October 1999

Abstract

Perfluorocarbons are a promising new class of molecules without direct toxicity, and possessing interesting properties on the respiratory gases as well as on the other inert gases. Their principal indications are related to their oxygen-carrying capacity. Pure solutions can be used in the injured lungs where they improve both oxygenation and lung mechanics. Their application as a treatment of the Adult Respiratory Distress Syndrome is currently investigated.
Perfluorocarbon emulsions are administered intravenously for their oxygen-carrying capacity, and utilized as blood substitutes. Although perfluorocarbon emulsions are often compared to hemoglobin-based solutions, these two families of compounds do not compete with each other because both of them act on a very different mode. Future investigations combining both types of molecules could take benefit from all of their advantages. Furthermore, perfluorochemicals have also interesting properties on the other gases that are of interest for instance in the treatment of acute carbon monoxide poisoning or decompression sickness.

Introduction

Since decades, the medical world tries to develop molecules that can act as blood without having any of its drawbacks, the so-called " blood substitutes". On the edge of the 21st century, two families of molecules represent those substitutes: the hemoglobin-based solutions and the perfluorocarbons. Blood substitute is a misnomer as both products only present an oxygen carrying property. They do not act on the coagulation, immunity or any other of the multiple properties of blood and it would therefore be more appropriate to call them " oxygen carriers ".
The present review will focus on one class of these oxygen carriers: the perfluorocarbons. These molecules are less efficient oxygen carriers than hemoglobins since their oxygen-carrying ability is linearly dependent on the inspired fraction of oxygen, but the perfluorocarbons also possess other interesting medical properties.

Once upon a time…

The story of perfluorocarbons (PFC) began in the middle of World War II (1). In the desert of Alamo, scientists trying to develop the hydrogen bomb needed a chemically inert molecule to manipulate highly sensitive components. They observed that substituting hydrogen atoms of carbohydrates with atoms of fluor produces electrovalent liaisons so strong that they prevent any kind of reaction with other molecules (figure 1) :

perfluorocarbons were born. Three conclusions at this point: perfluorocarbons consist in an infinity of molecules, they all present the same lack of chemical toxicity because of their chemical inertness, therefore their interaction with surrounding molecules depends only on their physical properties. One of these properties , summarized in table 1,

rapidly raised interest among the medical world: the high dissolution capacity of inert gases. Gaseous molecules could probably stay embedded in chemically inert perfluorocarbons significantly more than in the corresponding carbohydrates. Unfortunately, the relatively high volatility of some of these molecules -compared to their corresponding carbohydrate - reduces the number of those compatible with the living species (2). The remaining bio-compatible perfluorocarbons feature comparable oxygen contents, twice that of whole blood at room air ( table 2).

Besides, contents of other gases like CO2, N2, NO or CO are also significantly increased in perfluorocarbons, which might raise interest for other clinical indications (3).
The quest for biological and medical applications for these new molecules has never abatted since. In the mid-sixties, Clark and Golan published the result of a sensational experiment (4): a mouse immerged in an oxygenated bath of perfluorocarbons stayed alive for more than an hour and most of all, surviving without any damage (figure 2).

From this time on, scores of biological applications have been investigated, using two types of preparations: emulsions of perfluorocarbons and plain liquid perfluorocarbons. One of the constant physical properties of perfluorocarbons is their lake of solubility in water and - although less marked - in lipids (table1). Hence, they need to be emulsified in order to be administered in the blood compartment. Plain liquid perfluorocarbons are administered topically in the lungs or in the ocular globe. This categorization is the source for distinct indications and properties, and also for distinct limitations and toxicity, which may be misleading as one reads the abundant literature about perfluorocarbons.
Plain liquid perfluorocarbons.
Liquid ventilation directly originated from Clarke and Golan 's experiment. Initially, the authors utilized a specific ventilator and circuit, totally filled with perfluorocarbons, that assured the respiratory gas exchange and made the fluids circulate : they called it Total Liquid Ventilation (TLV). In the early eighties, studies comparing TLV to conventional ventilation reported better gas exchange and improved lung mechanics(5) in animal models of Acute Respiratory Distress Syndrome (ARDS). However, the use of such ventilators was depicted as cumbersome and needed a prolonged training. This is why investigators rapidly developped an easier mode of Partial Liquid Ventilation (PLV), that compares advantageously with TLV (6). It consists in a conventional pressure-controlled ventilation of a patient whose Functional Residual Capacity (CRF) is filled up with liquid perfluorocarbons. Animal studies and preliminary clinical investigations applying this technique in premature infants as well as in adults presenting with ARDS have demonstrated improved arterial oxygenation and lung compliance (7). If these studies do not demonstrate a clear amelioration of the outcome of these patients, they showed at least that this mode of ventilation is harmless. Postulated mechanisms of action of perfluorocarbons in ARDS are: recruitment of dependent atelectatic alveoles because of their high density and their surfactant-like effect (8), the presence of oxygen-carrying perfluorocarbons in the alveolus preventing it to collapse and maintaining oxygenation even during the expiratory phase (9), a lavage effect on inflammatory mediators in the lungs (10), a direct effect on neutrophils and platelets aggregation in pulmonary capillaries that reduces pulmonary hypertension (11), blood redistribution from dependent to better ventilated non-dependent areas of the injured lung - due to the elective presence of heavy perfluorocarbons in the dependent alveoles (11).
The perfluorocarbons properties in this category are essentially the same as those previously described. There is no redistribution in the systemic circulation but rather an elimination by evaporation. Therefore, side effects are limited to the mechanical effect observed when perfluorocarbons are poured into the lung, which may produce cardiovascular instability and transient hypoxemia (12). True toxicity is thus absent since liquid perfluorocarbons keep being chemically non-reactive.
A new mode of administration of plain perfluorocarbons has recently been published and consists in vaporizing one of them (13). This greatly simplifies the material needed and only requires a vaporizer with an anesthetic machine. The effects seems to persist several hours after administration and the system probably utilizes less perfluorocarbons than PLV. It also takes advantage of small perfluorochemicals rejected in the past because of their high vapor pressure, causing gaseous emboli when administered intravenously or limitating their half-life when poured into the lungs (14-15). The volatility of perfluorocarbons is indeed highly variable and is mainly dependent on their molecular weight (2).
Other indications for topical perfluorocarbons are to be found in ophtalmology, graft conservation or alveolar drug delivery but they do not concern anesthetic or intensive care utilization and will therefore not be addressed in this review.

Perfluorocarbon emulsions

The best known application of perfluorocarbon emulsion is their utilization as blood substitute. Indeed, the FDA only agrees to consider their approval in the perspective of "reducing the amount of allogeneic blood units transfused". As we will understand in this paragraph, this indication is probably not the best for perfluorocarbons, better suited for biological gas transport. Since these liquids are immiscible in water, an intravenous injection has the potential to form a (fatal) liquid embolus. In 1968, Geyer produced a micro-emulsion of a perfluorocarbon in normal saline, and conducted a complete exchange transfusion in a rat, which survived breathing 100% oxygen with a hematocrit of zero: perfluorocarbon emulsions were born (14). However, the primary technical problem in producing a clinically usable product is the development of a stable emulsion, and the major consequence of being an emulsion is the alteration of the perfluorocarbons properties. Emulsions have thus several shortcomings: their perfluorocarbon concentration is limited, which contributes to curtail the concentration that can be reached in blood, their ability to transport a given gas depends on its partial pressure, and they disappears rapidly from the intravascular compartment.
The first emulsion licensed for clinical use, Fluosol-DA ®, contained only 20% weight per volume (g/100 mL) perfluorocarbons. As summarized in table 3,

oxygen content in F-Decalin -the main component of Fluosol- is about five-fold the oxygen content in this emulsion at a FiO2 = 1.0, thereby exacerbating the already described dependency of plain perfluorocarbons for oxygen partial pressure (figure 3).

Another important limitation comes from the short intra-vascular residency time of the emulsion micelles mainly related to their elimination processes. Perfluorocarbons are not metabolized but are cleared from the vascular space by the reticuloendothelial system and collected in the liver and spleen, eventually leaving the body as a vapor in the respiratory gases (15). After 1 hour, about 20% of the total administered dose is cleared and intra-vascular half-time of the remaining varies from 12 to 24 hours. Reticuloendothelial half-time varies from 7 to 65 days, depending on each perfluorocarbon characteristics, such as molecular weight or the adjunction of the more lipophilic atom of bromide (2) Repeated administration is therefore unadvisable and causes hepatosplenomegaly.
The concentration of perfluorocarbons reached in plasma is measured by the fluorocrit that is defined as the relative height of the column of perfluorocarbons in spun whole blood. The toxicity inherent to the emulsifier limits its maximum value to 8%. The first experimental emulsions were very unstable and produced lethal gaseous emboli (16). In 1976, the Japanese introduced the first clinical emulsion: Fluosol-DA®. This decision was probably premature as the perfluorocarbon concentrations were too low to significantly affect oxygen transport and the product presented many side effects. After ten years of clinical experience, dose-dependent liver and spleen accumulation were reported as were troubles of coagulation (17). Dizziness, drops in platelet and white blood cells counts were noted during the first clinical trials in the U.K. (18). In the U.S., anaphylactoïd reactions and pulmonary hypertensions were reported (19). Investigators rapidly found that all these effects were mediated through complement activation by Pluronic F68, the emulsifier of Fluosol-DA®, but not by any of its two perfluorocarbon components (20). Direct activation of platelets by perfluorocarbons has also been demonstrated in specific experimental conditions in the early nineties (21). Fluosol-DA was also difficult to use because the emulsion had to be stocked frozen, and before it could be used it had to be thawed and mixed with two other components.
Fluosol-DA® was therefore refuted by the FDA as a blood substitute but gained access to approval for distal arterial perfusion during the balloon inflation phase of coronary artery angioplasty (22): being de facto recognized as an efficient oxygen carrier.
This decision renewed the interest for perfluorocarbons and a second generation of emulsions were developed in the mid-eighties, these new preparations were emulsified with complement-not-reactive lecithin, and were stable at room temperature for more than six years (23). In contrast to Fluosol-DA®, Oxyfluor®(HemaGen Inc, St Louis MO) and Oxygent® (Alliance Pharm, San Diego CA) are ready to use, more concentrated (respectively 72% and 60% by weight), present less accumulation in the reticuloendothelial system, and less toxicity (see table 4). However, there are still cleared within 24-36 hours and repeat dosing to animals is known to result in hepatosplenomegaly. In addition, there appears to be a dose-related effect on platelet counts, with counts decreasing by 15-25% on post-infusion day 2-4. No bleeding abnormalities have been noted with administration of these compounds to date (24). Given the general limitations of the emulsions, i.e. that they require high-inspired oxygen partial pressures and will have an intra-vascular half-life of less than one day, clinical studies currently focus on teir use as intra-operative oxygen supplements when used in conjunction with normovolemic hemodilution. Clinical phase II efficacy trials have been published in patients undergoing radical prostatectomy and/or cystectomy (25). They conclude that these emulsions used with 100% oxygen are significantly more effective than blood at reversing transfusion triggers (vital signs, cardiac output, oxygen extraction ratio and mixed venous oxygen tension). A phase III study is underway in the same population of patients, which tries to demonstrate a reduced need for allogeneic blood when normovolemic acute hemodilution with a perfluorocarbon emulsion is performed. The effectiveness of perfluorocarbon emulsions is probably attributable to their ability to rapidly increase tissue oxygen tension. Hence, their most attractive indications will be related to their action on gas exchange physiology at all levels of the body, rather than their ability to substitute blood.
Perfluorocarbons are thus a promising new class of molecules lacking direct toxicity, and possess interesting properties on the respiratory gases as well as on the other inert gases. Their principal indications are related to their improved oxygen-carrying capacity. Pure solutions can be used in the airways of injured lungs where they improve oxygen exchange and lung mechanics. Their application as a treatment of the Adult Respiratory Distress Syndrome is currently investigated.
Perfluorocarbon emulsions are intravenously administered for their oxygen-carrying capacity, and utilized as blood substitutes. Their toxicity results from the emulsifying agents. Although perfluorocarbon emulsions are often compared to hemoglobin-based solutions, these two families of compounds do not compete with each other because both of them act on a very different mode. On the contrary, future investigations should rather try to use both types of solutions in order to take full benefit from their combined advantages. Perfluorochemicals have also interesting properties on the other gases, which are of interest for instance in the treatment of acute carbon monoxide poisoning or decompression sickness.

References.
1. Biro GP. Perfluorocarbon blood substitutes. Crit Rev Oncol Hematol.Vol 6; 4: 311-74.
2. Riess JG. Reassessment of criteria for the selection of perfluorochemicals for 2nd generation blood substitutes: analysis of structure-properties relationships. Artif. Organs. 1984; 8 (1) : 44-56
3. Riess JG. Perfluorocarbon blood substitutes. Adv.Exp.Med.Biol. 1994; 345: 227-234
4. Clark LC, Golan F. Survival of mammals breathing organic liquids equilibrated with oxygen at atmospheric pressure. Science 1966; 152 : 155-156
5. Hirschl RB, Overbeck M.C., Parent A., et al. Liquid ventilation provides uniform distribution of perfluorocarbon in the settin of respiratory failure. Surgery 1994; 116: 159-168.
6. Fuhrman BP, et al. Perfluorocarbon-assisted-gas-exchange. Crit. Care Med. 1991; 19: 772-722.
7. Hirschl RB, et al. Liquid ventilation in adultss, children and full-term neonates. Lancet 1995; 346: 1201-1202.
8. Marini JJ. Down side up. A prone and partial liquid asset. Int. Care Med. 1995; 21: 963-965.
9. Smith TM, et al. A liquid perfluorocarbon decreases the invitro production of reactive oxygen species by alveolar macrophages. Crit Care Med 1995; 23: 1533-39
10. Varani J, et al. Perfluorocarbon protects lung epithelial cells from neutrophils-mediated lung injury. Shock 1975; 6: 339-44
11. Lowe C, Shaffer T. Pulmonary vascular resistance in the perfluorocarbon-filled lung. J. Appl. Physiol. 1986; 60: 154-59
12. Hernan LJ, et al. Cardiorespiratory effects of perfluorocarbon-associated-gas exchange at reduced oxygen concentrations. Crit Care Med 1995; 23: 553-559
13. Bleyl JU, et al. Vaporized perfluorocarbon improves oxygenation and pulmonary function in an ovine model of acute respiratory distress syndrome. Anesthesiology 1999; 91: 461-469.
14. Geyer RP, et al. Survival of rats having red cells totally replaced by emulsified fluorocarbons. Fed. Proc., 1968; 27: 384
15. Clark LC, et al. Response of rabbit lung as a criterion of safety for intravenous perfluorocarbons. Biomater. Artif. Cell Blood Subst. 1992; 20: 1085-1099
16. Geyer RP. Whole animal perfusion with fluorocarbon dispersions. Fed. Proc., 1970; 29: 1758-1763
17. Mitsuno.Fluosol-DA, a ten years clinical experience. International.Anesth. Clin.1985;23:169
18. Tremper KK. The preoperative treatment of severely anemic patients with a perfluorochemical oxygen-transport fluid, Fluosol-DA. NEJM 1986; 307 (5): 277-283
19. Police AM. Waxman KK. Pulmonary complications after Fluosol-DA administered to patients with life-threatening bmlood loss. Crit. Care Med. 1985 ; 13 (2) : 96-98.
20. Vercellotti GM, Hammerschmidt DE Craddock PR. Activation of plasma complement by perfluorocarbon artificial blood : a probable mechanism of adverse
pulmonary reactions in treated patients and rationale for corticosteroids prophylaxis. Blood. 1982; 59 (6) : 1299-1304
21. DJ Smith, Lane TA. Effect of a high concentration perflubron emulsion on platelet function. Biomat., Art.Cells & Biotech., 1993; 21 (2): 173-181
22. Kent KM, et al. Reduction of myocardial ischemia during percutaneous transluminal angioplasty with oxygenated Fluosol. Am. J. Cardiol.1990; 66: 279-284
23. Biro GP. Perfluorocarbon-based red blood cells substitutes. Transfus. Med. Rev. 1993; 7: 84-95
24. Goodnough LT, et al. A Randomized Trial of Acute Normovolemic Hemodilution Compared to Preoperative Autologous Blood Donation in Total Knee Arthroplasty. Vox Sang. 1999 Aug;77(1):11-16.
25. Monk TG, et al. Acute normovolemic hemodilution can replace preoperative autologous blood donation as a standard of care for autologous blood procurement in radical prostatectomy. Anesth Analg. 1997 Nov;85(5):953-8.