This approach remains experimental and a controlled study is presently under way. With experience, the principles and pitfalls of this approach are becoming clear. While we cannot yet claim success, a description is included for the benefit of readers. It is used when other support measures have failed.
Injury to the pulmonary capillary and alveolar membranes allows increased escape of plasma proteins, particularly albumin, which has a molecular weight of 69 kDa. All molecules in every endogenous protein fraction have the same molecular weight and the membrane injury affects them all equally. Conversely, artificial plasma expanders such as dextrans and polystarch solutions contain a continuous range of fractions, some of which may be retained by the injured membrane. Thus it may be possible to generate an oncotic gradient based on the retained fractions. There may be concerns about the potential adverse effects due to the accumulation of smaller fractions in the interstitium. These concerns have not been substantiated. Using hydroxyethyl starch solutions we have been able to reduce edema accumulation in an animal model of lung injury despite the accumulation of smaller fractions in the extravascular space. In our clinical studies we have used Pentaspan (Dupont), a hydroxyethyl starch solution containing fractions with molecular weights ranging from 10 to 450 kDa. The larger fractions are broken down into smaller fragments in the first 2 h. Thereafter, the average half-life of the effect is about 18 h. Fractions with molecular weights below 30 kDa are excreted in the urine and can be ultrafiltered. The macromolecules are ultimately broken down and metabolized.
Successful treatment will induce fluid reabsorption not only from the lungs but from all tissues. Intravascular fluid reabsorption results in increased filtration pressures and autodilution of the oncotic agent. Therefore it is essential to maintain the intravascular volume as constant as possible by vigorous diuresis or by an alternative avenue for fluid removal, such as an ultrafiltration circuit.
Monitoring and treatment protocol
We use hematocrit determinations to monitor changes in the circulating blood volume (DCBV):
where Hct(t) is the hematocrit at any time t. Hct(t) is referenced to the initial value. In ultrafiltration circuits, we use our own on-line monitor. We have found that HCT ^ is more useful than filling pressures for monitoring purposes. Increases in circulating blood volume frequently allow readjustment of inotropes and vasopressors, thus changing the filling pressures. Of course, hematocrit is only useful in the absence of bleeding. If frequent samples are taken, it is necessary to change the reference every few hours.
Was this article helpful?