Abstract 14403: Troponin as a Powerful Predictor of Mortality in Critically Ill Covid-19 Patientsmulticenter Analysis in 159 Patients

Introduction: Certain cardiac biomarkers have been shown to predict survival in patients infected with COVID-19. However. The use of Troponins, CPK and LDH in predicting the outcome in patients who are critically ill and require advanced respiratory support is less clear. Methods: We performed a multicenter analysis of 159 consecutive patients with confirmed COVID-19 who were admitted to Intensive Care Unit (ICU) between March 01, 2020 and April 30, 2020. Patients were then followed until May 23, 2020. Demographic data (age, sex, race, BMI) were recorded. Cardiac biomarkers (CPK, Troponins, BNP, LDH) were analyzed. Patient status was classified either alive or deceased at the end of follow up period. Results: Mean patient age was 66+/-15 and 53% were male. Mean BMI was 31+/- 9. Mean hospital ICU stay was 11+/-8 days. Mortality rate of this ICU cohort at the end of follow-up was 63%. Fifty-five (34%) patients were discharged from the hospital. A multivariate logistic regression analysis identified 2 factors (Age >62 HR 2.4, 95% CI 1.21-4.676, p<0.01 and elevated Troponin >0.5 HR 3.45, 95% CI 1.01-11.8, p 0.048) has significant and independent contributions to the likelihood of survival (Fig 1 and Fig 2). Fig 1 KM curve of different strata of Troponin levels Fig 2 KM curve of 2 strata of Age Conclusion: Elevated Troponins level are common in critically ill COVID-19 patients. The highly predictive value of Troponins in survival may indicate possible cardiac involvement of COVID-19 infection as a determinant of mortality. Early and serial determination of Troponin values can provide risk stratification for these patients and timely aggressive intervention may decrease mortality.

An Empirical Biomarker-based Calculator for Autosomal Recessive Polycystic Kidney Disease The Nieto-Narayan Formula

Autosomal polycystic kidney disease (ARPKD) is associated with progressive enlargement of the kidneys fuelled by the formation and expansion of fluid-filled cysts. The disease is congenital and children that do not succumb to it during the neonatal period will, by age 10 years, more often than not, require nephrectomy+renal replacement therapy for management of both pain and renal insufficiency. Since increasing cystic index (CI; percent of kidney occupied by cysts) drives both renal expansion and organ dysfunction, management of these patients, including decisions such as elective nephrectomy and prioritization on the transplant waitlist, could clearly benefit from serial determination of CI. So also, clinical trials in ARPKD evaluating efficacy of novel drug candidates could benefit from serial determination of CI. Although ultrasound is currently the imaging modality of choice for diagnosis of ARPKD, its utilization for assessing disease progression is highly limited. Magnetic resonance imaging or computed tomography, although more reliable for determination of CI, are expensive, time-consuming and somewhat impractical in the pediatric population. Using a well-established mammalian model of ARPKD, we undertook a big data-like analysis of minimally- or non-invasive serum and urine biomarkers of renal injury/dysfunction to derive a family of equations for estimating CI. We then applied a signal averaging protocol to distil these equations to a single empirical formula for calculation of CI. Such a formula will eventually find use in identifying and monitoring patients at high risk for progressing to end-stage renal disease and aid in the conduct of clinical trials.