Endocrine Abstracts

Introduction: To date, there are no clear biochemical parameters to identify early COVID-19 cases at risk of complications in the Intensive Care Unit (ICU).

Aim: To evaluate the prognostic potential of endocrine biomarkers associated with acute inflammatory conditions in ICU patients for COVID-19.

Methods: Between 1/3/2020 and 31/12/2020 we recruited 126 consecutive patients at the admission to ICU 1U, Molinette University Hospital of Turin. Within 24 hours (T1), we calculated: SAPS II (Simplified Acute Physiological Score II), SOFA (Sequential Assessment of Organ Failure) and MuLBSTA (Multinodular infiltration, hypoLymphocytosis, Bacterial co-infection, Smoking history, hyperTension and Age). At T1, at 72 hours (T3) and after 7 days (T7), we measured the plasma levels of copeptin and MR-proADM. Duration of extracorporeal membrane oxygenation (ECMO) and mechanical ventilation (MV), ICU and hospital length of stay (LOS), ICU (IM) and hospital mortality (HM) were recorded. We present the results of the first 69 patients.

Results: 53 males and 16 females (median age 63) were enrolled. The ICU and hospital LOS were 13 [6.7-22.2] days and 20 [13-30] days, respectively. The median time free from MV was 1 day. The need for pronation was significantly predicted by a higher MuLBSTA classification (OR 4.01, P=0.025). The median duration for ECMO was 17 days. The SOFA score was higher in patients requiring ECMO (P=0.019) and performed better when corrected for copeptin at T1 (HR 1.25, P=0.02). IM and HM were 62.3% and 68.1% (median survival time 18 and 24 days), respectively. MR-proADM was higher in dying patients with increasing statistical significance over time (T1 1.3 vs 0.91 nmol/l, P=0.028; T7 1.77 vs 0.9 nmol/l; P=0.0001). The MR-proADM T7/T1 ratio was significantly higher in those who died during hospitalization (P=0.0003) predicting HM (HR 1.35, P=0.025). SAPS II class at the admission also predicted HM (HR 1.39, P=0.013), but correcting for MRproADM (HR 1.43, P=0.012) or copeptin measured at T1 improved its statistical significance (HR 1.56, P=0.003). Likewise, SOFA score, easier to collect in ICU, significantly predicted HM (HR 1.09, P=0.041), but improved its performance if corrected for MR-proADM (HR 1.12, P=0.019) or copeptin at T1 (HR 1.11, P=0.018). Neither gender, estimated glomerular filtration rate, C-reactive protein, or procalcitonin affected the previous regression analyses.

Conclusions: Endocrine biomarkers evaluated at ICU admission improve the ability of prognostic scores to predict mortality and severe adverse outcomes. The increase in MR-proADM after 7 days of hospitalization in the ICU also predicts hospital mortality.