Microvascular Fluid Exchange: Implications of the Revised Starling Model for Resuscitation of Dengue Shock Syndrome

Abstract

Dengue is the most common mosquito-borne viral infection in the world. The most feared complication is a poorly understood vasculopathy that occurs in only a small minority of symptomatic individuals, especially children and young adults, but can result in potentially fatal dengue shock syndrome (DSS). Based mainly on expert opinion, WHO management guidelines for DSS recommend prompt infusion of a crystalloid fluid bolus followed by a tapering crystalloid fluid regimen, supplemented if necessary by boluses of synthetic colloid solutions. However, following publication of a number of major trials undertaken in other, primarily adult, critical care scenarios, use of both synthetic colloid solutions and of fluid boluses for volume expansion have become controversial. Synthetic colloids tend to be used for severe DSS cases in order to boost intravascular oncotic pressure, based on the classic Starling hypothesis in which opposing hydrostatic and oncotic forces determine fluid flow across the microvascular barrier. However, the revised Starling model emphasizes the critical contribution of the endothelial glycocalyx layer (EGL), indicating that it is the effective oncotic pressure gradient across the EGL not endothelial cells per se that opposes filtration. Based on several novel concepts that are integral to the revised Starling model, we review the clinical features of DSS and discuss a number of implications that are relevant for fluid management. We also highlight the need for context-specific clinical trials that address crucially important questions around the management of DSS.

Keywords: colloid; crystalloid; dengue shock syndrome; endothelial glycocalyx layer; fluid management; revised Starling model; tight junction.

Figure 1

Cartoon depicting factors influencing fluid flow through the endothelial barrier [adapted from (37)]. EGL, endothelial glycocalyx layer. Gray dots represent protein molecules. Fluid flow (Jv) is proportional to the hydrostatic gradient (i.e., the difference between the capillary {P_c} and interstitial {P_i} hydrostatic pressures) minus the product of the reflection coefficient {σ} and the oncotic pressure gradient [i.e., the difference between the capillary (π_c) and interstitial or sub-glycocalyx {π_i or π_g} oncotic pressures]. The reflection coefficient indicates the ease with which molecules either penetrate the EGL or are deflected back into free-flowing plasma, and varies between 0 and 1. In health σ is close to 1 for albumin and most other plasma proteins. In dengue σ is likely to be reduced, reflecting changes to the EGL structure. The EGL forms the primary barrier to movement of fluid and plasma contents from the intravascular to the interstitial spaces. Fluid that enters the sub-glycocalyx space flows toward the inter-endothelial clefts and is channeled at high velocity to the interstitium, due to the physical characteristics of the cleft architecture and the narrow width of the gaps in the structures that bind the cells together. This high velocity fluid movement prevents back diffusion of interstitial proteins toward the protected sub-glycocalyx space, which remains effectively protein free. Fluid and proteins that get to the interstitial space are reabsorbed back into the intravascular space via the lymphatic system. In line with the revised Starling model the oncotic pressure gradient from the protein free sub-glycocalyx space to plasma (broad green arrow) is greater than expected with classical Starling model (narrow green arrow), and fluid losses to the interstitium are much lower than previously thought.

Figure 2

Cartoon depicting the expected changes associated with profound shock. In profound shock the hydrostatic pressure gradient from the vascular lumen to the interstitial space is abolished (P_c = P_i). Fluid and proteins are able to diffuse back through the inter-endothelial cleft to the sub-glycocalyx space, allowing reabsorption of about 500 ml (autotransfusion) to occur. The capillary pressure rises and a new hydrostatic pressure gradient is established within about 30 min. Fluid filtration restarts but at a low level. However, the oncotic pressure gradient opposing filtration is also reduced since some protein is now present in the sub-glycocalyx protected space. Following successful resuscitation, hydrostatic pressure improves, intercellular cleft fluid flow increases, and protein is washed out from the sub-glycocalyx space until both P_c and π_g are restored to normal.