PERFLUOROCHEMICAL (PFC) liquids are hydrocarbon molecules of 8 to 10 carbons in length where the hydrogens have been replaced by fluorine. These liquids have several unique properties; they are immiscible in water, have a density of approximately twice that of water, are chemically inert, and have nearly twenty times the solubility for gases as does water. Because of their chemical inertness and high solubility for oxygen, emulsions of PFCs in normal saline have been studied over the past 30 yr as oxygen-carrying colloids that may temporarily supplement oxygen transport in lieu of erythrocytes. 1,2,3,4,5Since PFCs carry oxygen by direct solubility, their contribution to oxygen content, like plasma, is directly proportional to the arterial oxygen tension (Pao2) and requires a high Pao2(>300 mmHg) to be effective. Another limitation of these emulsions is that they are cleared from the vascular space by the reticuloendothelial system relatively quickly, having a half-life in the range of 12–18 h. 3,4,5Once out of the vascular space, the PFC remains in the liver and spleen for a prolonged period of time, with a half-life measured in weeks, thereby limiting the advisability of multiple doses over a short period of time. 5Early emulsions in clinical trials suffered another limitation of emulsion instability, requiring the product to be stored frozen and only containing an “anemic” 10% PFC by volume. 3,4These last two limitations have been overcome by a more recent product, although the primary limitations of requiring high inspired oxygen to carry limited amounts of oxygen for a short period of time remain. 6Given these physical and physiologic constraints, using a PFC emulsion as a temporary replacement for erythrocytes in the traditional sense has not been found to be effective. 3,4It has been demonstrated that treatment with a PFC emulsion does contributed to oxygen transport, but the contribution has not been sufficient to either change outcome or reduce the need for a blood transfusion. 3,4,6In this current issue of the journal, Spahn et al. present a study attempting to demonstrate a reduced need for a erythrocytes transfusion by combining acute normovolemic hemodilution (ANH) and treatment with the PFC emulsion, perflubron. 7
PFC (and plasma) pick up and release oxygen in direct proportion to oxygen partial pressure. Therefore, in spite of the fact they do not carry a large volume of oxygen; the oxygen they do carry will largely be released in the tissue. For example, if the Pao2is 400 mmHg and the mixed venous Po2is 40 mmHg, 90% of the dissolved arterial oxygen will be released in the tissue, whereas only 25% of the hemoglobin bound oxygen will be released. Unfortunately with the doses of PFC used in this and other clinical studies, the maximum of obtainable fluorocrit (Fct) of 3%, (volume % of fluorocarbon in the blood, Fct, analogous to hematocrit), would contribute just under 1.0 volume percent (vol%) of oxygen content. One g/dl of hemoglobin (hematocrit ≅3%) will carry 1.34 vol% of oxygen. These comparisons would make PFC treatment seem analogous to blood treatment on a Fct versus hematocrit basis, given a Pao2of less than 300 mmHg, except for the dosing limitations and short intravascular half-life of the PFC. It has been demonstrated, in animal and clinical studies, that treatment with PFC will produce an immediate rise in mixed venous oxygen tension (Pvo2) implying a contribution to oxygen delivery. 6,8,9Understanding the limitations of dose, half-life, and relatively small contribution to oxygen transport, the authors have selected ANH as the clinical situation where a temporary supplementation to oxygen transport may ultimately reduce the overall need for erythrocytes. 7,9
It is the authors supposition that treatment with PFC will allow patients to undergo a more severe degree of intraoperative anemia, due to ANH and surgical blood loss, resulting in a decreased need for red cell transfusions compared with a control group. Unfortunately, in this study the control group is not a parallel control. The treatment group undergoes ANH to a hemoglobin of 8 g/dl then receives 1.8 g/kg of PFC emulsion (about 0.9 ml/kg or 67 ml of PFC). During surgery the hemoglobin is then allowed to drop to 6 g/dl at which point another 0.9 g/kg of PFC is given. The patient's hemoglobin then is allowed to drop to 5.5 g/dl before they are transfused with the ANH blood. The patients randomized to the controls are transfused with blood when their hemoglobin reaches 8 g/dl. In addition, the PFC group has the Fio2increased to 1.0, whereas the control group has a Fio2of 0.4. At the end of surgery the target hemoglobin for both groups is greater than 8.5 g/dl. At 24-h the ANH/PFC group did receive fewer transfused units than the control group, (1.5 vs. 2.1 units). But by postoperative day 3 this difference is no longer significant. In a subgroup of patients whose intraoperative blood loss was greater than 20 ml/kg the difference remained significant throughout the hospital stay (3.4 units vs. 4.9 units at 21 days or day of discharge). It is not clear that these findings were due to the PFC emulsion for the results are predictable from the ANH alone. 10
Both groups were transfused when they reached the protocol-defined hemoglobin level or if one of the following physiologic transfusion triggers was achieved: heart rate greater than 100 beats/min., mean arterial pressure less than 60 mmHg, Pvo2less than 38 mmHg, or ST segment changes. One way of assessing safety may have been to determine how many of these physiologic transfusion triggers were encountered in each of the groups. If, for example, the PFC group noted more ST segment changes associated with lower hemoglobin levels it might bring into question whether the PFC treatment and the high Fio2indeed made the treatment group as safe as the control group. On the other hand, if more of these physiologic triggers were encountered in the control group it may help provide evidence that the PFC and high Fio2treatment allowed a lower hemoglobin with less physiologic consequence.
More adverse events were noted in the treatment group with respect to cardiovascular events, (40%vs. 30%), and digestive system events, (7%vs. 2%). The authors speculate that the cardiovascular events may have been due to the unfamiliarity of some of the study sites with the process of ANH. They then describe the technical issues regarding ANH and the subsequent retransfusion with associated volume shifts. This is indeed a risk associated with ANH, which should be taken into account when one is trying to lower the overall risk of perioperative blood and fluid management.
Ultimately the mortality was twice as high in the PFC group, (10 out of 195, vs. 5 out of 195), although this did not reach significance. Given the current risk of HIV is approaching 1:1,000,000 and Hepatitis C less than 1:100,000; it would take a very large study to determine if ANH/PFC treatment improves safety. 11,12
It is always far easier to criticize a large complex clinical trial than it is to actually perform one, particularly one this large. The primary conclusion that the authors draw from their study, “that the use of perflubron emulsion as an intravenous oxygen therapeutic to augment autologous blood harvesting may represent a new alternative for…patients seeking to avoid or minimize the risks of allogeneic RBS transfusions…,” is intriguing, but is likely to trigger disagreement and a skeptical response from many. However, regardless of such criticisms, the authors must be congratulated for what is clearly the largest and most thorough effort to date to examine the utility of these compounds in operative medicine. I am not yet convinced, but I remain hopeful that the broader utility of PFC emulsions may be proven. This study is a reasonable step in that direction.