, presenting symptoms, creatine kinase (CK) levels and antibody profiles. Muscle biopsy reports were reviewed in all available cases, and histological slides were re-examined when necessary.

Diagnostic classification

Diagnostic classification was established by integrating clinical, serological, and pathological data.^2,6,16 All cases were systematically re-reviewed, independent of the original diagnostic label, and classified according to contemporary clinicopathological and serological standards. Patients were classified using the 2017 EULAR/ACR classification criteria as a reference framework. Although these criteria include anti-Jo-1 status, they do not incorporate the broader spectrum of myositis-specific and myositis-associated autoantibodies (MSA/MAAs) now routinely used in clinical practice, nor do they recognize IMNM or ASyS as distinct entities. Therefore, subclassification was refined by integrating comprehensive autoantibody profiles, clinical features, and muscle biopsy findings, in line with contemporary consensus definitions and subsequent literature.^6,17,18

DM was defined by characteristic cutaneous manifestations and/or DM-associated autoantibodies with supportive clinical, serological, or pathological findings.^1,2,12,18 IMNM was defined by a compatible clinical phenotype together with a necrotizing muscle biopsy pattern and/or anti-HMGCR or anti-SRP positivity.^11,19 IBM was diagnosed according to the revised 2024 ENMC framework, integrating clinical and pathological findings. ³ ASyS was defined by the presence of anti-synthetase antibodies together with compatible clinical manifestations, including inflammatory myopathy, interstitial lung disease, arthritis, or related extramuscular features, in accordance with contemporary consensus recommendations. ¹³ OM was defined according to the operational definition used in the EuroMyositis registry, namely inflammatory myopathy occurring in the presence of a well-characterized connective tissue disease (systemic sclerosis, SLE, Sjögren’s syndrome, mixed CTD, or RA), provided that patients did not fulfil criteria for DM, IMNM, ASyS, or IBM.^2,20,21 PM was recognized as a diagnosis of exclusion and applied only when patients had no coexisting connective tissue disease and no clinical, serological, or pathological features consistent with DM, IMNM, IBM, ASyS, or OM. ¹⁷ Patients not fulfilling criteria for any defined subtype were categorized as non-classifiable IIM.

Serological testing

Testing for myositis-specific and myositis-associated autoantibodies was performed using standardized commercial immunoblot and ELISA assays in accredited reference laboratories. The myositis-specific autoantibody (MSA) panel included antibodies against Mi-2α, Mi-2β, transcription intermediary factor 1 gamma (TIF1γ), melanoma differentiation-associated protein 5 (MDA5), nuclear matrix protein 2 (NXP2), small ubiquitin-like modifier activating enzyme 1 (SAE1), signal recognition particle (SRP), histidyl-tRNA synthetase (Jo-1), threonyl-tRNA synthetase (PL-7), alanyl-tRNA synthetase (PL-12), isoleucyl-tRNA synthetase (OJ), and glycyl-tRNA synthetase (EJ). The myositis-associated autoantibody (MAA) panel included antibodies against Ro52, PM-Scl75, PM-Scl100, and Ku. Antibodies against 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGCR) and cytosolic 5′-nucleotidase 1A (cN1A/NT5C1A; Mup44) were assessed separately using ELISA-based assays. For the purposes of this retrospective study, only moderate (++) and strong (+++) autoantibody reactivity was recorded as positive, whereas weak positive (+) results were not considered positive. In cases with more than one positive autoantibody, final subtype assignment was based on integrated clinical, serological, and pathological assessment rather than antibody status alone.^{8,11–13,18,22}

Muscle biopsy

All muscle biopsies included in this study were obtained by open biopsy. Biopsy specimens were processed in our Neuropathology Laboratory using standard histological, histochemical, immunohistochemical, and immunofluorescence protocols. Frozen sections were stained with H&E, modified Gömöri trichrome, ATPases (pH 4.2, 4.6 and 9.4), acid phosphatase, α-glycerophosphate dehydrogenase, oxidative enzyme histochemistry (NADH-TR, SDH, and COX), PAS, and Oil Red O. Immunostaining included MHC class I and complement (C5b-9). Paraffin-embedded sections were stained with H&E, CD3, CD4, CD8, CD68, and CD20. The complete staining panel was performed in all biopsies included in the study. All samples were reviewed by neuropathologists with expertise in neuromuscular disease. ²³

Statistical analysis

Data were summarized using descriptive statistics. Continuous variables are presented as mean (range), and categorical variables as number (percentage).

Results

Overall subtype distribution

Thirty-six (36) patients were included: IMNM 14 (39%), DM 7 (19%), IBM 4 (11%), OM 5 (14%), ASyS 3 (8%), PM 2 (6%), and non-classifiable idiopathic inflammatory myopathy 1 (3%). The clinical and serological characteristics of the cohort stratified by disease subtype are summarized in Table 1.

Table 1.

Demographic and serological characteristics of patients with idiopathic inflammatory myopathies stratified by disease subtype.

Subtype	Female:Male	Age at onset (years)	MSA/anti-NT5C1A	MAA
IMNM (n=14, 39%)	10:4	58 (21–83)	anti-HMGCR (n=9); anti-SRP (n=2); anti-NT5C1A (n=1); negative (n=2); not tested (n=1)	—
DM (n=7, 19%)	4:3	54 (27–71)	anti-TIF1γ (n=4); anti-MDA5 (n=1); not tested (n=2)	anti-Ro52 (n=2); anti-PM-Scl (n=1); anti-Ku (n=1)
IBM (n=4, 11%)	2:2	64 (55–74)	anti-NT5C1A (n=1); not tested (n=3)	—
OM (n=5, 14%)	5:0	62 (39–69)	negative (n=2); not tested (n=3)	anti-PM-Scl (n=3); anti-Ro52 (n=1)
ASyS (n=3, 8%)	2:1	52 (37–65)	anti-Jo-1 (n=2); anti-PL-7 (n=1)	—
PM (n=2, 6%)	1:1	53 (42–63)	anti-NT5C1A (n=1); not tested (n=1)	—
Non-classifiable IIM (n=1, 3%)	1:0	68	negative (n=1)	anti-Ro52 (n=1)

Age at onset is presented as mean (range). MSA, myositis-specific autoantibody; MAA, myositis-associated autoantibody; IMNM, immune-mediated necrotizing myopathy; DM, dermatomyositis; IBM, inclusion body myositis; OM, overlap myositis; ASyS, antisynthetase syndrome; PM, polymyositis. Negative indicates absence of an autoantibody; not tested indicates that MSA testing was not performed. An em dash (—) indicates that MAA testing was unavailable or not systematically documented. †One patient with IMNM was positive for both anti-HMGCR and anti-NT5C1A.

MSA testing was performed in 26/36 patients (72%). MSAs were detected in 19 patients, corresponding to 19/36 of the total cohort (53%) and 19/26 of those tested (73%). Anti-NT5C1A antibodies were detected in three additional patients, including one who was also positive for anti-HMGCR. Five patients were negative for MSAs (5/36, 14%; 5/26 tested, 19%), while 10/36 patients (28%) were not tested. Among patients positive for MSAs (n=19), anti-HMGCR was the most frequent antibody (n=9), followed by anti-TIF1γ (n=4). Anti-SRP and anti-Jo-1 were each detected in two patients, while single cases were positive for anti-MDA5 and anti-PL7. MAAs were detected in a subset of patients, most frequently anti-Ro52 and anti-PM-Scl, with anti-Ku identified in one case. Muscle biopsy was performed in 30/36 patients (83%).

Immune-mediated necrotizing myopathy (IMNM)

Fourteen patients (39%) were classified as IMNM, with a female-to-male ratio of 10:4 and a mean age at onset of 58 years (range 21–83 years). Anti-HMGCR antibodies were detected in nine patients (including one patient who was also positive for anti-NT5C1A), while anti-SRP antibodies were identified in two. Two patients were seronegative, and antibody testing was not performed in one case. The patient with dual anti-HMGCR and anti-NT5C1A positivity demonstrated a necrotizing myopathy with minimal inflammatory changes and no rimmed vacuoles on muscle biopsy, and the clinical phenotype lacked features characteristic of IBM. The case was therefore classified as IMNM.

Serum CK levels were markedly elevated in IMNM patients, ranging from 1,100 to 15,000 IU/L, with a mean of 6,800 IU/L. Muscle biopsies in most patients revealed numerous necrotic and regenerating fibers with minimal inflammatory cell infiltrates (Figure 1(m)–(p)), although more notable endomysial inflammation was present in three cases (two anti-HMGCR–positive and one anti-SRP–positive). Clinically, proximal weakness was the predominant feature, often of subacute onset. Some patients also developed truncal muscle weakness, while none exhibited extramuscular manifestations.

Figure 1.

(a–d): Images from quadriceps biopsy in a patient with IBM. (a): H&E stain (20x magnification) shows variation in fiber size, mostly angulated, atrophic, many degenerating and some necrotic fibers. Some fibers show rimmed vacuoles (black arrows). Diffuse mononuclear inflammatory cell infiltrates in the endomysium. (b): ATPase 4.2 (10x) shows variation in fiber size, some vacuolated fibers. (c): SDH/COX double staining (10x) shows several COX-negative fibers (asterisks). (d): CD8 staining (20x) shows T cell infiltrates in the endomysium (red arrows). (e–h): Images from quadriceps biopsy in a patient with PM. (e): H&E staining (20x) shows marked fiber size variation, atrophic and necrotic fibers, no rimmed vacuoles. Extensive mononuclear cell infiltrates invading non-necrotic fibers (white arrows). (f): ATPase 4.2 staining (10x) shows fiber size variation with many atrophic and necrotic fibers. (g): Gömöri Trichrome staining (20x) also shows severe inflammatory infiltrates with non-necrotic fiber invasion (white arrows). (h): CD3 staining (20x) shows prominent T cell infiltrates with fiber invasion. (i–l): Images from quadriceps biopsy in a patient with DM. (i): H&E staining (5x) shows severe inflammatory infiltrates that are most prominent in a perifascicular distribution, invading fibers of the 2-3 first layers at the margin of the muscle fascicle. (j): ATPase 4.2 staining (5x) shows prominent perifascicular atrophy without type-specific atrophy or type grouping. (k): CD3 staining (10x) shows intense T cell inflammatory infiltrates mostly localized in perifascicular and perivascular areas of connective tissue. (l): C5b-9 immunofluorescence staining (20x, green) shows strong upregulation in the cytoplasm of some fibers mostly in perifascicular localization, as well as deposits in capillaries (open arrowheads). (m–p): Images from biceps brachii biopsy in a patient with HMGCR-Ab+ IMNM. (m): H&E staining shows fiber size variation, frequently atrophic fibers and several necrotic fibers. No prominent inflammatory infiltrates are present within the endomysium. (n): Kpi/CD68 staining (20x) shows the presence of scattered macrophages throughout the endomysium, that are prominently surrounding and clearing necrotic fibers. (o): MHC class I immunofluorescence staining (green, 10x) shows upregulation on muscle fiber membranes beyond capillaries and within connective tissue, (p): C5b-9 terminal complement immunofluorescence staining (green, 10x) shows characteristic upregulation on the sarcolemma of non-necrotic fibers in a fine punctuated pattern (open arrowheads).

Three patients had a history of malignancy (hematological, gastrointestinal, and urological cancers), two of whom were anti-HMGCR positive and one not tested. All were in remission at IIM onset.

Dermatomyositis (DM)

Seven patients (19%) were classified as DM, with a female-to-male ratio of 4:3 and a mean age at onset of 54 years (range 27–71 years). Serum CK levels were normal or only mildly elevated in most cases. MSAs were detected in five patients: four were anti-TIF1γ positive and one anti-MDA5 positive. MAAs included anti-Ro52, anti-PM/Scl, and anti-Ku. Two patients were not tested for MSAs.

Cutaneous manifestations across the DM cohort predominantly included heliotrope rash and/or Gottron’s papules, with additional truncal erythematous involvement observed in some cases. ILD was documented in one patient, who was anti-MDA5 positive and developed a severe pulmonary phenotype.

Muscle biopsy was performed in four patients and demonstrated similar pathological findings, with features characteristic of DM, including perivascular and perimysial inflammatory infiltrates with atrophy and complement deposition (Figure 1(i)–(l)), consistent with hallmark DM pathology. ² In the remaining patients, the diagnosis of DM was established on clinical and serological grounds, supported by characteristic cutaneous manifestations and DM-specific autoantibodies. None of the DM patients, including those positive for anti-TIF1γ, developed malignancy during follow-up.

Inclusion body myositis (IBM)

Four patients (11%) were classified as IBM, with a female-to-male ratio of 2:2 and a mean age at onset of 64 years (range 55–74 years). Serum CK levels were mildly to moderately elevated (530–2,000 IU/L). Three patients exhibited characteristic finger flexor weakness, while two had quadriceps weakness with associated atrophy. All patients had proximal muscle weakness, and one also developed truncal weakness. Anti-NT5C1A antibodies were detected in one patient, while the remaining three had not undergone testing.

Muscle biopsies in all cases showed endomysial inflammatory infiltrates with lymphocyte invasion of non-necrotic fibers and diffuse MHC class I upregulation. Among the three patients without antibody testing, numerous COX-negative fibers were observed; one biopsy revealed rimmed vacuoles and another showed ragged-red fibers (Figure 1(a)–(d)).^3,24–26 One patient had initially been diagnosed with PM based on biopsy findings but was subsequently reclassified as IBM because of the characteristic pattern of muscle weakness, including finger flexor involvement.

One patient (onset at 74 years) presented with slowly progressive, predominantly proximal muscle weakness. Muscle biopsy demonstrated inflammatory changes without rimmed vacuoles but with ragged-red fibers, several COX-negative fibers, and anti-NT5C1A antibody positivity, thereby fulfilling the revised 2024 ENMC diagnostic criteria for IBM in the setting of an uncommon clinical presentation.^3,26

Anti-synthetase syndrome (ASyS)

Three patients (8%) were classified as ASyS, comprising two females and one male, with a mean age at onset of 52 years (range 37–65). Two patients were positive for anti-Jo1 antibodies, and one for anti-PL7. Serum CK levels were moderately elevated, with a mean of 1,850 U/L (range 1,800–1,900 U/L), though baseline values were unavailable in one patient.

Muscle biopsies, available in all three, showed inflammatory pathology with endomysial or perimysial infiltrates, necrotic fibers, and diffuse MHC class I upregulation. Clinically, all patients presented with proximal muscle weakness. Extramuscular features included arthritis and erythema in one case and ILD in the patient with anti-PL7 positivity.

Notably, this PL7-positive patient had initially been labelled as PM. Re-evaluation, integrating the antibody profile together with the presence of ILD, established the diagnosis of ASyS.^13,27,28

Overlap myositis (OM)

Five patients (14%) were classified as OM, all female, with a mean age at onset of 62 years (range 39–69). Each patient had a well-defined connective tissue disease (CTD) coexisting with inflammatory myopathy: two with systemic lupus erythematosus (SLE), two with rheumatoid arthritis (RA), and one with systemic sclerosis.

Serological findings supported the overlap nature of the disease. One patient with RA was anti-PM-Scl positive, the patient with systemic sclerosis was positive for both anti-PM-Scl and anti-Ro52, and one SLE patient was also anti-PM-Scl positive. Antibody testing was not available in the remaining two cases. Serum CK levels were mildly to moderately elevated (range 600–1,500 IU/L); baseline values were unavailable in two patients.

Muscle biopsies were performed in four patients and demonstrated endomysial and/or perimysial inflammatory infiltrates with variable degrees of myofiber necrosis. The remaining patient did not undergo biopsy.

Clinically, all patients presented with proximal muscle weakness. Extramuscular manifestations were prominent and reflected the underlying CTDs, including arthritis, ILD, and scleroderma-related features. Notably, the two patients with RA had initially been diagnosed as PM based on muscle involvement alone. Subsequent identification of rheumatologic features and CTD-associated autoantibodies prompted reclassification as OM. This approach is consistent with published operational definitions.^2,20,21

Polymyositis (PM)

At presentation, six patients were initially diagnosed with PM. After detailed re-evaluation, four were reclassified (two as OM, one as ASyS, and one as IBM), leaving two patients (6%) with a final PM label. One was female and one male, with a mean age at onset of 53 years (range 42–63).

Serum CK levels were moderately elevated, ranging from 750 to 1,310 IU/L (mean 1,030 IU/L). Muscle biopsies in both patients showed endomysial inflammatory infiltrates with invasion of non-necrotic fibers, consistent with a PM pattern (Figure 1(e)–(h)); MHC class I was upregulated in both, and one exhibited COX-negative fibers, with additional C5b-9 deposition in one. Clinically, both presented with symmetric proximal weakness without skin involvement or extramuscular features. One patient had not undergone autoantibody testing, while the other was anti-NT5C1A–positive.

The anti-NT5C1A–positive patient had disease onset at 42 years of age and presented with predominantly proximal muscle weakness, without selective finger flexor or quadriceps weakness. Muscle biopsy did not demonstrate rimmed vacuoles or other characteristic degenerative features, although COX-negative fibers were present. Following treatment with intravenous immunoglobulins (IVIG) and mycophenolate mofetil, the patient experienced mild improvement in proximal muscle strength and has remained clinically stable over approximately four years from disease onset while continuing both therapies.

The other patient had a history of gastrointestinal cancer several years prior to IIM onset, with no evidence of active or recurrent disease during follow-up.

Non-classifiable idiopathic inflammatory myopathy (IIM)

One patient (3%) was classified as non-classifiable IIM. The patient was female, with disease onset at 68 years, and presented with predominantly proximal muscle weakness accompanied by truncal involvement. MSAs were negative, while anti-Ro52 antibodies were present. The patient had associated ILD and did not have a defined CTD.

Muscle biopsy demonstrated nonspecific inflammatory changes, including fiber size variation and mild endomysial inflammatory infiltrates, with increased MHC class I expression and focal complement (C5b-9) deposition. However, the biopsy lacked defining inflammatory, necrotizing, perifascicular, or degenerative patterns, and in the absence of MSAs, the overall clinicopathological profile did not permit confident classification as DM, IMNM, IBM, ASyS, OM, or PM; therefore, the case remained non-classifiable IIM.

Discussion

Complementary roles of clinical presentation, antibodies and muscle biopsy in diagnosis

MSAs played a decisive role in refining diagnosis in our cohort. This is in line with international experience showing that MSAs act as powerful diagnostic anchors for defining myositis subtypes.^2,5–7,9 Beyond subtype classification, selected MSAs also convey important prognostic information, particularly with respect to extramuscular involvement.

Anti-MDA5 and anti-PL7 antibodies are strongly associated with ILD, with MDA5 in particular often predicting a rapidly progressive course.^13,18,28,29 In our cohort, both antibody-positive patients developed ILD, severe in the MDA5 case, highlighting the prognostic relevance of these markers. Another patient with milder ILD lacked MSAs but carried anti-Ro52, consistent with reports associating anti-Ro52 with pulmonary involvement. ³⁰

Anti-TIF1γ antibodies are among the strongest serological markers of cancer-associated DM. ³¹ None of the four anti-TIF1γ–positive patients developed malignancy during follow-up. This absence of cancer likely reflects the small sample size rather than a true lack of association, and underscores the importance of continued individualized cancer surveillance despite negative initial screening. Recent studies suggest that additional autoantibodies, including anti-Sp4 and anti-CCAR1, may further refine malignancy risk stratification among anti-TIF1γ-positive patients. ³²

One patient demonstrated both anti-HMGCR and anti-NT5C1A reactivity, yet the biopsy and clinical presentation were typical of necrotizing myopathy, and the case was therefore classified as IMNM rather than IBM. Another patient classified as PM was anti-NT5C1A positive and demonstrated some pathological features that may be seen in IBM, including COX-negative fibers. Although the revised ENMC criteria recognize that IBM may present with uncommon phenotypes, including younger age at onset and predominantly proximal weakness, ³ the patient’s longitudinal clinical course has remained atypical, with sustained clinical stability over approximately four years while receiving therapy. Accordingly, an early or atypical IBM phenotype cannot be completely excluded, and continued clinical follow-up will be important to determine whether more characteristic features of IBM emerge over time. Together, these cases highlight the importance of interpreting anti-NT5C1A antibody positivity within the broader clinical, serological, and histopathological context rather than in isolation.^2,3,6

In addition, although IMNM is most commonly associated with anti-HMGCR or anti-SRP antibodies, seronegative IMNM has been described, and diagnosis in such cases relies on a characteristic clinicopathological phenotype rather than serology alone. ¹¹ In our cohort, two seronegative patients fulfilled diagnostic criteria for IMNM based on markedly elevated CK levels and a necrotizing muscle biopsy pattern.

One patient remained non-classifiable, reflecting the limitations of current criteria when serological and pathological findings are inconclusive. Such cases highlight the need for continued refinement of classification systems and underscore the value of long-term follow-up and re-evaluation.^5,33

The relatively high rate (83%) of muscle biopsy in our cohort primarily reflects historical diagnostic practice during earlier phases of the study period, when access to comprehensive MSA testing was limited and biopsy constituted a central component of diagnostic evaluation. In addition, the availability of specialized neuromuscular pathology expertise within our institution may have contributed to a lower threshold for biopsy in routine practice.

Diagnostic reclassification in the antibody era and polymyositis as a diagnosis of exclusion

Of six patients initially diagnosed with PM in our cohort, only two (6%) retained this classification after comprehensive re-evaluation, while the remaining cases were reclassified as ASyS, IBM, or OM.

These observations align with several large series indicating that a substantial proportion of patients historically labelled as PM may be reclassified within other IIM subgroups when contemporary serological and pathological criteria are applied.^15,17 In this context, PM is increasingly viewed as a diagnosis of exclusion, typically considered in patients who lack defining serological markers of other subtypes and whose muscle biopsy shows a classical PM pattern without features characteristic of DM, IMNM, or IBM. ¹⁷

Accordingly, reclassification to OM is appropriate when inflammatory myopathy occurs in the context of a well-defined systemic autoimmune disease, provided that patients do not fulfil criteria for DM, IMNM, ASyS, or IBM. In such cases, muscle biopsy serves to confirm inflammatory myopathy and exclude alternative IIM subtypes.^2,20,21

Comparison with international cohorts

Our cohort was dominated by IMNM (39%), followed by DM (19%), IBM (11%), OM (14%), ASyS (8%), and PM (6%). This subtype distribution is consistent with that reported in contemporary international cohorts, in which IMNM has emerged as one of the most frequent subtypes and PM is uncommon.

Similar patterns have been reported internationally. In a French multicenter cohort (n=260), data-driven clustering identified IMNM (35%), IBM (30%), DM (20%), and ASyS (15%), with no distinct PM cluster. ⁴ In South Korea, reclassification of 122 patients using comprehensive serology showed IMNM (44%) and DM (39%) as the dominant subtypes, with only 6% classified as PM. ³⁴ Studies from Australia and the UK have likewise highlighted anti-HMGCR–positive IMNM as a frequent diagnosis in contemporary cohorts. ³⁵ A UK-based study from Salford demonstrated the impact of classification strategy, whereby reliance on the 2017 EULAR/ACR criteria alone resulted in a high PM proportion, while expert clinicopathological reassessment substantially reduced PM and increased recognition of ASyS, OM, and IMNM. ³⁶

By contrast, older European registries reported markedly different subtype distributions. The German National Database (1997–2017) listed PM as the most common diagnosis (44%), with IMNM rarely recognized (0.5%), ³⁷ while the EuroMyositis registry reported PM in 27% of cases and IMNM in only 5–8%. ²⁰ These discrepancies largely reflect pre-MSA-era diagnostic practices, which tended to overestimate PM and underrecognize IMNM.

Taken together, the overrepresentation of IMNM in our cohort likely reflects referral patterns within a neurology-led center evaluating predominantly muscle-dominant disease, while the strong concordance between serological and pathological findings within IMNM supports the robustness and reliability of this classification.^11,19 In parallel, the low retention of PM after re-evaluation mirrors trends observed in contemporary cohorts and contrasts with historical registries.

Malignancy in our cohort

In our cohort, four patients had a history of cancer: three within the IMNM subgroup, and one with PM. None of these malignancies were temporally related to IIM onset or appeared causally linked. Nevertheless, awareness of potential cancer associations in IIM remains crucial, especially in subgroups with established risk. ³¹ Furthermore, international guidelines have recently proposed risk-adapted cancer screening strategies in patients with IIM based on autoantibody status and clinical phenotype.^38,39

Integration of serology and pathology

A pragmatic diagnostic strategy for IIMs emerges: When MSAs are positive and align with phenotype, they provide sufficient evidence to guide classification and subtype-specific management.^2,6–8,40 However, autoantibody results should always be interpreted in the context of the clinical phenotype, as both false-positive and false-negative results may occur and assay performance varies according to antibody specificity. ⁴¹ When serology is negative, inconclusive, discordant, or not fully compatible with the clinical phenotype, muscle biopsy remains critical to ensure diagnostic accuracy and frequently provides decisive information for subtype assignment and exclusion of important mimics.^2,3,11–13 This integrated approach not only improves diagnostic precision but also has immediate clinical implications for prognosis, cancer screening, and ILD surveillance.^29,31,38 In light of these findings, we strongly support the routine adoption of MSA testing in patients with suspected IIM, ensuring early and accurate classification across subtypes.^6,7,42

In settings where comprehensive MSA panels are not readily available, targeted antibody testing may still be informative in selected clinical scenarios. For example, anti-HMGCR testing should be considered in patients without prominent extramuscular manifestations who present with markedly elevated CK levels, particularly in the setting of statin exposure.^19,43

In IBM, however, muscle biopsy retains a central role within the current 2024 ENMC diagnostic framework, particularly in atypical presentations. ³ Although anti-cN1A antibodies may provide supportive diagnostic evidence, their limited sensitivity and imperfect specificity preclude their use as a standalone diagnostic marker.^44–46

Strengths and limitations

Strengths of this study include its status as the first systematic characterization of IIM in Cyprus, the systematic review of all available muscle biopsies, systematic autoantibody testing, and the explicit documentation of diagnostic reclassification according to contemporary clinicopathological and serological criteria.

Several limitations should be acknowledged. The retrospective design and single-center nature of the study limit formal epidemiological inferences. As this was a retrospective, single-center, referral-based study, the cohort does not represent an epidemiological estimate of IIM in Cyprus. Rather, the included patients reflect those referred to our tertiary neuromuscular center for evaluation and fulfilling the study inclusion criteria. Antibody testing was incomplete in a subset of patients. In addition, in routine clinical practice, access to comprehensive autoantibody testing may be delayed by administrative approval processes related to test reimbursement, which can influence the sequencing of diagnostic investigations. Furthermore, muscle imaging modalities, such as magnetic resonance imaging or ultrasound, were not routinely incorporated into the diagnostic work-up, which may have limited assessment of disease distribution, activity, and chronicity in some cases. Certain histochemical and immunohistochemical markers, including alkaline phosphatase, MHC class II, MxA, and p62/LC3, were also not part of the standardized muscle biopsy protocol used during the study period. The neurology-based referral setting likely contributed to an overrepresentation of IMNM relative to population prevalence. Finally, the exclusion of patients with incomplete records may have introduced bias in subtype distribution, underscoring the need for standardized, prospective data collection in future studies.

Conclusion

This study provides the first systematic, antibody- and biopsy-integrated characterization of idiopathic inflammatory myopathies in Cyprus, offering real-world insight into contemporary diagnostic practice within a national neuromuscular referral center.

Beyond describing subtype distribution, our findings demonstrate how integration of autoantibodies with clinicopathological assessment reshapes diagnostic labels in routine practice and highlights the complementary—and rarely discordant—roles of serology and muscle biopsy, with biopsy remaining essential when serological findings are delayed, inconclusive, or discordant with the clinical phenotype.

Importantly, this diagnostic approach has direct implications for patient management, including malignancy surveillance and ILD monitoring.

Future prospective, multicenter collaborations across neurology and rheumatology services are warranted to further refine classification strategies and to better characterize disease patterns and outcomes in IIM in Cyprus and worldwide.

ORCID iDs

Tasos Tsokkos https://orcid.org/0009-0007-9961-4174

Christodoulos Taliadoros https://orcid.org/0009-0008-4267-4843

Ethical considerations

The study was approved by the Cyprus National Bioethics Committee (Approval No. ΕΕΒΚ ΕΠ 2025.01.300).

Consent to participate

The requirement for individual informed consent was waived due to the retrospective and anonymized study design.

Consent for publication

Not applicable. No identifiable patient data are included in this study.

Data Availability Statement

Anonymized data supporting the findings of this study are available from the corresponding author upon reasonable request.*