Complement activation by artificial blood substitute Fluosol: in vitro and in vivo studies F. HONG, K.A. SHASTRI, G.L. LOGUE,AND M.B. SPAULDING A perfluorocarbon blood substitute, Fluosol, is undergoing clinical trials as an ad’unct to chemotherapy. The adverse effects associated with its administration have been postulated to result from complement activation. When gel electrophoresis and Western blotting of Fluosol are used after its incubation with serum, activated C3 and factors Bb and H are bound to the Fluosol particles in a timedependent fashion, which suggests that complement activation with Fluosol, as does that with mosan, occurs on the surface of the particles. Paradoxically, it is found, both by t e measurement of Fluosol-bound C3d and by fluid-phase C5a, that lower concentrations of Fluosol cause greater amounts of complement activation, which suggests a complex interaction of activators and inhibitors that changes as the available surface area is decreased. Studies performed with bystander red cell-bound C3d demonstrated in vivo complement activation occurring in six patients receiving fluosol as an adjunct to chemotherap for colon cancer. In two patients, there was a marked increase in red cell-bound 23d after Fluosol infusion; these two patients also developed adverse reactions during fluosol infusion. These studies suggest that the Fluosol surface plays a major role in the initiation and re ulation of complement activation that is seen during Fluosol infusion. TAANSFUSlbN 1991;31;642-647.
x
Abbrevlatlons: MRI = magnetlc resonance Imaging; VBS = Vrronal-buffered rallne.
FLUOSOL-DA (Green Cross, Osaka, Japan [Fluosol, Alpha Therapeutics, Los Angeles, CAI) is a perfluorocarbon blood substitute that is undergoing phase I and I1 clinical trials in cancer patients to determine its efficacy when combined with radiation or chemotherapy. Other potential uses for perfluorocarbonsare in magnetic resonance imaging (MRI), transplant organ perfusion, drug delivery systems, and the treatment of shock.’ Recently, Fluosol was approved by the United States Food and Drug Administration for use in coronary angioplasty. Up to 30 percent of patients receiving Fluosol in the United States develop adverse reactions consisting of shortness of breath, chest tightness, and wheezing.* (However, Fluosol has been administered to a large number of patients in Japan without significant adverse effects3) On the basis of studies with human serum in , ~ adverse reactions have vitro and on animal ~ t u d i e sthe been attributed to the activation of the alternative path-
way of complement. In this study, we describe in vitro and in vivo experiments performed to define the mechanism by which whole Fluosol particles activate complement. We initially examined perfluorocarbons’ in vitro activation of complement by using a monoclonal antibody to C3d bound to the Fluosol particle and by assays for fluid-phase C3a des Arg (C3a) and C5a des Arg (C5a). To explore further the mechanism of the activation of complement on the surface of the Fluosol, we subjected proteins bound to the Fluosol to electrophoresis and identified them with Western blotting. The addition of red cells to the in vitro system during complement activation resulted in increased C3d binding to the red cells. This finding was applied to patient studies, and red cells were assayed before and after infusion of Fluosol for changes in C3d. Materials and Methods C3a and C5a assays We measured C3a and C5a by a radioimmunoassay system according to manufactur(Amersham, Arlington Heights, L), er’s instructions. In this system, C5 or C3 that docs not react is precipitatedout, and known amounts of antibody and labeled ligand are added. The reaction mixture is allowed to incubate and then centrifuged. Measurement of the radioactivity in the pellet allows calculation of the amount of labeled C5a or C3a. The amount of competitive inhibition of labeled ligand by unknown concentrations of C5a or C3a was compared to that produced by known standard amounts.
From the Department of Medicine, State University of New York at Buffalo; the Transfusion Services, Buffalo General Hospital; and the Veterans Administration Medical Center, Buffalo, New York. Supported in part by the Margaret Duffy and Robert Cameron Troup Memorial Fund for Cancer Research of the Buffalo General Hospital, the Research Service of the Veterans Administration Medical Center, the Ralph Hcchstetter Medical Research Fund in Honor of Dr. Henry C. and Bertha H. Buswell of the State University of New York at Buffalo, and Alpha Therapeutics. Received for publication January 11, 1991; revision received April 15, 1991, and accepted April 16, 1991.
642
‘IRANSNSION 1991-Vd. 31. No. 7
643
COMPLEMENT ACTIVATION BY FLUOSOL
Assay for C3d deposition on red cells or Fluosol We radiolabeled the murine monoclonal C3d antibody (Cytotech, San Diego, CA)with *=I using chloramine T and measured red cell- or Fluosol-bound C3d as described previou~ly.~ Briefly, either washed red cells (5% suspension by volume in barbital [Veronall-buffered saline [VBS], pH 7.4, with 0.15 mM [0.15 mmoVL] CaCI, and 0.5 mM [0.5 mmoVL] MgCl,) or Fluosol was incubated with radiolabeled anti-C3d that had been layered Over phthalate oils in microfuge tubes. At the end of the incubation, the tubes were spun for 10 minutes. We cut the tips and counted them in a gamma counter. The results were expressed in fg of anti-C3d bound per cell, per volume of Fluosol, or per total amount of Fluosol in a reaction mixture as described below.
Electroelution We performed electroelution of proteins bound to Fluosol by a modification of the methods used to study complement components bound to dialyzer membranes.6 Fluosol was incubated with serum for a specified time. Subsequently, the Fluosol was spun at 970 x g for 10 minutes and washed with VBS. After resuspension, we reduced the Fluosol with 6-mercaptocthanol for 15 minutes. The sample was layered Over a 5-percent sodium dodecyl sulfate stacking gel with a 10-percent SDS resolving gel by the Laemmli method. We performed Western blotting for 1 hour in 0.20 mM (0.20 mmol/L) Tris buffer with plyclonal goat antibodies that were specific for complement proteins of interest. We washed the gel three times with Tris buffer and incubated it for 1hour with alkaline phosphatase-labeled anti-goat antibody. We then washed the nitrocellulose with Tris buffer three more times and carried out the enzymatic reaction by using a substrate system (LaboratoriesPhosphatase, Kirkegaard and Perry, Gaithersburg, MD).
Nuclear MRI to detect the presence of Fluosol on cells We carried out MRI studies to determine the presence of Fluosol on red cells.’ With the MRI device (Model Fx270, Joel, Peabody, MA) set at a frequency of 253.75 mHz, we
scanned each sample for lo00 degradations. Using known concentrations of Fluosol, we obtained a standard curve correlating the height of the measured peak with the concentration of Fluosol present in each sample.
Patient studies All patient studies were done as approved by the institutional review board at the Veterans Administration Medical Center. All patients were being treated for metastatic adenocarcinoma to the liver, and they received Fluosol at a dosc of 400 mL per m2. After an initial test dose of 0.5 mL, all patients received Fluosol at an infusion rate of 1 mL per minute for the first 5 minutes, 5 mL per minute for the next 5 minutes, and then 10 mL per minute until completion of the total dose of Fluosol. Samples were obtained in EDTA tubes 30 minutes before the infusion of Fluosol and 30 minutes after completion of the infusion. After collection, the samples were immediately centrifuged and separated. Red cells were refrigerated and assayed the following day. We performed studies that showed that red cells obtained from patients after Fluosol infusion and assayed after 24- and &-hour storage did not differ in C3d content from red cells that were assayed immediately. Control patients received chemotherapy only. Their blood samples were obtained in a similar fashion.
In vitro studies of complement activation by Fluosol We prepared human scrum by centrifugation of clotted whole blood from normal donors. Serum was either used fresh or stored at -70°C until used. The amount of Fluosol used in the in vitro experiments approximated that administered in vivo except where indicated. An amount of 400 mL per mzproduced a plasma concentration of approximately 20 percent. For most
A R C 0 .IE F
800 r
2W.Y”
0
10
20
30 40 Time (minutes)
50
60
Ro. 1. Complement activation by Fluosol as measured by deposition of Qd on Fluosol. Fluosol was incubated in Serum in the pnsence (- 0 -) or absence (- 0 -) of EDTA. In the absence of EDTA. then was significant complement activation. Data points arc the mean of triplicate determinations with SD
x
Abbrevlatlons: MRI = magnetlc resonance Imaging; VBS = Vrronal-buffered rallne.
FLUOSOL-DA (Green Cross, Osaka, Japan [Fluosol, Alpha Therapeutics, Los Angeles, CAI) is a perfluorocarbon blood substitute that is undergoing phase I and I1 clinical trials in cancer patients to determine its efficacy when combined with radiation or chemotherapy. Other potential uses for perfluorocarbonsare in magnetic resonance imaging (MRI), transplant organ perfusion, drug delivery systems, and the treatment of shock.’ Recently, Fluosol was approved by the United States Food and Drug Administration for use in coronary angioplasty. Up to 30 percent of patients receiving Fluosol in the United States develop adverse reactions consisting of shortness of breath, chest tightness, and wheezing.* (However, Fluosol has been administered to a large number of patients in Japan without significant adverse effects3) On the basis of studies with human serum in , ~ adverse reactions have vitro and on animal ~ t u d i e sthe been attributed to the activation of the alternative path-
way of complement. In this study, we describe in vitro and in vivo experiments performed to define the mechanism by which whole Fluosol particles activate complement. We initially examined perfluorocarbons’ in vitro activation of complement by using a monoclonal antibody to C3d bound to the Fluosol particle and by assays for fluid-phase C3a des Arg (C3a) and C5a des Arg (C5a). To explore further the mechanism of the activation of complement on the surface of the Fluosol, we subjected proteins bound to the Fluosol to electrophoresis and identified them with Western blotting. The addition of red cells to the in vitro system during complement activation resulted in increased C3d binding to the red cells. This finding was applied to patient studies, and red cells were assayed before and after infusion of Fluosol for changes in C3d. Materials and Methods C3a and C5a assays We measured C3a and C5a by a radioimmunoassay system according to manufactur(Amersham, Arlington Heights, L), er’s instructions. In this system, C5 or C3 that docs not react is precipitatedout, and known amounts of antibody and labeled ligand are added. The reaction mixture is allowed to incubate and then centrifuged. Measurement of the radioactivity in the pellet allows calculation of the amount of labeled C5a or C3a. The amount of competitive inhibition of labeled ligand by unknown concentrations of C5a or C3a was compared to that produced by known standard amounts.
From the Department of Medicine, State University of New York at Buffalo; the Transfusion Services, Buffalo General Hospital; and the Veterans Administration Medical Center, Buffalo, New York. Supported in part by the Margaret Duffy and Robert Cameron Troup Memorial Fund for Cancer Research of the Buffalo General Hospital, the Research Service of the Veterans Administration Medical Center, the Ralph Hcchstetter Medical Research Fund in Honor of Dr. Henry C. and Bertha H. Buswell of the State University of New York at Buffalo, and Alpha Therapeutics. Received for publication January 11, 1991; revision received April 15, 1991, and accepted April 16, 1991.
642
‘IRANSNSION 1991-Vd. 31. No. 7
643
COMPLEMENT ACTIVATION BY FLUOSOL
Assay for C3d deposition on red cells or Fluosol We radiolabeled the murine monoclonal C3d antibody (Cytotech, San Diego, CA)with *=I using chloramine T and measured red cell- or Fluosol-bound C3d as described previou~ly.~ Briefly, either washed red cells (5% suspension by volume in barbital [Veronall-buffered saline [VBS], pH 7.4, with 0.15 mM [0.15 mmoVL] CaCI, and 0.5 mM [0.5 mmoVL] MgCl,) or Fluosol was incubated with radiolabeled anti-C3d that had been layered Over phthalate oils in microfuge tubes. At the end of the incubation, the tubes were spun for 10 minutes. We cut the tips and counted them in a gamma counter. The results were expressed in fg of anti-C3d bound per cell, per volume of Fluosol, or per total amount of Fluosol in a reaction mixture as described below.
Electroelution We performed electroelution of proteins bound to Fluosol by a modification of the methods used to study complement components bound to dialyzer membranes.6 Fluosol was incubated with serum for a specified time. Subsequently, the Fluosol was spun at 970 x g for 10 minutes and washed with VBS. After resuspension, we reduced the Fluosol with 6-mercaptocthanol for 15 minutes. The sample was layered Over a 5-percent sodium dodecyl sulfate stacking gel with a 10-percent SDS resolving gel by the Laemmli method. We performed Western blotting for 1 hour in 0.20 mM (0.20 mmol/L) Tris buffer with plyclonal goat antibodies that were specific for complement proteins of interest. We washed the gel three times with Tris buffer and incubated it for 1hour with alkaline phosphatase-labeled anti-goat antibody. We then washed the nitrocellulose with Tris buffer three more times and carried out the enzymatic reaction by using a substrate system (LaboratoriesPhosphatase, Kirkegaard and Perry, Gaithersburg, MD).
Nuclear MRI to detect the presence of Fluosol on cells We carried out MRI studies to determine the presence of Fluosol on red cells.’ With the MRI device (Model Fx270, Joel, Peabody, MA) set at a frequency of 253.75 mHz, we
scanned each sample for lo00 degradations. Using known concentrations of Fluosol, we obtained a standard curve correlating the height of the measured peak with the concentration of Fluosol present in each sample.
Patient studies All patient studies were done as approved by the institutional review board at the Veterans Administration Medical Center. All patients were being treated for metastatic adenocarcinoma to the liver, and they received Fluosol at a dosc of 400 mL per m2. After an initial test dose of 0.5 mL, all patients received Fluosol at an infusion rate of 1 mL per minute for the first 5 minutes, 5 mL per minute for the next 5 minutes, and then 10 mL per minute until completion of the total dose of Fluosol. Samples were obtained in EDTA tubes 30 minutes before the infusion of Fluosol and 30 minutes after completion of the infusion. After collection, the samples were immediately centrifuged and separated. Red cells were refrigerated and assayed the following day. We performed studies that showed that red cells obtained from patients after Fluosol infusion and assayed after 24- and &-hour storage did not differ in C3d content from red cells that were assayed immediately. Control patients received chemotherapy only. Their blood samples were obtained in a similar fashion.
In vitro studies of complement activation by Fluosol We prepared human scrum by centrifugation of clotted whole blood from normal donors. Serum was either used fresh or stored at -70°C until used. The amount of Fluosol used in the in vitro experiments approximated that administered in vivo except where indicated. An amount of 400 mL per mzproduced a plasma concentration of approximately 20 percent. For most
A R C 0 .IE F
800 r
2W.Y”
0
10
20
30 40 Time (minutes)
50
60
Ro. 1. Complement activation by Fluosol as measured by deposition of Qd on Fluosol. Fluosol was incubated in Serum in the pnsence (- 0 -) or absence (- 0 -) of EDTA. In the absence of EDTA. then was significant complement activation. Data points arc the mean of triplicate determinations with SD
