Biotechnol. Prog. 1992, 8, 454-457
454
Perfluorooctyl Bromide Dispersions in Aqueous Media for Biomedical Applications Stephane S. Habif, Pascal E. Normand, Christian B. Oleksiak, and Henri L. Rosano* Department of Chemistry, The City College of the City University of New York, Convent Avenue at 138th Street, New York, New York 10031 In studying perfluorooctyl bromide (PFOB)dispersions in aqueous media, we have used two types of surfactant: egg yolk phospholipids (EYP) and polyglycerol esters (PGE). Our interest in these dispersions arises from their potential biomedical applications as imaging solutions and oxygen-carrying solutions (i.e., blood substitutes). For EYP systems, we have identified the dispersion structure as consisting of (a) PFOB droplets (250-nm diameter) stabilized by a phospholipid monolayer adsorbed irreversibly at the o/w interface and (b) small empty phospholipid vesicles. With both surfactants commercial preparations yielded stable systems, while purified samples, being nondispersible, could not be made t o act as emulsifiers. In both cases, minor components in the commercial surfactant were found to be necessary for the formation of a stable dispersion, enabling the transport of the pure surfactant to the PFOB/water interface.
1. Introduction The increasingly difficult and critical task of assuring sufficient quantities of safe human blood for medical uses has spurred two distinct lines of research into blood substitutes: one focused on “red blood”, based on human hemoglobin, the other on “white blood”, perfluorocarbon(PFC-)based emulsions whose synthesis involves no blood substances. Most recently, Hilts (1991) reported the development, by DNX Inc. of Princeton, NJ, of three genetically engineered pigs capable of producing human hemoglobin. Despite DNX’s claims for its proprietary purification system, serious questions remain as to the safety of such hemoglobin, which could conceivably retain animal pathogens. For this reason alone, it is essential that research into the inherently safer-because they are synthetic-white bloods continue. But PFC emulsions offer many other potential applications as well. Alliance Pharmaceutical Corp., based in Otisville, NY, has developed an injectable, stable 100 % weight per volume (64% w/w) PFOB in saline emulsion. This emulsion is currently being used in clinical trials as a contrast agent in imaging systems (X-ray, Magnetic Resonance Imaging (MRI), Computed Tomography (CT), ultrasound) and as an oxygen carrier. The emulsion’s possible applications as an oxygenating agent include the treatment of cancer (by enhancing the efficacy of radiation therapy) and myocardial or cerebral infarctions and the extended preservation of organs for transplant. A very similar chemical, developed by Japan’s Green Cross Corp., has already been approved by the Food and Drug Administration (FDA) for use during percutaneous transluminal coronary angioplasty (PTCA). This landmark decision in December 1989, the first approval by the FDA of a PFC emulsion for medical use, lays the groundwork for future FDA reviews of other PFC-based products.
2. PFOB Dispersions in Aqueous Media Lipid emulsions have been previously prepared for intravenous nutrition (Hansrani et al., 1983). In our emulsion systems, we are using a perfluorocarbon oil instead of a hydrocarbon oil. 8756-7938/92/3008-0454$03.00/0
Emulsions of perfluorooctyl bromide (PFOB, CBFITBr, 1-bromoheptadecafluorooctane, density 1.93) in either saline or 5 % dextrose solutions were prepared using either egg yolk phospholipids (EYP) or polyglycerol esters (PGE) as a surfactant. Our preparation of these dispersions was guided by several constraints. The desired injectability determined our choice of surfactants: the two used are both metabolized in the human body. Successful preparation of a dispersion requires both an optimum quantity of surfactant and an optimum PFOB/aqueous solution ratio. Finally, EYP’s susceptibility to oxidation required that dispersion preparation with EYP be carried out in the absence of air. Rosano et al. (1991) have reported in detail the results obtained in their investigation of PFOB dispersions prepared using EYP as surfactant. To summarize the important points: (1)The method and conditions of preparation influence the dispersion stability. A large amount of mechanical work was found to be needed to prepare dispersions that remained stable for several months at room temperature. The composition of EYP varies substantially, and samples with different minor-component profiles yielded dispersions with different degrees of stability. (2) A picture obtained for a typical 100% w/v (64% w/w) PFOB/saline dispersion (Figure 1) and taken by transmission electron microscopy (TEM) reveals a polydispersed collection of concave and convex semispheres (averaging 250 nm in diameter as determined by photon correlation spectroscopy (PCS)). Both big droplets and smaller ones in between them are visible. (3) A sedimentation field flow fractionation (SFFF)fractogram of the same system exhibits two peaks (Figure 2). The first peak, which corresponds to 33% of the total EYP fraction by mass, as calculated by Tarara et ai. (19911, represents empty EYP vesicles, and the second larger peak (67% of the total EYP fraction by mass) represents the PFOB droplets. The small particles (80 nm) visible on the electron micrograph are probably empty vesicles. Emulsion characterization by the combined SFFF-PCS methods was described by Caldwell and Li (1989).
0 1992 American Chemical Society and American Institute of Chemical Engineers
m. 13.00.. 1892, Vd. 0, NO. 5 er-
4110 I-
-
Figure I. Transmimian electron micrograph of a 100% w/v (a?wiw) PFQB d i n e dispemion (0.9cm = 0.133 pm).
snmple
mntinuawr pharw
commercinl 10-2-0
p u n water ?B~;dcxtMav snlinc Mlution
~~
purified 10-2-0 nonpalnr ftcrction p l a t frwtion
: 11'
Figure2. Mimentation field flow fractionationftactog" of R
100ci w/v (M7 w/ w) PFQR/rrrrline dispersian.
(4) Given that 67 % of the EYP isadsorbed a t the PFOW water interface and that the average diameter of a PFOR droplet is 2.50 nm, we can calculate that there is a monolayer of EYP ad.sorbed at the interface. The dispersion is thus composed of PPOB droplets stabilized by a phospholipid monolayer adrrorhed at the interface and of small empty phospholipid vesicles. In a m n d series of experimenb, we aubtituted for EY P another surfactant, decaglyceroldioleate (10-2-0) (Figure 31, a polyglycerolester that is structurally similar to EY B but less sensitiveto oxidation. Polyglycerol esters are important f d emulsifiers available commercially as a mixture of variow isomers differing by the size of the polyglycerol, by the degree of esterification of the polyol, by the chain length and the degree of insaturation of the fatty acids, and by positional i-somerismof the fatty acids on the polyol. 10-2-0is a polyglycerol with an average of 10 molecules of glycerol esterified by two molecules of oleate (on average). We found (Table I) that commercial cramples of 10-2-0 yielded stable sterile 100% w/Pv (645 w!w) BFOB dispersions when pure water or a 5% d e x t r w solution waa used as a continuous phaae. With a saline solution the system wm unstable. The presence of ions was found to be detrimental to stability. A tentativeexplanation is the formation of soaps of the free fatty acids (2.8''C as oleic) found in the 10-2-0 sample in the presence of salts. lye purified the 10-2-0 by elution on a column of silica gel (Merck, grade 60) in order to separate the polar compoundsoriginally present. We obtained two fractions: one nonpolar and one polar. With the nonpolar fraction, we were unable to obtain a dispersion at all. M?th the polar fraction, we obtained a dispersion easilv, but i t waq unstable. The polar fraction, then, i R neoessary for
S S dextrose SC; dcxtmbc
tmuk atable dhpcmion s ~ h l dispemion e unstable diqxmian
no dinpcrnion
enrrily dinpcmiblt unstable emulnion
dispersion formation, while the nonpolar fraction is n-ry for dispersion stability. Since 'nonpolar" 10-2-0 and pure phmpholipids (PC, PE)are not dispemible in water, they cannot act o surfactants by them-selves, being unable to ad.sorb a t the oil/ water interface. Davis and Hansrani (1979) recognized the importance of the minor components of EYP (e.g., lysolecithin) in determining its ability to produce stable oillwater emulsions. To suclceed in making an emulsion, then, we mwt have wme polar compounds p m n t that will somehow disperae the molecules of the insoluble surfactant and enable them to adsorb a t the PFOR/ water interface. The question now is %hy do them insoluble aurfactants provide stability?"
3. Surface Properties of EYP and PCE We camed out wme surface and interfacial tension experiments. For purpows of calibration, we began by measuring the surface tension of aqueous solutions of increasing concentrationof .sodium dodecyl sulfate (SBS), a typical water-soluble surfactant, using a sandblasted platinum blade suspended from a tensiometer; we a h measured the interfacial tension between PFOW and SDS solutions using a .sandblasted Teflon blade. The SDS solution PFOH interfacial tension decrea!! from .WmW m for the pure water)BFBW interface to 10 mN/ m at the SBS critical micellar concentration (CMC).This is clear evidence for the molecular adsorptionof SDSat the BFOR! water interface (Figure 4). Then we u . d a micmyringe to form a drop of PFOB near the interface between pure PFOR and SBSsolutions (Figure 5). The volume of the drop and itar lifetime (i.e., the time before i t maid with the bulk of WOW) were recorded. From the volume of the PFOB drop and the calibration curve obtained with the Teflon blade method, we calculated the interfacial tension. Using this method, we measured the interfacial tension in the presence of dipalmitoylphmphatidylcholine(BPPC, a pure phospholipid) and purified 'nonpolar" 10-2-0. DPPC and 10-2-0 are water-insolublesurfactants, so the samples were solubilized in P F 0 8 using hexane and a minimum amount of ethanol. The results, presented in Table II, ahow that DPPC and 10-2-0 greatly reduce the PFOR/water interfacial tension provided they can be made to adsorb molecularly
Biotechnol. hog., 1992, Vol. 8, No. 5
456 7r;
- .~--.-\.
._. .__ ._. _. - .._.
0
Surf. Tens. (mNlm)
waterlair
5
4 8o ‘ 70
Interf. Tens. (mN/m)
60
-
- 400
-
60-
waterlPFOB
_....-_.......
E e
-300 4030
15
I
0
i
. O O O 0
350
50-
- 250
-
S
g
%
.
O 0
0
50
100
150
200
surface area (A2 /molecule) 80
WDSI
Figure 4. Surface tension of SDS solutions and interfacial tension between PFOB and SDS solutions versus SDS concentration at 24 O C .
PFOe
I
(M) 60 h
microsyringe tip
I/
0
50
IC0
150
200
surface area (A2 /molecule)
f
=
SDS
Figure 6. (Top) Surface isotherms of commercial EYP on aqueous substrate containing Ca2+at pH 8.15 and 24 OC and (bottom) isotherms (compression-decompression) of commercial EYP on aqueous substrate containing Ca2+at pH 8.15 and 24 “C.
J
Table 11. Interfacial Tension Measurements at 25 “C SamDle interfacial tension PFOB/hexane/ethanol, 7112514 w/w 50 mN/m DPPC/PFOB/hexane/ethanol,1/71/2414w/w
454
Perfluorooctyl Bromide Dispersions in Aqueous Media for Biomedical Applications Stephane S. Habif, Pascal E. Normand, Christian B. Oleksiak, and Henri L. Rosano* Department of Chemistry, The City College of the City University of New York, Convent Avenue at 138th Street, New York, New York 10031 In studying perfluorooctyl bromide (PFOB)dispersions in aqueous media, we have used two types of surfactant: egg yolk phospholipids (EYP) and polyglycerol esters (PGE). Our interest in these dispersions arises from their potential biomedical applications as imaging solutions and oxygen-carrying solutions (i.e., blood substitutes). For EYP systems, we have identified the dispersion structure as consisting of (a) PFOB droplets (250-nm diameter) stabilized by a phospholipid monolayer adsorbed irreversibly at the o/w interface and (b) small empty phospholipid vesicles. With both surfactants commercial preparations yielded stable systems, while purified samples, being nondispersible, could not be made t o act as emulsifiers. In both cases, minor components in the commercial surfactant were found to be necessary for the formation of a stable dispersion, enabling the transport of the pure surfactant to the PFOB/water interface.
1. Introduction The increasingly difficult and critical task of assuring sufficient quantities of safe human blood for medical uses has spurred two distinct lines of research into blood substitutes: one focused on “red blood”, based on human hemoglobin, the other on “white blood”, perfluorocarbon(PFC-)based emulsions whose synthesis involves no blood substances. Most recently, Hilts (1991) reported the development, by DNX Inc. of Princeton, NJ, of three genetically engineered pigs capable of producing human hemoglobin. Despite DNX’s claims for its proprietary purification system, serious questions remain as to the safety of such hemoglobin, which could conceivably retain animal pathogens. For this reason alone, it is essential that research into the inherently safer-because they are synthetic-white bloods continue. But PFC emulsions offer many other potential applications as well. Alliance Pharmaceutical Corp., based in Otisville, NY, has developed an injectable, stable 100 % weight per volume (64% w/w) PFOB in saline emulsion. This emulsion is currently being used in clinical trials as a contrast agent in imaging systems (X-ray, Magnetic Resonance Imaging (MRI), Computed Tomography (CT), ultrasound) and as an oxygen carrier. The emulsion’s possible applications as an oxygenating agent include the treatment of cancer (by enhancing the efficacy of radiation therapy) and myocardial or cerebral infarctions and the extended preservation of organs for transplant. A very similar chemical, developed by Japan’s Green Cross Corp., has already been approved by the Food and Drug Administration (FDA) for use during percutaneous transluminal coronary angioplasty (PTCA). This landmark decision in December 1989, the first approval by the FDA of a PFC emulsion for medical use, lays the groundwork for future FDA reviews of other PFC-based products.
2. PFOB Dispersions in Aqueous Media Lipid emulsions have been previously prepared for intravenous nutrition (Hansrani et al., 1983). In our emulsion systems, we are using a perfluorocarbon oil instead of a hydrocarbon oil. 8756-7938/92/3008-0454$03.00/0
Emulsions of perfluorooctyl bromide (PFOB, CBFITBr, 1-bromoheptadecafluorooctane, density 1.93) in either saline or 5 % dextrose solutions were prepared using either egg yolk phospholipids (EYP) or polyglycerol esters (PGE) as a surfactant. Our preparation of these dispersions was guided by several constraints. The desired injectability determined our choice of surfactants: the two used are both metabolized in the human body. Successful preparation of a dispersion requires both an optimum quantity of surfactant and an optimum PFOB/aqueous solution ratio. Finally, EYP’s susceptibility to oxidation required that dispersion preparation with EYP be carried out in the absence of air. Rosano et al. (1991) have reported in detail the results obtained in their investigation of PFOB dispersions prepared using EYP as surfactant. To summarize the important points: (1)The method and conditions of preparation influence the dispersion stability. A large amount of mechanical work was found to be needed to prepare dispersions that remained stable for several months at room temperature. The composition of EYP varies substantially, and samples with different minor-component profiles yielded dispersions with different degrees of stability. (2) A picture obtained for a typical 100% w/v (64% w/w) PFOB/saline dispersion (Figure 1) and taken by transmission electron microscopy (TEM) reveals a polydispersed collection of concave and convex semispheres (averaging 250 nm in diameter as determined by photon correlation spectroscopy (PCS)). Both big droplets and smaller ones in between them are visible. (3) A sedimentation field flow fractionation (SFFF)fractogram of the same system exhibits two peaks (Figure 2). The first peak, which corresponds to 33% of the total EYP fraction by mass, as calculated by Tarara et ai. (19911, represents empty EYP vesicles, and the second larger peak (67% of the total EYP fraction by mass) represents the PFOB droplets. The small particles (80 nm) visible on the electron micrograph are probably empty vesicles. Emulsion characterization by the combined SFFF-PCS methods was described by Caldwell and Li (1989).
0 1992 American Chemical Society and American Institute of Chemical Engineers
m. 13.00.. 1892, Vd. 0, NO. 5 er-
4110 I-
-
Figure I. Transmimian electron micrograph of a 100% w/v (a?wiw) PFQB d i n e dispemion (0.9cm = 0.133 pm).
snmple
mntinuawr pharw
commercinl 10-2-0
p u n water ?B~;dcxtMav snlinc Mlution
~~
purified 10-2-0 nonpalnr ftcrction p l a t frwtion
: 11'
Figure2. Mimentation field flow fractionationftactog" of R
100ci w/v (M7 w/ w) PFQR/rrrrline dispersian.
(4) Given that 67 % of the EYP isadsorbed a t the PFOW water interface and that the average diameter of a PFOR droplet is 2.50 nm, we can calculate that there is a monolayer of EYP ad.sorbed at the interface. The dispersion is thus composed of PPOB droplets stabilized by a phospholipid monolayer adrrorhed at the interface and of small empty phospholipid vesicles. In a m n d series of experimenb, we aubtituted for EY P another surfactant, decaglyceroldioleate (10-2-0) (Figure 31, a polyglycerolester that is structurally similar to EY B but less sensitiveto oxidation. Polyglycerol esters are important f d emulsifiers available commercially as a mixture of variow isomers differing by the size of the polyglycerol, by the degree of esterification of the polyol, by the chain length and the degree of insaturation of the fatty acids, and by positional i-somerismof the fatty acids on the polyol. 10-2-0is a polyglycerol with an average of 10 molecules of glycerol esterified by two molecules of oleate (on average). We found (Table I) that commercial cramples of 10-2-0 yielded stable sterile 100% w/Pv (645 w!w) BFOB dispersions when pure water or a 5% d e x t r w solution waa used as a continuous phaae. With a saline solution the system wm unstable. The presence of ions was found to be detrimental to stability. A tentativeexplanation is the formation of soaps of the free fatty acids (2.8''C as oleic) found in the 10-2-0 sample in the presence of salts. lye purified the 10-2-0 by elution on a column of silica gel (Merck, grade 60) in order to separate the polar compoundsoriginally present. We obtained two fractions: one nonpolar and one polar. With the nonpolar fraction, we were unable to obtain a dispersion at all. M?th the polar fraction, we obtained a dispersion easilv, but i t waq unstable. The polar fraction, then, i R neoessary for
S S dextrose SC; dcxtmbc
tmuk atable dhpcmion s ~ h l dispemion e unstable diqxmian
no dinpcrnion
enrrily dinpcmiblt unstable emulnion
dispersion formation, while the nonpolar fraction is n-ry for dispersion stability. Since 'nonpolar" 10-2-0 and pure phmpholipids (PC, PE)are not dispemible in water, they cannot act o surfactants by them-selves, being unable to ad.sorb a t the oil/ water interface. Davis and Hansrani (1979) recognized the importance of the minor components of EYP (e.g., lysolecithin) in determining its ability to produce stable oillwater emulsions. To suclceed in making an emulsion, then, we mwt have wme polar compounds p m n t that will somehow disperae the molecules of the insoluble surfactant and enable them to adsorb a t the PFOR/ water interface. The question now is %hy do them insoluble aurfactants provide stability?"
3. Surface Properties of EYP and PCE We camed out wme surface and interfacial tension experiments. For purpows of calibration, we began by measuring the surface tension of aqueous solutions of increasing concentrationof .sodium dodecyl sulfate (SBS), a typical water-soluble surfactant, using a sandblasted platinum blade suspended from a tensiometer; we a h measured the interfacial tension between PFOW and SDS solutions using a .sandblasted Teflon blade. The SDS solution PFOH interfacial tension decrea!! from .WmW m for the pure water)BFBW interface to 10 mN/ m at the SBS critical micellar concentration (CMC).This is clear evidence for the molecular adsorptionof SDSat the BFOR! water interface (Figure 4). Then we u . d a micmyringe to form a drop of PFOB near the interface between pure PFOR and SBSsolutions (Figure 5). The volume of the drop and itar lifetime (i.e., the time before i t maid with the bulk of WOW) were recorded. From the volume of the PFOB drop and the calibration curve obtained with the Teflon blade method, we calculated the interfacial tension. Using this method, we measured the interfacial tension in the presence of dipalmitoylphmphatidylcholine(BPPC, a pure phospholipid) and purified 'nonpolar" 10-2-0. DPPC and 10-2-0 are water-insolublesurfactants, so the samples were solubilized in P F 0 8 using hexane and a minimum amount of ethanol. The results, presented in Table II, ahow that DPPC and 10-2-0 greatly reduce the PFOR/water interfacial tension provided they can be made to adsorb molecularly
Biotechnol. hog., 1992, Vol. 8, No. 5
456 7r;
- .~--.-\.
._. .__ ._. _. - .._.
0
Surf. Tens. (mNlm)
waterlair
5
4 8o ‘ 70
Interf. Tens. (mN/m)
60
-
- 400
-
60-
waterlPFOB
_....-_.......
E e
-300 4030
15
I
0
i
. O O O 0
350
50-
- 250
-
S
g
%
.
O 0
0
50
100
150
200
surface area (A2 /molecule) 80
WDSI
Figure 4. Surface tension of SDS solutions and interfacial tension between PFOB and SDS solutions versus SDS concentration at 24 O C .
PFOe
I
(M) 60 h
microsyringe tip
I/
0
50
IC0
150
200
surface area (A2 /molecule)
f
=
SDS
Figure 6. (Top) Surface isotherms of commercial EYP on aqueous substrate containing Ca2+at pH 8.15 and 24 OC and (bottom) isotherms (compression-decompression) of commercial EYP on aqueous substrate containing Ca2+at pH 8.15 and 24 “C.
J
Table 11. Interfacial Tension Measurements at 25 “C SamDle interfacial tension PFOB/hexane/ethanol, 7112514 w/w 50 mN/m DPPC/PFOB/hexane/ethanol,1/71/2414w/w
