Background Pulmonary alterations are the major cause of mortality in SSc. The most pts develop some degree of lung injury over the course of the disease.
Objectives to assess predictors of poor prognosis in a single center cohort of SSc pts with interstitial lung disease (ILD) and to build a predictive model of prognosis disease SSc-ILD.
Methods 77 pts with SSc-ILD were involved in a longitudinal study over a five-year period. We used discriminant analysis and our selection involved missing value. So discriminant analysis was held at 58 pts (4 were men). The mean age was 47±13 years. Pts. were divided into 3 subgroups based on HRCT changes estimated at the entry of the study and at last visit. The mean time between two evaluations was 57±10 months. The 1st group included 13pts with improvement (n=13), 2nd –pts without any changes (n=30) and 3rd – pts with deterioration negative dynamics (n=15). ROC analysis was used for determination of prognostic intervals.
Results We got the discrimination function that included: ground glass opacities (GGO) (1 – the absents of GGO; 2 – the presents of GGO); index EScSG (the digital value); FVC (FVC ≥75 = 0, FVC <75 = 1); the maximum daily dose of glucocorticoids (the digital value, mg); Gamma globulins (the digital value, %); cyclophosphamide (CYC) (the absence of CYC =0, the treatment of CYC =1); DLco (DLco <52% =0, DLco ≥52 =1). We assessed all parameters at the entry in the study except DLco evaluated after one year of follow up. Based on this analysis the equation was developed.
The equation of prognosis is 3.14 GGO + 0,70 index EScSG – 1,326 FVC - 0,1 the maximum daily dose of glucocorticoids + 0,136 Gamma globulins + 1.066 CYC – 1.075 DLCO ≤ 5,8.
Value of equation of prognosis to 5.8 corresponds to stabilization or good prognosis and value after 5.8 corresponds to poor prognosis. The equation of prognosis had 86% sensitivity and 56% specificity. The cut-off point 4.7 was represented 65% sensitivity and 67% specificity. The prognosis intervals were following: value of equetion to 4.7 corresponds to good prognosis; value from 4.8 to 5.8 corresponds to stabilization and vaule after 5.8 corresponds to poor prognosis.
The mean value of discrimination function for each group is shown in table.
Conclusions using discrimination analysis we got the prognosis intervals for the pts with SSc. It can give an opportunity to determine the prognosis of SSc-ILD.
Disclosure of Interest None declared