The assessment of left ventricular function in sepsis and septic shock; comparison of ejection fraction measurement, global longitudinal strain, troponin T and N Terminal-ProBNP

Abstract Background Sepsis and septic shock, and the sequelae of septic cardiomyopathy (SCM), are a common yet major concern in the intensive care unit and are associated with high mortality (Zakynthinos, 2025). Traditionally, echocardiography using two-dimensional ejection fraction (2DEF) is used to diagnose SCM. However, these measurements may be affected by other confounding factors (Berman, 2022). Different methods of detecting cardiac dysfunction such as global longitudinal strain (GLS) or biomarkers (NT-ProBNP or Troponin T) are not routinely used. Furthermore, 2DEF measurements are subject to poor sensitivity and specificity, with documented inter-operator errors (Kim, 2022). Purpose To evaluate relationship between LVEF, GLS and biomarkers across 90 days in patients with sepsis/septic shock on ICU. Method A single UK site, prospective cohort study of patients on ICU meeting criteria for sepsis/septic shock. Simpson’s biplane 2DEF and GLS performed by single operator. Biomarker samples taken within 24hrs of echo. Data collection took place at days 1, 3 (no biomarker), 30 and 90 as inpatient or returning outpatient. Normal values were defined by British Society Echocardiography and National Institute Clinical Excellence guidelines respectively. Statistical significance was determined using Pearson Correlation and Principal Component Analysis (PCA). Results From 118 patients enrolled, 105 met inclusion criteria at time of echo, all were on inotropes or vasopressors. Mean age 60yrs, 61% male, 38% female. Days 1, 3, 30 and 90 results shown in table 1. Loss to follow up over the study was 33%. At day 1 there was no significant correlation between 2DEF and any biomarker (NT-ProBNP r =0.15,p =0.55, Troponin T, r =0.25 ,p=0.27). Nor any significant relationship between GLS and Troponin T (r = 0.20, p =0.390). Conversely there was a borderline relationship between GLS and NT-pro-BNP, (r =0.49, p=0.05) at day 1. Over 90 days,despite a normal 2DEF in many cases, GLS and biomarkers were frequently abnormal and persisted beyond recovery (see figure 1). PCA mimic these findings with no 2DEF and GLS relationship but some degree of inter-relationship between GLS and NT ProBNP. Conclusion Generally, 2DEF is the chosen method for the diagnosis of SCM. Given that many patients have normal 2DEF despite abnormal GLS and biomarkers, this study raises the question of ‘what is a normal LVEF’ for ICU patients with sepsis. Perhaps given the fluid loading, sepsis-driven vasodilatation, and inotropic support, an otherwise normal 2DEF of 50-60% may be impaired in this cohort. As such, GLS and biomarkers may identifying subtle (potentially subclinical) dysfunction in sepsis or septic shock. These changes persist to day 90, despite clinical recovery. While 2DEF is the primary factor influencing decision-making in the ICU, the relationship between GLS and NT ProBNP post day 3 requires further investigation with regards to genuine persisting myocardial dysfunctionTable 1 Figure 1