Evaluation of usefulness in surfactant protein D as a predictor of mortality in myositis-associated interstitial lung disease

Objective Surfactant protein D (SP-D) is considered a serum biomarker of various forms of interstitial lung disease (ILD). In this study, we examined the utility of SP-D as a predictive biomarker for mortality in patients with ILD associated with polymyositis/dermatomyositis (PM/DM) using large-scale multicentre cohort data. Methods We enrolled 381 patients with incident PM/DM-associated ILD in a multicentre retrospective cohort based on the availability of serum SP-D at the baseline. Demographic and clinical characteristics as well as the presence of autoantibodies to melanoma differentiation-associated gene 5 (MDA5) and aminoacyl tRNA synthetase were measured at the time of diagnosis, and follow-up survival data were collected prospectively. Results Seventy-eight patients died during the median observation period of 18 months, and the majority of patients died of ILD. The SP-D levels at baseline were significantly lower (P = 0.02) in a non-survivor subset than in a survivor subset among the entire enrolled patients. However, the SP-D levels were higher in the non-survivor subset than in the survivor subset based on the stratification by anti-MDA5-positive, anti-ARS-positive and, double-negativity, although there was an only statistically significant difference (P = 0.01) in the double-negative group. Surprisingly, the SP-D levels were within the upper limit of normal, 110 ng/mL, in 54 (87%) of 62 anti-MDA5-positive patients who died. In the double-negative group, the mortality rates were significantly higher (P = 0.002) in a subset with SP-D ≥127.6 ng/mL, the cut-off value for mortality calculated by the receiver operating characteristic curve, than the other subset. All of patients with SP-D <127.6 ng/mL survived. Conclusion Serum SP-D levels behave differently among patients with stratified by anti-MDA5 antibody, anti-ARS antibody and both negativity in PM/DM-associated ILD. Its use in clinical practice should be applied with caution on the basis of the presence or absence of anti-MDA5 antibody or anti-ARS antibody.