GAP (gender-age-physiology) and TORVAN are multi-parametric prognostication scores for idiopathic pulmonary fibrosis (IPF). We compared their prognostic value in patients treated with nintedanib or pirfenidone and explored their effect on patient survival in relation to disease staging.

Retrospective evaluation of 235 naïve IPF patients (M = 179; mean age 69.8 yrs±7.1; 102 treated with nintedanib and 133 with pirfenidone), referred to two Italian academic centers between February 2012 and December 2019.

During a median follow-up of 4.2 years, the incidence rate of death was 14.5 per 100 person-years (95% CI: 12 to 17.4), with no differences between nintedanib and pirfenidone (log-rank p = 0.771). According to time-ROC analysis, GAP and TORVAN showed a similar discrimination performance at 1, 2, and 5 years. Survival of GAP-2/GAP-3 IPF patients treated with nintedanib was worse than that of patients in GAP-1 (HR 4.8, 95% CI: 2.2 to 10.5 and HR 9.4, 95% CI: 3.8 to 23.2). TORVAN I patients treated with nintedanib exhibited better survival than those in stages III (HR 3.1, 95% CI: 1.4 to 6.6) and IV (HR 10.5, 95% CI: 3.5 to 31.6). A significant treatment x stage interaction was observed for both disease staging indexes (p = 0.042 for treatment by GAP interaction and p = 0.046 for treatment by TORVAN interaction). A better survival was associated with nintedanib in patients with mild disease (GAP-1 or TORVAN I stage) and with pirfenidone in GAP-3 or TORVAN IV cases, although these findings did not always reach statistical significance.

GAP and TORVAN similarly perform in IPF patients on anti-fibrotic therapy. However, the survival of patients treated with nintedanib and pirfenidone appears to be differently affected by disease staging.