Anti-SAE autoantibody in dermatomyositis: original comparative study and review of the literature

Abstract

Introduction

DM is a rare idiopathic inflammatory disorder characterized by the association of typical skin and skeletal muscle features [1]. This systemic disease may be associated with interstitial lung disease (ILD) and/or malignancy, which are the leading causes of death in myositis patients [2].

Five mutually exclusive DM-specific autoantibodies [anti-Mi2, anti-melanoma differentiation-associated protein 5 (MDA5), anti-nuclear matrix protein-2 (NXP2), anti-transcriptional intermediary factor-1 (TIF1γ), anti-small ubiquitin like modifier activating enzyme (SAE)] have been described thus far. In addition to their importance for diagnosis [3], DM-specific autoantibodies help delineate more homogeneous subgroups of patients associated with distinct clinical phenotypes and prognosis (e.g. regarding presence of ILD or cancer) [4].

The anti-SAE antibody was first described in 2007 by Betteridge et al. [5]. Its prevalence is low, ranging from <1% to 8% of DM patients [5, 6]. Only a few case reports and case series have been published on it, reporting association with late-onset myositis with mild severity, but with a high frequency of dysphagia [5, 7–10], sometimes in association with ILD [5, 6, 8, 11–16]. These studies have raised awareness concerning a diffuse skin rash [8, 10, 16, 17] and potential ethnic disparities [7, 8, 10], and no reliable phenotypic description can yet be made [18]. Findings regarding the prevalence of cancer vary between studies [6, 7, 12, 16, 19], and the prognosis of DM associated with the anti-SAE antibody remains uncertain. Finally, its myopathological features have yet to be described. This study aimed to characterize the clinical and pathological features of anti-SAE antibody–positive DM (anti-SAE + DM) patients, clarify its association with cancer, and evaluate its outcome.

Patients and methods

We identified anti-SAE + DM patients through the French networks in systemic diseases and dermatology, and data collection occurred between January 2018 and April 2021, with clinical follow-up until May 2021.

Patients were enrolled if (i) they fulfilled the 239th European NeuroMuscular Centre (ENMC) DM classification criteria [18] and (ii) had positive anti-SAE autoantibody test results. Patients with anti-SAE–negative DM (Anti-SAE – DM) were used as controls.

Testing for anti-SAE antibodies and other DM-specific autoantibodies (anti-NXP-2, anti-Mi2, anti-TIF1-γ and anti-MDA5) was undertaken using commercial dot immunoassays (D-Tek ^®, Belgium, or Euroline ^®, Germany).

Data collection

The following data were collected for each patient using a standardized form: demographic data, clinical presentation data, laboratory test results and paraclinical evaluation data [chest CT, pulmonary function tests (PFTs), MRI, and skin and muscle biopsies) at diagnosis, longitudinal clinical evolution, and lines of treatment.

Cutaneous manifestations were classified according to the ENMC criteria [20]. Typical DM lesions were defined as a heliotrope (red to violaceous) rash and/or eyelid oedema, Gottron’s sign or papules at bony prominences, an erythema of the chest and neck (V-sign), or an erythema of the upper back (shawl sign). Classical DM lesions included a photosensitive erythema [face and/or trunk and/or extension side of the arms and/or lateral hips (holster sign)], a periungual telangiectatic erythema or dystrophic cuticles; and secondary DM lesions were panniculitis, alopecia, poikilodermia, ulceration, cutaneous necrosis, and calcinosis.

Muscle involvement was defined as proximal limb muscle weakness [evaluated by Medical Research Council (MRC) scale of ≤5), and/or dysphagia, and/or increase of creatinine phosphokinase (CK) levels and/or T2/Short-TI Inversion Recovery hypersignal on MRI, and/or pathological DM features on muscle biopsy, including either perifascicular atrophy, perimysial inflammatory infiltrate, diffuse positive HLA1 immunostaining, capillary dropout and/or dilatation and/or capillary C5b-9 deposition, or muscle infarction.

The diagnosis of ILD was based on radiological assessment of a chest CT scan by a radiologist (expert in chest radiology), and features were classified according to the American Thoracic Society/European Respiratory Society International Multidisciplinary Consensus Classification of the Idiopathic Interstitial Pneumonias [21]. A restrictive pulmonary pattern was defined by a decrease in total lung capacity to under 80% of the theoretical value on pulmonary function tests. Cancer-associated DM was diagnosed if a malignancy occurred within 3 years (before or after) of the diagnosis of DM [22].

Pathological analysis

Muscle biopsies were reviewed by muscle expert pathologists using a standardized approach. Muscle involvement was assessed in three domains (inflammatory, vascular, and muscle fibre), using the following stains: haematoxylin and eosin, anti‐HLA class I (Mouse, clone CR3/43, Dako), C5b-9 (Mouse, clone aE11, Dako), CD31 (clone JC70A, Dako^®), MxA [Mx1/2/3 (H‐285, sc‐50509), 1:100 dilution, Santa Cruz Biotechnology, USA], CD3 (Rabbit, clone 2GV6, Roche), CD20 (Mouse, clone L26, Dako) and CD68 (Mouse, clone KP1, Dako).

Comparison with cancer prevalence in the general population

Incidence of malignancy in the general population of France was obtained from the Association of the French Cancer Registries (FRANCIM; a French registry was chosen as only two patients were recruited from another country). The number of cases of cancer-associated DM in anti-SAE + DM patients was compared with the expected number of malignancies in the general population matched for sex and age, and reported by a standardized incidence ratio (SIR; number of observed incident cancers divided by the number of the expected incident cancers). Deceased patients and malignancies occurring within the 3 years before or after the diagnosis of DM were censored at the time of death or malignancy diagnosis; other patients (death or diagnosis of cancer >3 years after DM diagnosis, alive and cancer-free patients) accounted for 6 person-years. Non-French patients and patients with cancer diagnosed >3 years before DM diagnosis were excluded from the SIR analysis.

Statistical analysis

Statistical analyses were performed using Prism 9.0 (GraphPad Software). Quantitative variables are reported as median and interquartile range [IQR]. Qualitative variables are reported as percentage of the total of the available data. The characteristics of anti-SAE + DM and anti-SAE – DM patients were compared using Fisher’s exact test for categorical variables and Student’s t test for continuous variables. The P-value estimate for the SIR of cancer used Poisson modelling (appropriate for uncommon events). Tests were two-sided, and differences were considered statistically significant for P-values of ≤0.05.

Review of the literature

We searched the PubMed, EMBASE, EM-Premium, LiSSa, BDSP, and Cochrane Library databases for articles related to the association of DM and the anti-SAE autoantibody up until January 2022. We included all papers with anti-SAE + DM case(s) description and/or comparison with anti-SAE – DM patients.

Ethics

This work has been approved by the French Ministry of Research (CCTIRS no. 14.323 and AC-2013–1868) and the Research Ethics Committee of the Pitié Salpêtrière University Hospital (Paris, France) and is in compliance with the Declaration of Helsinki. The registration number in ClinicalTrials.gov is NCT04637672. Written informed consent is not necessary for this type of study under current legislation in France.

Results

Patients’ characteristics

A total of 49 anti-SAE+ DM patients from 18 French centres and 1 Finnish centre were included in this study. Anti-SAE autoantibody subtypes were represented as follows: anti-SAE type 1 n = 48/49, anti-SAE type 2 n = 2/49. The control group was composed of 85 anti-SAE–negative DM patients, made up of 32 anti-Mi2 DM, 5 anti-TIF1γ DM, 5 anti- NXP2 DM, 9 anti-MDA5 DM and 34 seronegative DM patients.

Patients were mostly females (83.7%, n = 41/49) and were median of 53 years old (43–62 years), including 3 juvenile cases. In comparison, females comprised 65.9% of the control group (anti-SAE – DM) (n = 56/85, P = 0.03), and age at diagnosis was not statistically different (anti-SAE – DM 50 years, P = 0.39). A comparison of anti-SAE + DM patients with anti-SAE – DM patients is shown in Table 1.

Skin manifestations

All patients, but two, presented with DM skin lesions. Two patients presented a sine dermatitis DM and were diagnosed as DM based on myopathological criteria. All anti-SAE + DM patients with skin changes (n = 47/49) had a typical DM skin rash; 91.8% (n = 45/49) of patients had a classical skin rash. Details of the skin manifestations are shown in Table 2.

In relation to the secondary DM lesions, 18.4% (n = 9/49) of anti-SAE + DM patients presented with cutaneous (n = 6/9) or mucosal (n = 3/9) ulcerations (Fig. 1a). The prevalence of ulceration was not statistically different from that in the control group [anti-SAE + DM patients 18.4% vs anti-SAE – DM patients 11.8% (n = 10/85), P = 0.31) (Table 1). Skin necrosis and calcinosis were each observed in >10% of anti-SAE + DM patients (Fig. 1b and c). The prevalence of calcinosis was not statistically different in the control group [anti-SAE + DM patients 10.2% vs anti-SAE – DM patients 5.9% (n = 5/85), P = 0.50). Widespread skin involvement (beyond photoexposed areas) was described in 34.7% (n = 17/49) of anti-SAE + DM patients (Fig. 1d).

Figure 1.

Cutaneous lesions in anti-SAE + DM patients. (A) Ulceration located on the upper thigh. (B) Skin necrosis. (C) Calcinosis in a juvenile DM patient. (D) Diffuse erythema of the back

Open in new tabDownload slide

Muscular manifestations

Two-third (69.4%, n = 34/49) of patients had proximal weakness. Muscle weakness was mild in most patients, and only 30.6% (n = 15/49) were scored ≤3 on the MRC scale (including 4 patients who scored ≤2). Compared with controls, anti-SAE + DM patients presented more frequently with normal strength [anti-SAE + DM patients 30.6% vs anti-SAE – DM patients 12.9% (n = 11/85), P = 0.02), and muscle weakness, when present, was milder [worst proximal MRC in anti-SAE + DM patients was 4 (3–5) vs antiSAE – DM patients was 3 (3–4), P = 0.006] (Table 1). Dysphagia was reported in 38.8% of patients (n = 19/49), including 17 patients at baseline and 2 other patients during follow-up. Frequency of dysphagia was not statistically different from that in the control group [anti-SAE + DM patients 38.8% vs anti-SAE – DM patients 41.2% (n = 35/85), P = 0.85]. Of note, when data was available, 86.7% of this group of patients (n = 13/15) had muscle disease that occurred after skin lesions, with a delay of 92 days (11–212).

The median CK level was 200 UI/l (94–440), with that of 59.2% (n = 29/49) of patients remaining less than twice the normal level (N < 130 UI/l). The CK level was statistically lower in the anti-SAE + DM patients compared with the anti-SAE – DM patients [the anti-SAE – DM patients’ CK level was 1967 (786–7592), P < 0.0001; the anti-SAE – DM patients CK was <2N in 16.4% of patients (n = 13/79), P < 0.0001] (Table 1). Detailed muscle features of the anti-SAE + DM patients are listed in Table 3.

Fifteen muscular biopsies were reviewed. Most (80.0%, n = 12/15) showed myofibre atrophy, prevailing in perifascicular areas (53.3%; n = 8/15), and 60% (n = 9/15) presented diffuse HLA-1 immunostaining. Myofibre necrosis was observed in 53.3% of the patients (n = 8/15) (Fig. 2a and b). More than a half the cases showed capillary C5b-9 deposition (57.1%, n = 8/14), capillary dilatation (53.3%, n = 8/15) and dropout (58.3%, n = 7/12) (Fig. 2c); infarction was found in 20.0% of patients (n = 3/15). There was a perimysial inflammatory infiltrate in 46.7% of the biopsies (n = 7/15) (Table 3).

Figure 2.

Muscle pathology in anti-SAE + DM patients. (A) Perifascicular atrophy and necrosis. (B) HLA1 perifascicular positivity. (C) Capillary microthrombosis with C5b9 endocapillaries; the small number of thrombi is explained by a global capillary rarefaction seen on CD31 immunostaining

Open in new tabDownload slide

Extra-cutaneous or muscular manifestations

Among patients who underwent chest CT, 21% (n = 8/39) were shown to have an ILD, associated with clinical dyspnoea in 5 out of the 8 patients. All ILDs except one (non-specific interstitial pneumonia) presented with an organizing pneumonia pattern [87.5% (n = 7/8)] (Fig. 3). The prevalence of ILD in the anti-SAE + DM patients did not differ compared with that in the control group [anti-SAE – ILD 27.9% (n = 19/68), P = 0.49]. The prevalence of dyspnoea was, however, statistically higher in the control group than in the anti-SAE + DM [anti-SAE + DM patients 26.5% (n = 13/49) vs anti-SAE – DM patients 56.1% (n = 32/57), P = 0.003].

Figure 3.

Thorax CT scans with interstitial lung disease in anti-SAE+ DM patients. Thorax CT scan evaluation revealed an interstitial lung disease defining an organizing pneumonia pattern

Open in new tabDownload slide

A minority of patients (12.5%, n = 4/32) displayed a restricted pulmonary disease. Forced vital capacity (FVC) was decreased in 15.4% of patients (FCV <80%, n = 4/26). Diffusing capacity for carbon monoxide (DLCO) was reduced in 26.7% of patients (DLCO <70%, n = 8/30). Of note, patients with anti-SAE + DM with ILD were more frequently African (62.5%, n = 5/8), compared with anti-SAE–positive DM patients without ILD (0%, n = 0/39; P < 0.0001).

Articular manifestations (arthralgia and/or synovitis) were found in 24.5% (n = 12/49) of patients, most commonly in distal joints of the upper limbs, but the difference from the control group was not statistically significant [anti-SAE – DM patients 50% (n = 31/62), P = 0.07].

Malignancy association

The median follow-up duration was 39 months (27–62). Eight out of ten malignancies fulfilled the definition of cancer-associated DM, representing 16.3% (n = 8/49) of our anti-SAE + DM patients [colorectal (n = 2, aged 57 and 67 years at diagnosis of cancer), lung (n = 2, aged 62 and 77  years at diagnosis of cancer), ovarian (n = 2, aged 71 and 76 years at diagnosis of cancer), melanoma (n = 1, aged 37 years at diagnosis of cancer), haematologic (n = 1, aged 81 years at diagnosis of cancer)].

A comparison of the cancer incidence with that of the age- and sex-matched general population in France found an increased risk of cancer in our anti-SAE + DM patients, with a SIR of 5.0 [observed n = 8 vs expected n = 1.59 in n = 229.5 person-years, (2.5–10.1) P < 0.001]. This association was present in women [SIR = 6.6 (3.3–13.2) P < 0.001], but this study could not estimate the risk in the male subset because of the limited number of cases (0 cancer cases in 8 patients).

Of note as expected, patients with anti-SAE + DM with malignancy were older (median age at diagnosis of DM of 70 years old) compared with patients with anti-SAE–positive DM without cancer (median age at diagnosis of DM of 50 years old; P = 0.02).

Treatments and outcomes

The first line of treatment consisted of oral CSs for the majority of patients (87.7%, n = 43/49), and another systemic therapy in 91.8% (n = 45/49) (MTX n = 27, HCQ n = 21, AZA n = 6, rituximab n = 2, MMF n = 2, CYC n = 1, and IVIG n = 18). Overall, 51% (n = 25/49) of patients received IVIG, the clinical indication being dysphagia in 64% of these patients. The lines of systemic treatments used are detailed in Supplementary Table S1, available at Rheumatology online.

At the final follow-up (a median of 39 months since diagnosis), 58.3% (n = 28/48) of patients were in physician-assessed complete remission, including 9 patients who were without any systemic treatment. Eighteen percent (n = 9/48) had a persistent muscular disease, and 35.4% (n = 17/48) had a persistent cutaneous disease (presence of at least one typical skin lesion). Among the half of the patients who had follow-up of at least 39 months, no patients had active muscular disease but three patients had persistent cutaneous disease.

Five patients died during the follow-up period: one from disseminated tuberculosis, three from evolutive malignancy, one from evolution of DM (a bedridden patient aged 76 years). The death rate did not differ statistically significantly from that in the control group [anti-SAE – DM patients: 8.2% (n = 7/85), P = 0.76].

Review of the literature

Twenty-seven articles met our inclusion criteria. Three case reports were excluded due to insufficient data for diagnosis of DM (n = 2) or insufficient description of the case (n = 1). The criteria of interest in each selected study are reported in Supplementary Table S2, available at Rheumatology online. The results are discussed in the following section.

Discussion

This cohort of 49 cases of anti-SAE + DM shows that (i) cutaneous manifestations are ‘typical’ and rarely severe, however, sometimes associated with a diffuse erythema; (ii) muscle involvement is frequently mild, with expected myopathological changes, (iii) ILD occurs in the same proportion as in other DM, with a characteristic pattern of organizing pneumonia and (iv) the risk of malignancy is increased compared with the general population, and malignancy was responsible for the majority of fatal outcomes.

Our results are consistent with previously published data, including the findings for 166 cases reported in 24 articles [4–17, 19, 23–31] (Supplementary Table S2, available at Rheumatology online). All reported case series, but one (Betteridge et al., 2019 [4]), reported findings for no more than 20 patients. The largest of the case series (n = 42), comparing anti-SAE + myositis vs anti-SAE – myositis (not only DM) demonstrated that the anti-SAE antibody is DM specific (rash and periungual erythema), but no further description of the cases was available.

According to our review of the literature, skin features are reported as follows: 100% cutaneous involvement (n = 111/111), with a prevalence of 80% heliotrope rash (n = 88/110), 93% Gottron’s sign or papules (n = 102/110), 66% periungual changes (n = 49/74), 69% V-neck sign (n = 51/74), 65% shawl sign (n = 44/68), 62% diffuse erythema (n = 26/42), 27% alopecia (n = 8/30), 11% calcinosis (n = 5/44) and 28% ulceration (n = 12/42) (Supplementary Table S2, available at Rheumatology online). The proportion of calcinosis was consistent among cohorts, although seldom recorded (total n = 44), and ranged between 0 and 40% [7, 25].

Despite the discrepancies among cohorts, a diffuse erythema was consistently reported. Among studies reporting this feature, ours had the lowest prevalence, and the findings of Albayda et al. [16] in a North American cohort found a similar prevalence of 42%. Inoue et al. [17] recently reported an erythema of the back excluding the interscapular region (n = 6); characterization of this manifestation would benefit a prospective setting.

With respect to the the muscle domain, our review shows a global prevalence of myopathy of 79% (n = 69/87), with 45% myalgia (n = 10/22), 71% paresia (n = 58/82), 51% having elevation of CK levels (n = 49/97), and 50% dysphagia (n = 42/84). Dysphagia prevalence may reach 64% [8] or even 78% [6] in some cases series. Frequently (58%, n = 57/98), patients presented either with clinically amyopathic DM or delayed muscle disease onset after skin involvement (‘skin first’). Amyopathic DM affected 20.7% (n = 18/87) of anti-SAE + DM patients. The severity of muscle weakness found in our study was also found by Peterson et al. (36% moderate to severe) [25], but researchers of the other series did not report a semiquantitative measure of the severity of muscle weakness. IVIG use in our study was correlated with its expected efficacy in muscle involvement according to Aggarwal et al.'s trial in a selected population [32].

One previous study described muscle biopsy findings in anti-SAE + DM (n = 7) [8]. It reported perimysial inflammatory infiltrates without the characteristic DM pattern [18]. Our findings delineated perifascicular atrophy, vasculopathic changes, and inflammatory infiltrate in only about half of the anti-SAE + DM patients. These results suggest that in patients with anti-SAE + DM, muscle damage could be moderate, which would be consistent with the milder clinical muscle features. Of note, only one patient who was biopsied was amyopathic.

Our review identified a total prevalence of 53% (n = 49/92) of ILD in anti-SAE + DM patients, ranging from 17% to 77% (excluding studies in which pulmonary involvement was an inclusion criteria, and studies reporting 2 or fewer cases). In particular, evidence of preserved pulmonary function, the absence of respiratory failure, and the presence of organizing pneumonia pattern (statistically more frequent than with other myositis-specific autoantibodies [15]) has been reported by Zuo et al. [15], Gono et al. [11] and Albayda et al. [16]. The majority (79.6%) of the patients we studied underwent a CT scan. The discrepancy between the ILD rate in our study compared with that reported in the literature may be explained by the fact that no study systematically conducted a CT evaluation; in addition, some authors have raised queries about potential ethnic disparities between Asian and Caucasian patients in the phenotypic expression of anti-SAE + DM [8, 10]. Our study population only included 5 patients each of Asian and African ethnicity, and thus lacked the power to draw any conclusions regarding this.

The cancer incidence found in our study is consistent with that reported for other cohorts, ranging from 6% to 25% [6–9, 16, 19, 25], excluding the 57% rate reported by Muro et al. [12] for their small series. The total malignancy rate found in our review of the literature was 18% (n = 17/96). In our series, we were able to measure the SIR, showing for the first time an increased risk of malignancy in patients with anti-SAE–positive DM compared with that of the general population. Patients with cancer-associated DM were older, which is concordant with previous population-based studies in DM [33, 34].

We conducted a retrospective study, with all the related limitations. Nevertheless, our results are consistent with those published in the literature. The multicentre design allowed for minimization of possible referral bias, and the prolonged follow-up duration likely limited the risk of incomplete phenotypic description (e.g. delayed muscle signs). Our review identified only one small (n = 6) prospective study [23]. The phenotypic description of the disease constructed from our review of the literature was limited by the amount of missing data for each feature, and this needs to be taken into account; our study, however, endeavoured to minimize missing data.

Conclusion

The anti-SAE autoantibody in DM is associated with ‘typical’ skin disease and potentially diffuse erythema, with a mild muscle involvement that may present with a delayed onset. ILD is most commonly present with an organizing pneumonia pattern, yet it does not worsen the overall prognosis. Cancer prevalence, however, is higher than in the general population.

Supplementary material

Supplementary material is available at Rheumatology online.

Data availability

The data that support the findings of this study are available from the corresponding author, Y.A, upon reasonable request. The data are not publicly available in accordance with French legislation.

Funding

No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article.

Disclosure statement: The authors have declared no conflicts of interest.

Acknowledgements

We thank the Study Group of Systemic Diseases in Dermatology (Étude des Maladies Systémiques en Dermatologie, EMSED group), who provided help in identification of cases. This work was presented as an abstract at the 83rd Société Nationale Française de Médecine Interne (SNFMI) Conference 2021.

References

Author notes