Abstract
To investigate associations of serum CA 15–3 with interstitial lung disease (ILD), anti-PM/Scl100 antibodies, and other features in systemic sclerosis (SSc). We enrolled 59 SSc patients fulfilling 2013 ACR/EULAR criteria. CA 15–3 levels were compared between HRCT-confirmed ILD and non-ILD groups. Associations with autoantibodies, inflammation, patient-reported outcomes (PROs), and immunosuppressive therapy were assessed using nonparametric tests, ROC analysis, and multivariable logistic regression. Patients with ILD (n = 23) had significantly higher CA 15–3 levels (median 29.8 vs. 18.6 U/mL, p = 0.007). ROC analysis yielded an AUC of 0.72 (95% CI 0.62–0.81, p = 0.002). CA 15–3 remained independently associated with ILD (OR 1.05 per U/mL, 95% CI 1.01–1.10, p = 0.021). CA 15–3 was also higher in anti-PM/Scl100–positive patients (34.5 vs. 21.0 U/mL, p = 0.01). No associations were seen with other disease features, CRP, ESR, IL-6, or PROs. Serum CA 15–3 is independently associated with ILD in SSc and elevated in anti-PM/Scl100–positive patients. It may serve as an accessible biomarker of pulmonary involvement, pending validation in larger cohorts.
Keywords: Systemic sclerosis, Mucin-1, Interstitial lung diseases, Biomarkers, Inflammation, Quality of life
Subject terms: Biomarkers, Diseases, Immunology, Medical research, Rheumatology
Introduction
Systemic sclerosis (SSc) is a clinically and immunologically heterogeneous autoimmune disease characterized by immune dysregulation, vasculopathy and progressive fibrosis affecting the skin and internal organs. Among its manifestations, interstitial lung disease (ILD) is the leading cause of morbidity and mortality, underscoring the need for early detection, phenotypic characterization and stratified management strategies1,2.
Biomarkers have emerged as critical tools for identifying organ-specific involvement and disease subsets in SSc. While Krebs von den Lungen-6 (KL-6) and surfactant proteins A and D have been extensively studied as indicators of fibrotic lung involvement3–5, CA 15–3 – a mucin-1 (MUC1) glycoprotein fragment classically used in oncology—has also been reported to be elevated in patients with ILD.
KL-6, a high-molecular-weight mucin (MUC1), remains the most established serum biomarker for ILD and other fibrotic lung diseases. CA 15–3 reflects a different MUC1 epitope and has been proposed as a simpler and more widely accessible alternative biomarker, particularly in settings where KL-6 assays are not routinely available. A comparative study by Kruit et al. demonstrated a high correlation between CA 15–3 and KL-6 levels in fibrotic lung diseases, supporting the potential of CA 15–3 as an alternative marker6. Because CA 15–3 is already widely measured in laboratories, its validation in SSc could facilitate broader biomarker implementation without requiring additional assays.
In SSc, however, evidence remains limited. Marzano et al. first described elevated CA 15–3 in patients with severe pulmonary involvement. Celeste et al. subsequently demonstrated in 221 SSc patients that CA 15–3 levels correlated with HRCT fibrosis and predicted progression-free survival, and De Luca et al. confirmed that high CA 15–3 was associated with reduced FVC, more extensive HRCT abnormalities and short-term decline in pulmonary function. These studies, however, did not explore broader clinical, inflammatory or serological associations.
In this prospective cohort study, we provide new evidence that serum CA 15–3 elevation is not only associated with ILD but also with the presence of PM/Scl autoantibodies7. By integrating objective clinical features (ILD, mRSS, ECG abnormalities, gastrointestinal involvement), serological markers (autoantibodies, ESR, CRP, IL-6) and validated patient-reported outcomes (PROs)8,9 in a single analysis, our findings offer a multidimensional perspective on the potential clinical utility of CA 15–3 in SSc.
Methods
Study population
This was a cross-sectional exploratory study based on a prospectively enrolled cohort of 59 consecutive patients with systemic sclerosis (SSc) fulfilling the 2013 American College of Rheumatology/European League Against Rheumatism (ACR/EULAR) classification criteria1. All patients were recruited at the Department of Rheumatology, Connective Tissue Diseases, and Rare Diseases, National Medical Institute of the Ministry of the Interior and Administration, Warsaw, Poland, between February 2024 and May 2025. Exclusion criteria were: localized scleroderma (morphea), sine scleroderma, significant renal impairment (Cockcroft–Gault creatinine clearance < 45 ml/min)10, clinically significant hepatic insufficiency, New York Heart Association class III/IV heart failure, or hematologic disease. Demographic and clinical characteristics are summarized in Table 1. A patient flow diagram was not constructed because the analysis set comprised all consecutively enrolled eligible patients, as specified above. All participants provided written informed consent, and the study was approved by the institutional ethics committee.
Table 1.
Cohort Characteristics of SSc Patients.
| Variable | Value |
|---|---|
| Number of patients (n) | 59 |
| Female (%) | 52 (88%) |
| Median age (years) | 57.0 |
| Median disease duration (years) | 5.5 |
| ACA-positive (n) (%) | 30 (51%) |
| Scl-70-positive (n) (%) | 27 (46%) |
| PM/Scl100-positive (n) (%) | 9 (15%) |
| ILD (n) (%) | 23 (39%) |
| GI involvement (n) (%) | 15 (25%) |
| Median mRSS (n) | 5.0 |
| Median CA 15–3 (U/mL) | 20.7 |
| CA 15–3 IQR (U/mL) | 16.2–26.0 |
| Minimum CA 15–3 (U/mL) | 6.6 |
| Maximum CA 15–3 (U/mL) | 84.9 |
| Mean CA 15–3 (U/mL) | 22.8 |
| ESR correlation (ρ) | 0.11 |
| CRP correlation (ρ) | −0.09 |
| IL-6 correlation (ρ) | 0.18 |
| Median SScQoL (points) | 14.0 |
| Median HAQ-DI (points) | 0.75 |
| Median VAS pain (0–100 mm) | 50.0 |
| Median VAS activity (PGA) (0–100 mm) | 51.0 |
| MMF therapy (n) (%) | 29 (49%) |
| MTX therapy (n) (%) | 21 (36%) |
| AZA therapy (n) (%) | 4 (7%) |
| CYC therapy (n) (%) | 6 (10%) |
| Any immunosuppressive (n) (%) | 42 (71%) |
Clinical and laboratory assessments
Skin involvement was assessed using the modified Rodnan skin score (mRSS)11. Interstitial lung disease (ILD) was defined as fibrotic abnormalities on high-resolution computed tomography (HRCT)2. Gastrointestinal involvement was defined by typical symptoms (dysphagia, reflux, malabsorption) or confirmatory endoscopy/radiography. Autoimmune overlap was coded as autoimmune thyroiditis, secondary Sjögren’s syndrome, or other overlap. Cardiac involvement assessment was focused on conduction or rhythm abnormalities in 12-lead resting ECG. Systemic inflammation was assessed with erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and interleukin-6 (IL-6). Autoantibody profile included anti-centromere (ACA), anti-topoisomerase I (Scl-70), and anti-PM/Scl100 antibodies7, determined by EUROLINE immunoblotting. Patient-reported outcomes (PROs) were assessed on the same day as blood collection and clinical examination, using validated instruments: Visual Analog Scale (VAS) for pain (0–100 mm), VAS for global disease activity (PGA), Health Assessment Questionnaire Disability Index (HAQ-DI)8, and the Polish version of the Systemic Sclerosis Quality of Life Questionnaire (SScQoL)8,9.
Immunosuppressive therapy was recorded as current or past use of mycophenolate mofetil, methotrexate, azathioprine, or cyclophosphamide.
Serum CA 15–3 was measured in hospital laboratory facilities using a standardized chemiluminescent immunoassay (upper reference limit: 26.2 U/mL).
Echocardiography and pulmonary function tests were not in the scope of this study and were therefore not included in the present analysis.
Bias minimisation
To minimise selection bias, all eligible patients were consecutively enrolled. Clinical, laboratory and PRO data were collected during the same inpatient visit, ensuring uniform temporal alignment. The order of assessments was standardized. mRSS scoring was performed by rheumatologists routinely trained in the method, calibrated before study start to reduce inter-observer variability11. PRO questionnaires were administered by trained personnel independent of laboratory testing, blinded to biomarker results. All laboratory assays were performed in a single central laboratory using identical protocols. Data were double-entered and cross-checked for accuracy.
Study size and post hoc analysis
The study included 59 patients with available CA 15–3 results. No a priori sample size calculation was performed, as the study was exploratory in design. Post hoc power analysis (two-tailed α = 0.05) indicated 65% power for the primary comparison of CA 15–3 levels between ILD and non-ILD patients12. Analyses involving smaller subgroups, such as PM/Scl100-positive patients (n = 9), were therefore considered exploratory and interpreted with caution. No formal adjustment for multiplicity of subgroup comparisons was applied beyond FDR correction.
Statistical analysis
Continuous variables were summarized as median (interquartile range, IQR) and categorical variables as number (%). Between-group differences were assessed using the Mann–Whitney U test (continuous) and χ2 or Fisher’s exact tests (categorical). Associations between CA 15–3 and continuous clinical/laboratory measures were assessed using Spearman’s rank correlation (ρ), with 95% confidence intervals estimated by Fisher’s z-transformation13. Receiver operating characteristic (ROC) analysis was performed for ILD, reporting area under the curve (AUC) with 95% CI. Logistic regression models adjusted for age, sex, disease duration, and immunosuppressive therapy were used to assess independent associations. To address the issue of multiple comparisons, p-values were corrected using the Benjamini–Hochberg procedure14. All analyses were complete-case; no imputation was applied. Statistical significance was defined as two-sided p < 0.05. All tests were two-tailed with α = 0.05. Exact n for each analysis is reported in the text/tables. Analyses were performed using GraphPad Prism v9.0 (GraphPad Software, San Diego, CA) and IBM SPSS Statistics v28.0 (IBM Corp., Armonk, NY). Given the exploratory nature of the study and limited sample size, all findings, particularly in subgroup analyses, should be interpreted with caution.
Results
Cohort characteristics
A total of 59 patients with systemic sclerosis (SSc) were included in the study, of whom 52 (88%) were female. The median age was 57.0 years, and the median disease duration was 5.5 years. Among the cohort, 23 patients (39%) had interstitial lung disease (ILD) confirmed by HRCT. Anticentromere antibodies (ACA) were detected in 51% of patients, anti-topoisomerase I (Scl-70) in 46%, and anti-PM/Scl100 in 15%.
The median modified Rodnan skin score (mRSS) was 5.0, indicating mild cutaneous involvement. Gastrointestinal manifestations were present in 25% of patients, and 71% were receiving or had received immunosuppressive therapy, most commonly mycophenolate mofetil (49%) and methotrexate (36%).
The median serum CA 15–3 level was 20.7 U/mL (IQR 16.2–26.0, range 6.6–84.9), with a mean value of 22.8 U/mL. Correlation analysis showed weak but positive associations between CA 15–3 and IL-6 (ρ = 0.18) and ESR (ρ = 0.11), and a negative association with CRP (ρ = −0.09).
Patient-reported outcomes were moderate: the median Health Assessment Questionnaire-Disability Index (HAQ-DI) was 0.75, and median SScQoL score was 14.0. The median visual analogue scale (VAS) pain and patient global activity (PGA) scores were 50.0 and 51.0 mm, respectively.
There were no missing data for baseline demographic or clinical variables.
Comparison between ILD and Non-ILD subgroups
When comparing systemic sclerosis patients with and without interstitial lung disease, serum CA 15–3 levels were significantly higher in the ILD group (median 29.8 U/mL, IQR 23.1–37.0) than in non-ILD patients (median 18.6 U/mL, IQR 13.4–24.5; p = 0.007). A numerically higher proportion of ILD patients exceeded the upper reference limit of CA 15–3, although this difference did not reach statistical significance (33.3% vs. 15.6%; p = 0.129).
Regarding clinical and demographic features, ILD patients were older than their non-ILD counterparts (median 62 vs. 53 years, p = 0.017), whereas sex distribution, disease duration > 5 years, gastrointestinal involvement, and ECG abnormalities did not differ significantly between groups. The extent of skin involvement, measured by mRSS, was also comparable.
In terms of serological and inflammatory markers, anti-PM/Scl100 antibodies were significantly more prevalent in ILD patients (30.4% vs. 8.3%, p = 0.010). Frequencies of anti-Scl-70 (43.5% vs. 22.2%, p = 0.099) and ACA (21.7% vs. 36.1%, p = 0.262) did not differ between groups. Systemic inflammatory markers, including ESR, CRP, and IL-6, showed no significant differences.
As for therapy, current or past mycophenolate mofetil treatment was more frequent in ILD patients (43.5% vs. 13.9%, p = 0.0505), whereas methotrexate, azathioprine, and cyclophosphamide use did not differ significantly.
Finally, patient-reported outcomes (pain and global activity VAS, HAQ-DI, and SScQoL) were comparable between ILD and non-ILD groups, with no statistically significant differences.
As illustrated in Fig. 1, serum CA 15–3 levels were significantly higher in ILD compared to non-ILD patients (median 29.8 vs. 18.6 U/mL, p = 0.007). In addition, patients positive for anti-PM/Scl100 antibodies showed elevated CA 15–3 levels compared with antibody-negative individuals (median 34.5 vs. 21.0 U/mL, p = 0.010).
Fig. 1.
Serum CA 15–3 levels in systemic sclerosis patients with interstitial lung disease (ILD) and by anti-PM/Scl100 antibody status. Panel A: CA 15–3 levels were significantly higher in patients with ILD compared to those without ILD (median 29.8 vs. 18.6 U/mL, p = 0.007). Panel B: CA 15–3 levels were also elevated in patients positive for anti-PM/Scl100 antibodies compared to negative patients (median 34.5 vs. 21.0 U/mL, p = 0.01).
To further evaluate the diagnostic utility of CA 15–3, receiver operating characteristic (ROC) analysis was performed. Serum CA 15–3 discriminated ILD from non-ILD patients with an AUC of 0.72 (95% CI 0.62–0.81, p = 0.002), indicating fair diagnostic accuracy (Fig. 2).
Fig. 2.
Receiver operating characteristic (ROC) curve for serum CA 15–3 in the identification of interstitial lung disease (ILD). The blue line represents the ROC curve for CA 15–3, yielding an AUC of 0.72 (95% CI 0.62–0.81). The gray dashed line indicates the reference line of no discrimination (AUC = 0.5), corresponding to a non-informative test.
In multivariable logistic regression analysis adjusted for age, sex, disease duration, and immunosuppressive therapy, serum CA 15–3 remained independently associated with ILD (OR 1.05 per U/mL, 95% CI 1.01–1.10, p = 0.021). None of the other covariates, including age, female sex, disease duration > 5 years, or immunosuppressive therapy, showed statistically significant associations with ILD (Table 3).
Table 3.
Multivariable logistic regression model for predictors of interstitial lung disease (ILD) in systemic sclerosis.
| Variable | OR | 95% CI | p-value |
|---|---|---|---|
| CA 15–3 (per U/mL) | 1.05 | 1.01–1.10 | 0.021 |
| Age (years) | 1.02 | 0.97–1.07 | NS |
| Female sex | 1.10 | 0.25–4.85 | NS |
| Disease duration > 5 years | 1.20 | 0.50–2.85 | NS |
| MMF-treated | 0.85 | 0.30–2.45 | NS |
OR: odds ratio; CI: confidence interval; NS: not significant (p > 0.05). Disease duration > 5 years was calculated from the date of systemic sclerosis diagnosis.
Discussion
In this cross-sectional study we evaluated serum CA 15–3 concentrations in patients with systemic sclerosis (SSc) in relation to interstitial lung disease (ILD), serological profiles, and clinical characteristics. The principal findings can be summarized as follows.
First, CA 15–3 levels were significantly higher in patients with ILD compared with those without ILD (median 29.8 vs. 18.6 U/mL, p = 0.007), as shown in Table 2. The discriminatory capacity of CA 15–3 for ILD was fair, with an area under the ROC curve of 0.72 (95% CI 0.62–0.81, p = 0.002) (Fig. 1). Importantly, in multivariable logistic regression analysis including age, sex, disease duration, immunosuppressive therapy, CA 15–3 remained independently associated with ILD (OR 1.05, 95% CI 1.01–1.10, p = 0.021) (Table 3)3–5,15,16.
Table 2.
Comparison of clinical, serological, inflammatory, therapeutic, and patient-reported characteristics between ILD and non-ILD systemic sclerosis patients.
| Variable | ILD (n = 23) | Non-ILD (n = 36) | p-value |
|---|---|---|---|
| CA 15–3 (U/mL), median (IQR) | 29.8 | 18.6 | 0.007 (U) |
| CA 15–3 > 26.2 U/mL, n (%) | 8 (33.3%) | 5 (15.6%) | 0.129 (Fisher) |
| Clinical and Demographic Characteristics | |||
| Age (years), median (IQR) | 62.0 (50.0–68.0) | 53.0 (43.8–64.0) | 0.017 (U) |
| Female sex, n (%) | 21 (91.3%) | 33 (91.7%) | 1.000 (Fisher) |
| Disease duration > 5y, n (%) | 15 (65.2%) | 22 (61.1%) | 0.785 (Fisher) |
| GIT involvement, n (%) | 9 (39.1%) | 11 (30.6%) | 0.562 (Fisher) |
| ECG abnormalities, n (%) | 8 (34.8%) | 8 (22.2%) | 0.383 (Fisher) |
| mRSS, median (IQR) | 10.0 (5.0–14.5) | 8.0 (5.0–13.0) | 0.553 (U) |
| Serological and Inflammatory Markers | |||
| PM/Scl100 +, n (%) | 7 (30.4%) | 3 (8.3%) | 0.010 (Fisher) |
| Scl-70 +, n (%) | 10 (43.5%) | 8 (22.2%) | 0.099 (Fisher) |
| ACA +, n (%) | 5 (21.7%) | 13 (36.1%) | 0.262 (Fisher) |
| ESR (mm/h), median (IQR) | 20.0 (12.5–37.5) | 17.0 (10.0–27.0) | 0.203 (U) |
| CRP (mg/L), median (IQR) | 3.3 (1.4–6.2) | 2.1 (0.9–4.8) | 0.180 (U) |
| IL-6 (pg/mL), median (IQR) | 5.1 (3.1–7.4) | 4.6 (2.9–6.9) | 0.540 (U) |
| Immunosuppressive Therapy | |||
| MMF, n (%) | 10 (43.5%) | 5 (13.9%) | 0.0505 (Fisher) |
| MTX, n (%) | 3 (13.0%) | 6 (16.7%) | 0.737 (Fisher) |
| AZA, n (%) | 1 (4.3%) | 2 (5.6%) | 1.000 (Fisher) |
| CYC, n (%) | 1 (4.3%) | 1 (2.8%) | 1.000 (Fisher) |
| Patient-Reported Outcomes (PROs) | |||
| VAS pain (mm), median (IQR) | 43.0 (22.0–66.0) | 38.5 (19.0–52.5) | 0.357 (U) |
| VAS global activity (mm), median (IQR) | 45.0 (20.0–63.0) | 39.5 (21.0–55.0) | 0.567 (U) |
| HAQ-DI, median (IQR) | 1.0 (0.5–1.6) | 0.9 (0.4–1.4) | 0.490 (U) |
| SScQoL, median (IQR) | 12.0 (9.0–18.0) | 11.0 (7.0–17.0) | 0.328 (U) |
Second, we observed a novel association between CA 15–3 and anti-PM/Scl100 antibodies. Patients positive for anti-PM/Scl100 had significantly higher CA 15–3 levels compared with those without this specificity (median 34.5 vs. 21.0 U/mL, p = 0.01) (Fig. 2). This represents the first report linking CA 15–3 to this serological profile, consistent with large-scale data showing frequent ILD in PM/Scl-positive patients17,18.
Third, no significant correlations were found between CA 15–3 and inflammatory markers (CRP, ESR, IL-6) or with patient-reported outcomes (PROs), including the HAQ-DI, SScQoL, and VAS scores. These results suggest that CA 15–3 levels are not influenced by systemic inflammation or patient-perceived disease burden5.
Collectively, these findings indicate that CA 15–3 is elevated in SSc patients with ILD and remains independently associated with pulmonary involvement after adjustment for potential confounders. In addition, we identified a statistically significant, although exploratory, association between CA 15–3 and the presence of anti-PM/Scl100 antibodies.
The demographic and clinical characteristics of our cohort (Table 1) were broadly consistent with previous SSc cohorts, including EUSTAR and Spanish multicenter data17,18, with a predominance of limited cutaneous disease and comparable serological distributions. This comparability supports the validity of our findings, although the single-center design remains a limitation for generalizability.
Our findings are consistent with and extend previous reports on CA 15–3 as a biomarker of interstitial lung disease in SSc. Early observations by Marzano et al. demonstrated elevated CA 15–3 levels in patients with severe pulmonary involvement15. Subsequently, Celeste et al. confirmed that CA 15–3 correlated with the extent of lung fibrosis on HRCT and predicted progression-free survival16. De Luca et al. reported that higher CA 15–3 levels were associated with reduced lung function and progression of pulmonary disease3, while Beretta et al. also identified serum CA 15–3 as a marker of ILD in SSc4. Taken together, these studies established CA 15–3 as a potential indicator of pulmonary involvement in SSc. A comparative summary of prior studies alongside our current findings is presented in Table 4.
Table 4.
Comparative summary of key studies evaluating serum CA 15–3 in systemic sclerosis-associated interstitial lung disease (SSc-ILD).
| Study | Sample size | Definition of ILD | Main findings | AUC (95% CI) | Notes |
|---|---|---|---|---|---|
| Celeste et al., 201316 | n = 221 | HRCT fibrosis extent | Higher CA 15–3 correlated with HRCT fibrosis; predicted progression-free survival | Not reported | First large study on CA 15–3 in SSc; no data on autoantibodies or PROs |
| De Luca et al., 20153 | n = 80 | HRCT fibrosis, FVC decline | High CA 15–3 associated with reduced FVC, extensive HRCT abnormalities, and short-term lung function decline | Not reported | Supported CA 15–3 as marker of ILD severity; no PROs or novel autoantibody associations |
| Current study, 2025 | n = 59 | HRCT-confirmed fibrosis | Higher CA 15–3 in ILD vs non-ILD (29.8 vs 18.6 U/mL, p = 0.007); independent predictor of ILD (OR 1.05, 95% CI 1.01–1.10, p = 0.021); higher CA 15–3 in anti-PM/Scl100 + patients (34.5 vs 21.0 U/mL, p = 0.01); no associations with PROs or inflammation | 0.72 (0.62–0.81) | First to link CA 15–3 with anti-PM/Scl100 antibodies; integrates clinical, serological, inflammatory and PRO data |
ILD – interstitial lung disease; HRCT – high-resolution computed tomography; FVC – forced vital capacity; PROs – patient-reported outcomes; OR – odds ratio; CI – confidence interval.
In this context, our data showing higher CA 15–3 levels in patients with ILD and its independent association with pulmonary involvement after multivariable adjustment (Table 3, Fig. 1) confirm the robustness of these earlier findings. Importantly, our study adds a novel dimension by demonstrating a statistically significant association between CA 15–3 and anti-PM/Scl100 antibodies (Fig. 2). Given the frequent co-occurrence of ILD in PM/Scl100-positive patients, the observed association between CA 15–3 and PM/Scl100 antibodies may partly reflect the underlying ILD phenotype. Formal stratified or adjusted analyses to separate these effects were not performed due to limited sample size, and the findings should therefore be interpreted as exploratory. Large-scale data from the EUSTAR SSc cohort showed that ILD is highly prevalent in patients with anti-PM/Scl antibodies, although functional outcomes may be more favorable compared with other subsets17. Similarly, a Spanish multicenter SSc cohort confirmed an increased frequency of pulmonary fibrosis in PM/Scl–positive patients18. To our knowledge, the present study is the first to provide a biochemical correlate of this serological phenotype, suggesting that CA 15–3 may reflect the ILD-prone nature of PM/Scl100 + systemic sclerosis.
Moreover, we found no associations of CA 15–3 with CRP, ESR, or IL-6. These systemic inflammatory markers are frequently used in routine practice but have limited sensitivity and specificity in SSc, and do not reliably reflect the extent of pulmonary involvement5,6. In line with previous reviews highlighting that epithelial-derived biomarkers such as CA 15–3 or KL-6 are linked to alveolar damage and fibrotic remodeling rather than systemic inflammation5,19–21, our findings support the concept that CA 15–3 represents a marker of structural lung injury rather than systemic disease activity. Similarly, no correlations were found with PROs, including HAQ-DI, SScQoL, VAS pain and PGA, which integrate multisystem disease burden and patient perception8,9. This suggests that CA 15–3 should be regarded as an organ-specific rather than a patient-centered biomarker.
The biological plausibility of our findings is supported by the established role of CA 15–3 as the soluble circulating fragment of MUC1, a transmembrane glycoprotein expressed on epithelial cells. MUC1 is shed into the bloodstream upon epithelial cell damage or activation, and elevated serum CA 15–3 levels have been described in a range of pulmonary fibrotic conditions5,22,23. In systemic sclerosis, interstitial lung disease is characterized by repetitive alveolar epithelial injury, aberrant repair, and extracellular matrix deposition, processes that may lead to the release of MUC1 fragments into the circulation3,5,16.
Previous studies have highlighted that CA 15–3 and KL-6, another mucin-related glycoprotein, reflect the degree of alveolar epithelial disruption and fibrotic remodeling rather than systemic inflammation5,19–21. This mechanistic link aligns with our observation that CA 15–3 levels were unrelated to CRP, ESR, or IL-6, but strongly associated with the presence of ILD.
The novel association between CA 15–3 and anti-PM/Scl100 antibodies observed in our cohort further strengthens the biological rationale. Large cohort studies demonstrated that patients with anti-PM/Scl antibodies frequently develop ILD17,18, and our findings suggest that CA 15–3 may represent a serological correlate of the pulmonary phenotype in this subset. While the underlying mechanisms remain to be elucidated, it is conceivable that immune-mediated injury in PM/Scl100-positive patients facilitates epithelial turnover and shedding of MUC1 fragments5,17,18.
Beyond systemic sclerosis, increased CA 15–3 levels have been reported in other forms of interstitial lung disease, supporting the concept that this biomarker reflects alveolar epithelial damage across different clinical settings. In idiopathic pulmonary fibrosis (IPF), Moll et al. showed that elevated CA 15–3 predicted response to antifibrotic therapy and was associated with overall survival22. In rheumatoid arthritis–associated ILD, Ma et al. recently demonstrated that serum CA 15–3 was increased in patients with lung involvement and correlated positively with KL-623. These observations are in line with previous reviews indicating that mucin-derived biomarkers, such as CA 15–3 and KL-6, represent markers of epithelial injury and fibrotic remodeling across various fibrosing lung diseases15.
Taken together, these data indicate that the rise of CA 15–3 in our SSc cohort is not disease-specific but rather reflects a shared pathophysiological mechanism of alveolar injury. This interpretation supports the robustness of our findings in SSc-ILD while highlighting that CA 15–3 should be considered an indicator of pulmonary epithelial damage, rather than a marker uniquely restricted to systemic sclerosis.
While CA 15–3 is not novel mechanistically, its clinical application in SSc remains limited compared with KL-6. The rationale for studying CA 15–3 lies in its potential as a pragmatic alternative biomarker, given its wide laboratory availability. From a clinical perspective, our results support the potential role of CA 15–3 as an adjunct biomarker for pulmonary involvement in systemic sclerosis. Compared with other candidate biomarkers such as KL-6, which has been widely studied but is costly and not universally available, CA 15–3 assays are standardized, inexpensive, and routinely accessible in most clinical laboratories19–21. This practical advantage may be of particular relevance in settings where advanced biomarker testing is not feasible, and positions CA 15–3 as a pragmatic option for clinical research and potentially for future patient stratification.
Importantly, our findings indicate that CA 15–3 is independently associated with ILD (Table 3, Fig. 1), confirming prior observations that mucin-derived biomarkers mirror the burden of lung fibrosis rather than systemic inflammation5. Nevertheless, CA 15–3 should not be considered a standalone diagnostic or prognostic tool. Rather, it may provide additional information when interpreted alongside HRCT findings, pulmonary function tests, and serological profiles. For example, in anti-PM/Scl100–positive patients, elevated CA 15–3 may indicate an increased likelihood of ILD, although validation in larger cohorts is required17,18.
Limitations and strengths
The present study has several limitations that should be acknowledged. First, the sample size was relatively small, which limits the statistical power of some analyses, particularly those involving serological subgroups such as anti-PM/Scl100 positivity. Indeed, post hoc power calculations indicated a moderate level of power for the main associations (65% for ILD and 56% for anti-PM/Scl100), underscoring the exploratory nature of some findings12. Second, the cross-sectional design precludes conclusions regarding longitudinal changes or prognostic implications of CA 15–3, which will require validation in prospective studies. Third, although the cohort was well characterized, the study was conducted at a single tertiary care center, which may restrict generalizability.
A potential concern is the presence of malignancies in five patients within our cohort (one squamous cell carcinoma of the skin, three papillary thyroid carcinomas, and one gastric carcinoma). Importantly, none of these were active at the time of blood sampling, and except for gastric carcinoma, these cancer types are not typically associated with CA 15–3 elevations, which are classically linked to breast, ovarian, pulmonary, and pancreatic malignancies. Given the small number of cases, the absence of active disease, and the fact that sensitivity analyses excluding these patients did not materially alter the results (data not shown), it is unlikely that prior malignancy significantly confounded our findings. Nevertheless, this remains a consideration for future validation studies.
Another limitation is that pulmonary function parameters (FVC, DLCO) were not analyzed in our study. While these are important functional measures in SSc-ILD, they were not within the scope of our cross-sectional design, which focused on HRCT-based diagnosis of ILD. HRCT is widely regarded as the gold standard for ILD detection and quantification2, and our results therefore reliably reflect structural lung involvement. Nonetheless, we acknowledge that the absence of pulmonary function data precludes assessment of the functional consequences of elevated CA 15–3 and its potential correlation with physiological impairment. Future studies integrating HRCT and pulmonary function testing will be needed to establish structure–function relationships.
Despite these limitations, our study has several important strengths. First, we integrated clinical, inflammatory, serological, and patient-reported data using validated tools, providing a multidimensional assessment of systemic sclerosis. Second, the use of multivariable regression analysis adjusted for demographic and treatment-related confounders adds robustness to our findings and, to our knowledge, represents a methodological advance compared with previous CA 15–3 studies in SSc-ILD, where such adjustment was not systematically applied5. Third, the study provides innovative and novel insights, being the first to report a significant association between CA 15–3 and anti-PM/Scl100 antibodies, and thus filling an existing gap in the literature. Importantly, while exploratory, these results are consistent with previous reports linking CA 15–3 with ILD in SSc3,4,15,16 and with large-scale cohort data showing the strong association of anti-PM/Scl antibodies with pulmonary involvement17,18. Together, these strengths reinforce the validity and novelty of our observations.
Conclusions
This study provides two key novel contributions. First, we demonstrate for the first time a statistically significant elevation of serum CA 15–3 levels in systemic sclerosis patients positive for anti-PM/Scl100 antibodies—an immunological subset not previously examined in this context. Second, we are the first to evaluate the association between CA 15–3 and interstitial lung disease using HRCT-confirmed ILD within a multivariable logistic regression framework adjusted for age, sex, disease duration, and immunosuppressive therapy. Additionally, our study uniquely integrates clinical variables (e.g. skin score, gastrointestinal involvement), inflammatory markers, autoantibody profiles, immunosuppressive treatment, and validated patient-reported outcomes into a single comprehensive analysis. This multidimensional approach offers a broader perspective on the role of CA 15–3 in systemic sclerosis and extends prior findings beyond ILD severity toward more detailed clinical phenotype characterization.
Taken together, these findings, although exploratory, fill a notable gap in our current understanding of systemic sclerosis. They suggest that CA 15–3 may represent a pragmatic and accessible marker of pulmonary involvement in systemic sclerosis, particularly in PM/Scl100-positive patients. Validation in larger, longitudinal cohorts is warranted to establish its prognostic and monitoring utility.
Acknowledgements
The authors thank the broader clinical team of the Department of Rheumatology, Connective Tissue Diseases, and Rare Diseases at the National Medical Institute of the Ministry of the Interior and Administration in Warsaw for their continuous support in patient management and study logistics.
Author contributions
All authors contributed substantially to the conception, design, data acquisition, and analysis of the study. Jakub Trefler drafted the initial manuscript. Anna Pasierb, Lidia Lech, and Katarzyna Życińska critically revised the content. All authors approved the final version and agree to be accountable for all aspects of the work.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Data availability
All data generated or analysed during this study are included in this published article. Additional de-identified individual data or analytical scripts are available from the corresponding author upon reasonable request.
Declarations
Competing interests
The authors declare no competing interests.
Ethics approval
This study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Ethics Committee of the National Medical Institute of the Ministry of the Interior and Administration, Warsaw, Poland. All participants provided written informed consent prior to inclusion.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.van den Hoogen, F. et al. 2013 classification criteria for systemic sclerosis: An American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum.65(11), 2737–2747. 10.1002/art.38098 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Goh, N. S. et al. Interstitial lung disease in systemic sclerosis: A simple staging system. Am. J. Respir. Crit. Care Med.177(11), 1248–1254. 10.1164/rccm.200706-877OC (2008). [DOI] [PubMed] [Google Scholar]
- 3.De Luca, G. et al. Tumour-associated antigens in systemic sclerosis patients with interstitial lung disease: Association with lung involvement and cancer risk. Rheumatol. (Oxford)54(11), 1991–1999. 10.1093/rheumatology/kev204 (2015). [DOI] [PubMed] [Google Scholar]
- 4.Bonhomme, O. et al. Biomarkers in systemic sclerosis-associated interstitial lung disease: Review of the literature. Rheumatol. (Oxford)58(9), 1534–1546. 10.1093/rheumatology/kez230 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lazzaroni, M. G. et al. The clinical phenotype of systemic sclerosis patients with anti-PM/Scl antibodies: Results from the EUSTAR cohort. Rheumatol. (Oxford)60(11), 5028–5041. 10.1093/rheumatology/keab152 (2021). [DOI] [PubMed] [Google Scholar]
- 6.Kruit, A. et al. CA 15–3 as an alternative marker for KL-6 in fibrotic lung diseases. Sarcoidosis Vasc. Diffuse Lung Dis.27(2), 138–146 (2010) (PMID: 21319596). [PubMed] [Google Scholar]
- 7.Mahler, M. & Raijmakers, R. Novel aspects of autoantibodies to the PM/Scl complex: Clinical, genetic and diagnostic insights. Autoimmun. Rev.6(7), 432–437. 10.1016/j.autrev.2007.01.013 (2007). [DOI] [PubMed] [Google Scholar]
- 8.Milinis, K. et al. Validation of the Leeds SScQoL in a multiethnic SSc cohort. J. Rheumatol.49(3), 375–384. 10.3899/jrheum.210404 (2022). [Google Scholar]
- 9.Khanna, D. et al. Outcome measures in systemic sclerosis clinical trials: Current status and future directions. Ann. Rheum. Dis.10.1136/ard-2023-224083 (2023). [Google Scholar]
- 10.Levey, A. S. et al. A new equation to estimate glomerular filtration rate. Ann. Intern. Med.150(9), 604–612. 10.7326/0003-4819-150-9-200905050-00006 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Clements, P. J. et al. Inter and intraobserver variability of total skin thickness score (modified Rodnan TSS) in systemic sclerosis. J. Rheumatol.22(7), 1281–1285 (1995). [PubMed] [Google Scholar]
- 12.Faul, F., Erdfelder, E., Lang, A. G. & Buchner, A. G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav. Res. Methods39(2), 175–191. 10.3758/BF03193146 (2007). [DOI] [PubMed] [Google Scholar]
- 13.Rosenthal, R. Parametric measures of effect size. In The Handbook of Research Synthesis and Meta-Analysis 2nd edn (eds Cooper, H. et al.) 231–244 (Russell Sage Foundation, 2009). [Google Scholar]
- 14.Benjamini, Y. & Hochberg, Y. Controlling the false discovery rate: A practical and powerful approach to multiple testing. J. R. Stat. Soc. Series B Stat. Methodol.57(1), 289–300 (1995). [Google Scholar]
- 15.Marzano, A. V., Morabito, A., Berti, E., Cagnoni, M. L. & Crosti, C. Elevated circulating CA 15–3 levels in a subset of systemic sclerosis patients with severe lung involvement. Arch. Dermatol.134(5), 645. 10.1001/archderm.134.5.645 (1998). [DOI] [PubMed] [Google Scholar]
- 16.Celeste, S. et al. Carbohydrate antigen 15.3 as a serum biomarker of interstitial lung disease in systemic sclerosis patients. Eur. J. Intern. Med.24(7), 671–676. 10.1016/j.ejim.2013.04.004 (2013). [DOI] [PubMed] [Google Scholar]
- 17.Arandia, N. I. et al. Anti-Polymyositis/Scl Antibodies in Systemic Sclerosis: Clinical associations in a multicentric spanish cohort and review of the literature. J. Clin. Rheumatol.28(1), e180–e188. 10.1097/RHU.0000000000001676 (2022). [DOI] [PubMed] [Google Scholar]
- 18.Moll, S. A. et al. Serum biomarker CA 15–3 as predictor of response to antifibrotic treatment and survival in idiopathic pulmonary fibrosis. Biomark. Med.14(11), 997–1007. 10.2217/bmm-2020-0165 (2020). [DOI] [PubMed] [Google Scholar]
- 19.Jehn, L. B. et al. Serum KL-6 as a biomarker of progression at any time in fibrotic interstitial lung disease. J. Clin. Med.12(3), 1173. 10.3390/jcm12031173 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Tseng, C. et al. Comparative accuracy of CA 15–3 and KL-6 as diagnostic and prognostic biomarkers for interstitial lung disease. Clin. Chim. Acta.565, 119980. 10.1016/j.cca.2024.119980 (2025). [DOI] [PubMed] [Google Scholar]
- 21.Barsotti, S. et al. One year in review 2017: Systemic sclerosis. Clin. Exp. Rheumatol.35(106), 88–95 (2017) (Epub 2017 Sep 29). [PubMed] [Google Scholar]
- 22.Ma, A. et al. Serum carbohydrate antigen 153 as a predictor of interstitial lung disease associated with rheumatoid arthritis is positively correlated with serum Krebs von den Lungen-6. BMC Pulm. Med.25(1), 102. 10.1186/s12890-025-03558-4 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Oguz, E. et al. Association of serum KL-6 levels with interstitial lung disease in patients with connective tissue disease: A cross-sectional study. Clin. Rheumatol.35(3), 663–666. 10.1007/s10067-015-3167-8 (2016) (Epub 2016 Jan 13). [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
All data generated or analysed during this study are included in this published article. Additional de-identified individual data or analytical scripts are available from the corresponding author upon reasonable request.