Activin/Follistatin-axis deregulation is independently associated with COVID-19 in-hospital mortality

Synolaki, E.; Papadopoulos, V.; Divolis, G.; Gavriilidis, E.; Loli, G.; Gavriil, A.; Tsigalou, C.; Tsachouridou, O.; Sertaridou, E.; Rafailidis, P.; Pasternack, A.; Boumpas, D.; Germanidis, G.; Ritvos, O.; Metalidis, S.; Skendros, P.; Sideras, P.

Rationale: Activins are inflammatory and tissue-repair-related members of the TGF{beta}-superfamily that have been implicated in the pathogenesis of several immuno-inflammatory disorders including sepsis/acute respiratory distress syndrome (ARDS). We hypothesized that they might be of particular relevance to COVID-19 pathophysiology. Objectives: To assess the involvement of the Activin-Follistatin-axis in COVID-19 pathophysiology. Methods: Levels of Activins -A, -B and their physiological inhibitor Follistatin, were retrospectively analyzed in 314 serum samples from 117 COVID-19 patients derived from two independent centers and compared with common demographic, clinical and laboratory parameters. Optimal-scaling with ridge-regression was used to screen variables and establish a prediction model. Main Results: The Activin/Follistatin-axis was significantly deregulated during the course of COVID-19 and was independently associated with severity and in-hospital mortality. FACT-CLINYCoD, a novel disease scoring system, adding one point for each of Follistatin >6235 pg/ml, Activin-A >591 pg/ml, Activin-B >249 pg/ml, CRP >10.3 mg/dL, LDH >427 U/L, Intensive Care Unit (ICU) admission, Neutrophil/Lymphocyte-Ratio >5.6, Years of Age >61, Comorbidities >1 and D-dimers >1097 ng/ml, efficiently predicted and monitored fatal outcome independently of multiplicity and timing of sampling (AUC: 0.951{+/-}0.032, p<10-6). Validation in 35 samples derived from a third hospital indicated comparable AUC (0.958{+/-}0.086, p=0.032). Conclusion: This study unravels the link between Activin/Folistatin-axis and COVID-19 mortality and introduces FACT-CLINYCoD, a novel pathophysiology-based tool that copes with the dynamic and heterogeneous nature of COCVID-19, predicts disease outcome and supports clinical decision making. Prospective large-scale validation of this calculator, as well as investigation of the mechanisms linking Activin/Folistatin-axis to COVID-19 pathogenesis is warranted.