Transcapillary IgG Exchange and Vascular Reflection

Why Capillary Exchange Has Two Components

IgG moves between plasma and tissue across capillary barriers by more than one mechanism. A concentration difference can drive passive diffusion, while water moving across the barrier can carry some IgG with it. The relative importance of these processes depends on the tissue barrier: a highly permeable capillary can support substantial exchange, whereas a restrictive barrier can limit both diffusion and solvent drag.

This tutorial represents those processes as a net transcapillary flux. The sign convention is essential: a positive flux means movement from vascular plasma into the tissue interstitium; a negative flux means movement from interstitium back to plasma. The flux is an amount per unit time. It can be placed with opposite signs in the vascular and tissue amount balances, without specifying what happens to tissue IgG after it enters the interstitium.

A Bidirectional Flux Model

Let \(C_v\) and \(C_i\) be IgG concentrations in vascular plasma and tissue interstitium, respectively. Let \(J_v\) be the net volume flow of water across the capillary, with positive values indicating vascular-to-interstitial flow. The diffusive component is proportional to the concentration difference: when \(C_v>C_i\), diffusion is outward; when \(C_i>C_v\), it is inward.

Convection carries IgG along with water, but the barrier can restrict that carriage. To calculate the convective component when water direction can reverse, use the concentration on the side from which the water originates. Define \(J_v^+=\max(J_v,0)\) and \(J_v^-=\max(-J_v,0)\), so both quantities are nonnegative. Then \(J_v^+\) is outward water flow and \(J_v^-\) is inward water flow.

The first term is net diffusion. The second is net solvent drag, with the reflection coefficient reducing the fraction of IgG carried by water. Concentration multiplied by volume per time gives amount per time in both terms. Thus \(PS(C_v-C_i)\) and \(J_v C\) have the same flux units.

When water flows outward, \(J_v>0\), the convective term is \((1-\sigma_{\mathrm{IgG}})J_vC_v\). When water flows inward, \(J_v<0\), it is \(-(1-\sigma_{\mathrm{IgG}})(-J_v)C_i\). Using the concentration at the water’s source prevents inward water from incorrectly carrying vascular-side IgG.

How Water Flow Enters the Model

Hydrostatic pressure can drive water across a capillary wall, but the net water flow may also depend on osmotic forces. A simplified Starling-type relation is

Here \(L_pS\) is hydraulic conductance, \(P_v\) and \(P_i\) are vascular and interstitial hydrostatic pressures, and \(\pi_{v,k}\) and \(\pi_{i,k}\) are osmotic pressures associated with solute or protein class \(k\). The corresponding \(\sigma_k\) describes reflection of that class for the osmotic term. This relation is a simplified description: actual tissue models may represent the relevant osmotic contributions and hydraulic properties in more detail.

Do not infer the direction of water flow from hydrostatic pressure alone if an osmotic term is included. The sign of the net \(J_v\) determines the direction used in the IgG convection term. Also, the reflection coefficient used for IgG convection need not equal a reflection coefficient used for another protein or solute’s osmotic effect.

Worked Flux Calculations

Why the Flux Formula Has This Form

The expression follows by adding two fluxes with the same sign convention. For diffusion, the net flux is the outward diffusive contribution \(PS C_v\) minus the inward contribution \(PS C_i\), giving \(PS(C_v-C_i)\). For convection, outward water moves at rate \(J_v^+\) and carries source-side concentration \(C_v\); inward water moves at rate \(J_v^-\) and carries source-side concentration \(C_i\). Multiplying each by the transmitted fraction \(1-\sigma_{\mathrm{IgG}}\) gives the net convective term.

Adding the diffusive and convective terms yields the stated model. This construction also checks its limiting cases: if \(J_v=0\), convection is zero; if \(C_v=C_i\), diffusion is zero; and if \(\sigma_{\mathrm{IgG}}=1\), convection is zero regardless of water direction. These checks are useful when implementing the equation.

Model Choices, Diagnostics, and Limitations

A common alternative is to represent exchange by a single first-order term, such as \(K(C_v-C_i)\). That form can describe net concentration-driven exchange, and it may be a useful approximation when convection is negligible or absorbed into an empirically estimated coefficient. It does not separately identify diffusion and solvent drag, however, and it cannot directly represent convection changing direction while the concentration gradient stays fixed. A perfusion-limited approximation instead treats capillary exchange as sufficiently rapid that barrier resistance is negligible; it is unsuitable when a restrictive barrier or finite \(PS\) materially limits transfer.

The combined model is still an approximation. It assumes well-mixed vascular and interstitial compartments, concentrations that represent the sides of the barrier, and effective parameters that adequately summarize capillary heterogeneity. It does not resolve spatial concentration gradients, time-varying barrier properties, or binding and cellular handling within tissue. In particular, an observed tissue concentration may reflect processes beyond transcapillary entry; this flux equation alone does not assign those effects to \(PS\) or \(\sigma_{\mathrm{IgG}}\).

Useful checks include verifying that \(PS\) and \(J_v\) have units of volume per time, that \(\sigma_{\mathrm{IgG}}\) lies between 0 and 1 under this convention, and that the net flux sign agrees with the calculated components. If the flux is unexpectedly large, inspect concentration units and the scale of \(PS\) and \(J_v\). If its direction is unexpected, check the sign convention, the water-flow direction, and which side supplies the concentration in convection. A zero concentration gradient does not imply zero total flux when water flows, and zero water flow does not imply zero total flux when a concentration gradient remains.

• Do not treat reflection as permeability. A high \(\sigma_{\mathrm{IgG}}\) suppresses convective carriage; it does not, by itself, set diffusive \(PS\) to zero.
• Do not use the wrong source concentration. Outward convection uses \(C_v\); inward convection uses \(C_i\).
• Do not assume hydrostatic pressure alone determines flow direction. If osmotic terms are modeled, use the sign of net \(J_v\).
• Do not interpret net flux as gross exchange in both directions. It is the signed difference between opposing contributions, not a measurement of either one separately.