Assessment of Microcirculatory Dysfunction in Septic Shock Patients by OCTA

Purpose and rationale: Sepsis is a life-threatening organ dysfunction caused by a dysregulated host response to infection. Sepsis and septic shock are major public health problems killing one in every three patients. Microcirculatory dysfunction is frequent in septic shock. The duration and severity of this dysfunction have a prognostic impact by being associated with organ failure and mortality. Our study purposes to demonstrate the feasibility of OCTA to improve assessment of microcirculatory dysfunction by showing that retinal and choroidal microcirculatory changes with prognostic impact are present during septic shock.

Primary objective (specific aim 1): To characterize the alterations of retinal and choroidal microcirculation in septic shock.

We will test the hypothesis that retinal and/or choroidal microcirculation shows dysfunctional changes (lower vascular density, lower percentage of perfused small vessel, lower blood flow index and higher vascular heterogeneity) in septic shock patients.

Secondary objective (specific aim 2): To test the prognostic value of retinal and choroidal microcirculatory dysfunction in septic shock.

We will test the hypothesis that higher magnitude and persistence of retinal and/or choroidal microcirculatory dysfunction beyond the successful macro-hemodynamic resuscitation are independent predictors of organ failure and mortality in septic shock patients.

Study type: Two sequential observational studies.

Study design: A cross-sectional case-control study followed by a prospective cohort study with a 90-days longitudinal follow-up period.

Study population: 165 septic shock patients and 30 healthy controls. Power and sample size calculations: Based on sublingual percentage of perfused small vessel (PPV) difference previously reported between septic shock patients and healthy controls (60% vs. 95%), we estimate that we will need to enrol 27 patients and 27 controls to demonstrate the same difference at the retina and choroid. To show a PPV difference between survivors and non-survivors (70% vs. 46%) in the septic shock group, also based on previous reports at the sublingual microcirculatory level, we estimate that we will need to enrol 150 septic shock patients. These calculations assumed an alpha level of 0.05 and a power of 80%. Our Department admits about 200 septic shock patients annually, so to account for drop-outs and limitations with enrolment (10-20% refusal to participate) we decided to increase our enrolment goal by 10% to 165 septic shock patients and 30 healthy controls.

Recruitment and inform consent: Recruitment will take place at the Intensive Care Medicine Department of Hospital da Luz Lisboa. Patients admitted to the Department with the diagnosis of septic shock will be screened for eligibility. If the patients are eligible to participate, they will be invited to enrol in the study by the principal investigator or other ICU medical team member. The study objectives and procedures will be explained to them. If the patient agrees to participate in the study, a written informed consent will be sign before enrolment. If the patient is unable to give informed consent to participate in the study, it will be asked to the reference next of kin using the same procedures as described before.

Study duration: 90 days from enrolment to final follow-up assessment. One to two years of enrolment.

Study procedures:

Baseline Assessment: After confirmation of eligibility and enrolment in the studies, demographic data (age, gender), type of admission (urgent/elective, medical/surgical/trauma), patient origin (emergency department, ward, operation room), source of infection (lung, urinary tract, intra-abdominal, skin and soft tissues, others), prognostic scores (APACHE II, SAPS II) and organic dysfunction scores (SOFA, quick SOFA) will be collected at baseline for septic shock patients.

Study Assessments:

OCTA: OCTA examination will be performed daily to septic shock patients from day 1 (less than 24 hours after diagnosis) until successful shock resolution (weaning from vasopressors) or until a maximum of 7 days. The healthy controls will be submitted to a single OCTA evaluation. We will use the Cirrus 6000® OCTA system (ZEISS, Germany) and the interpretation of OCTA images will be performed by two independent specialized ophthalmologists blinded to the clinical condition of patients. Images will be stored at the device working station.

Hemodynamic assessment: Every septic shock patient will have a central venous catheter and an arterial line in place according to the standard of care. Daily, during or within a maximum of 1 hour of OCTA examination, we will record the values of temperature, heart rate, mean arterial pressure, capillary refill time, central venous pressure, cardiac index, pH, PaCO2, PaO2, PvCO2, SaO2, SvO2, haemoglobin, serum lactate, fluid balance, oxygen delivery, oxygen consumption and oxygen extraction ratio.

Vasoactive and sedo-analgesic drugs: Daily we will record the vasoactive and sedo-analgesic drugs administered to septic shock patients and its total daily dose until complete weaning (defined as 24 hours free of vasopressors). Daily vasopressor score will also be calculated as suggested by Póvoa, P. et al.

Ventilatory and renal replacement therapy support: Daily we will record the type of ventilatory support (oxygen therapy, high flow oxygen therapy, non-invasive ventilation, invasive ventilation) and of renal replacement therapy (continuous, sustained low-efficiency, intermittent) used, if needed, until complete weaning.

Data analysis: In both studies, we will perform a descriptive statistical analysis of studied variables. Comparisons between healthy controls (single OCTA) and septic shock patients (first OCTA) to address endpoints related to specific aim 1 will be performed using t-Student test or Mann-Whitney test as appropriate. To address endpoints related to specific aim 2, we will compare survivors to non-survivors by doing a survival analysis using Kaplan-Meyer curves and Cox proportional hazards multivariable regression with microcirculatory measures as predictors. Additional comparisons will be performed using χ2 test, t-Student test, Mann-Whitney test and logistic regression as appropriate. We will assess confounding by adding significant variables in the regression models. We intend to perform a prespecified subgroup analysis of septic shock patients by serum lactate <4mmol/L vs. ≥4mmol/L and <2mmol/L vs. ≥2mmol/L. The correlation between retinal and choroidal microcirculation variables and macrohemodynamic variables will be assessed by the Spearman rank correlation coefficient. Area under Receiver Operating Characteristic curve of OCTA for prognostic outcomes will be calculated. A p-value < 0.05 will be considered statistically significant. Statistical analysis will be performed with STATA® 15 (StataCorp, Texas, USA).