Blood Test Flags ILD in Systemic Sclerosis

Serum levels of three markers of pulmonary fibrosis pathology were significantly associated with the presence of interstitial lung disease (ILD) in patients with systemic sclerosis (SSc) — strongly enough to serve as a clinical diagnostic-prognostic test, researchers said.

When combined into a four-stage categorical index, SSc with ILD patients were nearly 13 times more likely to have the highest value relative to patients without ILD (OR 12.72, 95% CI 4.59-35.21), according to Adelle Jee, PhD, of Royal Prince Alfred Hospital and the University of Sydney in Australia, and colleagues.

When the control group also encompassed healthy controls and the case group expanded to include non-SSc patients with idiopathic pulmonary fibrosis (IPF), the index was even more strongly associated with ILD, with an odds ratio of 15.61 (95% CI 5.81-41.61), the group reported in Arthritis & Rheumatology.

The three biomarkers were surfactant protein D, the cancer antigen CA15-3, and intercellular adhesion molecule-1. When examined individually for their association with ILD, Jee and colleagues found each association significant, with areas under their respective receiver operating characteristic curves (AUC) of 0.68-0.76.

A fourth marker, matrix metalloproteinase-7, had AUC values in the same range, but it didn’t contribute significantly to the overall index’s accuracy when factors such as age, sex, and smoking status were considered. It thus was dropped from the final model.

Currently, the standard for assessing ILD in SSc patients is periodic high-resolution CT scanning along with pulmonary function tests. No major questions have been raised about the effectiveness of such testing, but CT scans are not benign. “The cumulative radiation exposure from repeated [CT] screening poses an unacceptable risk in a predominantly young to middle-aged, female population,” Jee and colleagues explained. Accuracy of other imaging methods such as ultrasound has not been established.

Pulmonary function tests, meanwhile, only return positive results when impairment is already advanced and may reflect impacts from conditions other than ILD.

After reviewing the literature, Jee and colleagues identified 28 serum components showing some association with ILD and checked a variety of combinations to gauge their ability to discriminate ILD in different patient groups.

For developing the models, the researchers drew on data from 111 SSc patients with ILD, 77 SSc patients with no lung pathology, 74 non-SSc patients with IPF, and 30 healthy controls. The models were then tested in a validation cohort comprising 142 SSc-ILD patients, 102 SSC patients without ILD, and 98 IPF controls. Some 5 years of longitudinal data were available for these groups. For each model’s results, categorical index values of 0-3 were set.

More than 70% of the healthy controls had index values of 0 using the final three-marker panel and none had values of 3. In SSc patients without ILD, about one-quarter had index values of 0, about half had values of 1, and 20% had values of 2. Two-thirds of SSc-ILD patients scored 2 or 3, as did more than 80% of IPF controls.

Besides flagging ILD in patients with definite lung disease, the index was also associated with disease severity in the SSc group. Patients scoring in category 3 showed forced vital capacity values nearly 18 percentage points lower than patients in category 0; diffusion capacity for carbon monoxide was lower by 20 percentage points comparing the two groups.

Jee and colleagues noted that there clearly remains room for improvement, but the current results demonstrate that their approach is workable. They suggested that “further investigation of composite biomarker indices as objective, minimally-invasive, point-of-care tests to complement current diagnostic modalities” is warranted.

Limitations to the study included differences in how basic data were collected in the various patient groups and that many of those individuals were undergoing treatments that may have affected results.