Impact of sample processing delays on plasma markers of inflammation, chemotaxis, cell death, and blood coagulation

Abstract

Background: Biosampling studies in critically ill patients traditionally involve bedside collection of samples followed by local processing (ie. centrifugation, aliquotting, and freezing) and storage. However, community hospitals, which care for the majority of Canadian patients, often lack the infrastructure for local processing and storage of specimens. A potential solution is a "simplified" biosampling protocol whereby blood samples are collected at the bedside and then shipped to a central site for processing and storage. One potential limitation of this approach is that delayed processing may alter sample characteristics.

Objective: To determine whether delays in blood sample processing affect the stability of cytokines (IL-6, TNF, IL-10, IFN-γ), chemokines (IL-8, IP-10, MCP-1, MCP-4, MIP-1α, MIP-1β), cell-free DNA (cfDNA) (released by dying cells), and blood clotting potential in human blood samples.

Methods: Venous blood was collected into EDTA and citrate sample tubes and stored at room temperature (RT) or 4°C for progressive intervals up to 72 hours, prior to processing. Plasma cytokines and chemokines were quantified using single or multiplex immunoassays. cfDNA was measured using Picogreen DNA Quantification. Blood clotting potential was measured using a thrombin generation assay.

Results: Blood samples were collected from 9 intensive care unit (ICU) patients and 7 healthy volunteers. Admission diagnoses for the ICU patients included sepsis, trauma, ruptured abdominal aortic aneurysm, intracranial hemorrhage, gastrointestinal bleed, and hyperkalemia. After pre-processing delays of up to 72 hours at RT or 4°C, no significant changes were observed in plasma cytokines, chemokines, cfDNA, or thrombin formation.

Conclusions: Delayed sample processing for up to 72 hours at either RT or 4°C did not significantly affect cytokines, chemokines, cfDNA, or blood clotting potential in plasma samples from healthy volunteers and ICU patients. A "simplified" biosampling protocol is a feasible solution for conducting biosampling research at hospitals without local processing capacity.

Fig 1. Multiplex assay of cytokine levels in EDTA plasma samples from ICU patients.

Whole blood was collected from ICU patients (n = 9) into EDTA tubes and stored at room temperature (RT) for 0, 24, 48, or 72 hours prior to centrifugation. Plasma levels of IL-6, IL-10, IFN-γ, and TNF were measured using multiplex analysis.

Fig 2. Multiplex assay of chemokines levels in EDTA plasma samples from ICU patients.

Blood was collected from ICU patients (n = 9) into EDTA tubes and stored at room temperature (RT) for 0, 24, 48, or 72 hours prior to centrifugation. No significant changes were observed in levels of MCP-1, MCP-4, MIP-1α, MIP-1β, MCP-4, IP-10, or IL-8 with delayed processing at RT.

Fig 3. cfDNA levels in citrate and EDTA plasma samples from healthy volunteers and ICU patients.

Blood was collected from healthy volunteers (n = 7) and ICU patients (n = 9) into citrate or EDTA tubes and stored at room temperature (RT) for 0, 24, 48, or 72 hours prior to centrifugation. cfDNA levels were measured using a Picogreen assay.

Fig 4. Thrombin generation parameters (lag time and AUC) in citrated plasma samples from healthy volunteers and ICU patients.

Blood was collected from healthy volunteers (n = 7) and ICU patients (n = 9) into citrate tubes and stored at room temperature or at 4°C for 0, 24, 48, or 72 hours, prior to centrifugation. Blood clotting potential was measured using a thrombin generation assay. The lag time is expressed as minutes. The units of AUC are “nM thrombin x min”.