- Email: [email protected]
Effect of intensive plasmapheresis on the plasma cholesterol concentration with familial hypercholesterolemia
105
31 (1978) 105-115 0 Elsevier/North-Holland Scientific Publishers, Ltd.
Atherosclerosis,
EFFECT OF INTENSIVE PLASMAPHERESIS ON THE PLASMA CHOLESTEROL CONCENTRATION WITH FAMILIAL HYPERCHOLESTEROLEMIA
CARL S. APSTEIN *, DONALD B. ZILVERSMIT P.K. GEORGE Atherosclerosis and Clinical Research Cambridge, MA 02118 (U.S.A.)
Center,
**, ROBERT S. LEES and
Massachusetts
Institute
of Technology,
(Received 25 August, 1977) (Revised, received 24 July, 1978) (Accepted 31 July, 1978)
Summary Plasmapheresis was studied as a means of reducing the serum cholesterol concentration in 3 hypercholesterolemic patients who each underwent courses of intensive plasmapheresis with removal of 250-500 ml of plasma each day for 5-9 days. In one homozygous Type II patient, the serum cholesterol concentration decreased from 609 +_ 45 mg/lOO ml (mean + SEM) to 365 + 17 mg/lOO ml (40% decrease, P < 0.05) with two different courses of plasmapheresis. In the two other patients with non-homozygous hyperbetalipoproteinemia the serum cholesterol concentration decreased from 289 it 27 mg/ 100 ml to 205 + 19 mg/lOO ml (29% decrease, p < 0.05). After cessation of treatment, the cholesterol concentration returned to pre-treatment levels in IO-13 days in the homozygous patient and 7 days in one non-homozygous hyperbetalipoproteinemic patient; clofibrate (2 g/day) in this patient was associated with a smaller reduction of the cholesterol concentration with plasmapheresis and an increased rate of return of pre-treatment levels after plasmapheresis was stopped. Sustained plasmapheresis for 6 days in the other nonhomozygous hyperbetalipoproteinemic patient resulted in a new approximate “steady state” with a serum cholesterol concentration of 176-199 mg/lOO ml
Supported in part by NIH Grant HL-14209 for a Specialized Center of Research in Arteriosclerosis, and in part by General Clinical Research Center Grant RR88. * Dr. Apstein is a recipient of a Research Career Development Award from the N.I.H. * * Dr. Zilversmit is a Career Investigator of the American Heart Association. Present address: Cornell University, Division of Nutritional Sciences, 202 Savage Hill. Ithaca, NY 14850, U.S.A. Reprint requests should be sent to: Carl S. Apstein, M.D., Director, Cardiac Muscle Research Boston University School of Medicine, SO East Concord Street. Boston, MA 02118, U.S.A.
Lab,
compared with a pre-plasmapheresis value of 227 mg/l.OO ml. The response of the plasma cholesterol levels to plasmapheresis was subjected to kinetic analysis based on a current model of the regulation of lipoprotein metabolism. Key
words:
Cholesterol
~~ Hyperlipoproteinemia
Lipoproteins
Plasmapheresis
Introduction This study was undertaken to evaluate intensive plasmapheresis as a method of reducing the plasma cholesterol concentration and perturbing the steadystate cholesterol and lipoprotein levels and metabolism in order to study their kinetics. Type II hyperbetalipoproteinemia (familial hypercholesterolemia) is characterized by a high plasma concentration of cholestero1 and low density lipoprotein (LDL) and by premature atherosclerosis. Patients with homozygous familial hypercholesterolemia appear to have, at the same time, a reduction in the fractional catabolic rate of LDL and a 2-3-fold overproduction of LDL [l-4]. In heterozygous or mixed lipoprotein phenotypes of familial hypercholesterolemia normal rates of cholesterol and of LDL metabolism may occur, but at elevated plasma concentration [ 1,5]. The genetic mechanisms which lead to familial hypercholesterolemia have recently been reviewed [ 11; several sites appear to be involved in the ultimate determination of the circulating cholesterol and LDL level. In patients with familial hypercholesterolemia changes in plasma cholesterol concentration primarily reflect changes in LDL, as this protein carries most of the plasma cholesterol. Our data show that intensive plasmapheresis can lower the plasma cholesterol concentration; during the return of these concentrations to baseline levels after plasmapheresis is stopped the system is not in a “steady state” and the kinetics of cholesterol metabolism can be studied without the use of radioactive isotopes. A preliminary report of this study has appeared in abstract form [ 61. Methods Three patients with familial hypercholesterolemia were studied. One was a homozygous type IIa by tissue culture criteria [7] for a single dominant gene for hypercholesterolemia. The other two were presumably heterozygous or of a mixed lipoprotein phenotype. The patients were hospitalized as in-patients on the metabolic ward of the Massachusetts Institute of Technology Clinical Research Center. All hypolipidemic drugs were discontinued for 4 weeks prior to the start of plasmapheresis except in one patient (J.H.) in whom a second course of plasmapheresis was performed while the patient was taking clofibrate, 2 g/day. The patients were maintained on a 40% corn oil formula diet [8] for at least 5 weeks prior to plasmapheresis. Caloric consumption was adjusted to maintain the patients at constant weight during the period of study. Plasmapheresis was performed by collecting 500-600 ml of whole blood in a Fenwal Plasmapheresis Bag (Fenwal Laboratories, Division of Travenol Labora-
tories, Inc., Morton Grove, IL 60053) which contained 50 ml of ACD solution as anticoagulant. The citrate-whole blood mixture was centrifuged and about 350 ml of plasma was removed. The packed red cells were resuspended in a volume of saline equal to the volume of plasma which was removed and reinfused into the patient. Patient O.C. (see below) received human albumin in addition to saline when the packed red cells were reinfused during her second course of plasmapheresis. After the reinfusion of the packed red cells, a second 500--600 ml of blood was removed and the process repeated. The duration and periodicity of the course of intensive plasmapheresis varied slightly from patient to patient and is described for each case with the results of each course. Daily post-absorptive serum cholesterol and triglyceride determinations were performed as well as other laboratory tests on an individual basis. Cholesterol and triglyceride concentrations and lipoprotein phenotype were determined as described previously [ 91. A separate cholesterol determination was performed on each unit of plasma removed, as there was variable dilution by the ACD solution used as anticoagulant. The amount of cholesterol removed by plasmapheresis was calculated from the volume of plasma--citrate mixture removed, multiplied by its cholesterol concentration. Plasma volume was determined by the Evans’ blue dye dilution technique (Warner Lambert Corp., Morris Plains, NJ). Plasma volume was measured immediately prior to the start of the first 3-day course of plasmapheresis and after the completion of the 3-day course; the post-plasmapheresis plasma volume measurement was made at the time that the blood sample was drawn for the determination of the plasma cholesterol concentration. Because this was the time at which the plasma cholesterol concentration was the lowest, it was the most appropriate time to determine any dilution of plasma constituents due to an increase in plasma volume as a result of fluid shifts. The plasma volume determinations were performed on patients O.C. and J.H. before and after their first course of plasmapheresis. Subjects studied (1) O.C. was 16 years old at the time of study. She was documented to have receptor-negative homozygous familial hypercholesterolemia [ ‘71 by tissue culture of her skin fibroblasts, through the kindness of Drs. Michael Brown and Joseph Goldstein. She was known to have xanthomas since age 2, but had no symptoms of atherosclerosis at time of study. On examination, multiple tuberous xanthomas were seen on the dorsum of the hands, the finger folds, the elbows, knees and buttocks. Tendinous xanthomas were visible in the Achilles tendons and planar xanthomas on one palm. An early systolic murmur of aortic sclerosis was heard. All peripheral pulses were normal to palpation; bilateral systolic femoral bruits (2+/4+) were present. Prior to plasmapheresis the plasma cholesterol concentration was 689 mg/lOO ml, LDL cholesterol was 551 mg/lOO ml, HDL cholesterol was 26 mg/lOO ml, VLDL cholesterol was 112 mg/lOO ml and the triglyceride concentration was 83 mg/lOO ml. Lipoprotein electrophoresis demonstrated a type IIa pattern. (2) J.H. was a 40-year-old white female with documented hypercholesterolemia from the age of 28 when her serum cholesterol level was said to be 400 mg/lOO ml. At age 32, her serum cholesterol concentration was docu-
mented to be 500 mg/lOO ml. In the ensuing years, she was treated with thyroxine, caloric restriction, clofibrate, and colestipol, with variable reduction in the plasma cholesterol concentration. She gave a history of angina pectoris for approximately 6 months prior to the plasmapheresis study. Her ‘mother died of myocardial infarction at the age of 53 and a brother died at the age of 41 of his second myocardial infarction, a first such having occurred at the age of 29. One of the patient’s daughters also had hypercholesterolemia and a type II lipoprotein electrophoretic pattern. On examination, the patient had bilateral peri-orbital xanthelasmas but no tendon xanthomas. Prior to plasmapheresis and off drug therapy her plasma cholesterol concentration was 337 mg/lOO ml, LDL cholesterol was 247 mg/lOO ml, HDL cholesterol was 35 mg/lOO ml, VLDL cholesterol was 55 mg/lOO ml and her triglyceride concentration was 235 mg/lOO ml; she had a type II pattern on lipoprotein electrophoresis. (3) A.G., a 25-year-old black male had a history of xanthomas on both elbows and on the dorsum of both hands since adolescence, but no cardiovascular symptoms. Both parents were known to be hypercholesterolemic. On examination, tendon xanthomas were palpable in the fingers of both hands and in the Achilles tendons, with scars from xanthoma removal present bilaterally on the patellae and elbows. A faint arcus cornea was present bilaterally. His plasma cholesterol concentration on ad lib diet was 349 mg/lOO ml, LDL cholesterol was 286 mg/lOO ml, HDL cholesterol was 35 mg/lOO ml, and VLDL cholesterol was 28 mg/lOO ml and plasma triglyceride was 115 mg/lOO ml. Lipoprotein electrophoresis demonstrated a type II pattern. Patients J.H. and A.G. were considered to have familial hypercholesterolemia and hyperbetalipoproteinemia. They were clearly not homozygous and were presumably heterozygous or of a mixed phenotype lipoprotein family. This study was approved by the Committee on the Use of Humans as Experimental Subjects of the Massachusetts Institute of Technology Clinical Research Center. Results The results are summarized in Figures l-3. With homozygous patient O.C., plasmapheresis for 3 successive days caused a sharp drop in the plasma cholesterol concentration. During the first course of plasmapheresis, an average daily plasma removal of 642 ml resulted in the removal of 3.57, 2.93, and 2.59 g of cholesterol on 3 successive days so that 9.09 g cholesterol was removed from the circulating pool with a resultant decrease in plasma cholesterol concentration from 639 to 382 mg/lOO ml; with the second course of plasmapheresis an average daily plasma removal of 580 ml resulted in the removal of 3.80, 2.92, and 1.31 g of cholesterol on 3 successive days (total of 8.03 g of cholesterol) and reduced the plasma concentration from 578 to 348 mg/lOO ml. With cessation of plasmapheresis, the plasma cholesterol concentration returned towards baseline over the subsequent lo-13 days. Patient J.H. underwent two separate courses of plasmapheresis. No lipidlowering drugs were taken during the first course of plasmapheresis while the patient was maintained on a 40% corn oil formula diet [8]. An average of 550 ml of plasma were removed on each of 3 consecutive days with a resultant
P
1
‘ii
/
PLASMA COURSE
,).
f i
f
WERE% No
2
1 2
9
6
8
‘0
12
14
DAY
Fig. 1. Effect of plasmapheresis on plasma cholesterol in a homozygous patient decrease and return of the plasma cholesterol concentration after plasmapheresis with theoretical predictions baaed on equations 2 and 3. The arrows indicate the plasmapheresis. Abbreviations are those of equations 2 and 3 in text. The upper for the first course of plasmapheresis (Co = 639 mg/lOO ml. average Y = 0.159, panel shows the results of the second course of plaamapheresis for the first two plasmaphereses and v = 0.065 for plasmapheresis
OE.
(O.C.). The fractional is shown and compared time of each episode of panel showsthe results k = 0.061). The lower (Co = 578 mg/lOO ml, average v = 0.174 No. 3 of Course No. 2, k = 0.11).
PLAShfAPHERESfS
I
DAY
DAY
Fig. 2. Effect of plasmapheresis on plasma cholesterol in non-homozygous patient (J.H.). The format is the same as in Fig. 1. The upper panel shows the results for the first course of plasmapheresis (Co = 350 mg/lOO ml, average v = 0.144, k = 0.203) when the patient took no lipid-lowering drugs; the lower panel shows the results of the second course of plasmapheresis (Co = 286 mg/lOO ml. average v = 0.157. k = 0.582) when the patient was taking elofibrate, 2 g/day. Fig. 3. Effect of sustained plasmapheresis on the plasma cholesterol. Patient A.G.‘s response to sustained plasmapheresis is shown. The arrows indicate the time of each episode of plasmapheresis. On days 1, 2 and 3 an average of 517 ml of plasma was removed each day; on subsequent days an average daily plasma removal of 294 ml occurred.
removal of 1.60, 1.78 and 1.40 g of cholesterol each day (total of 4.78 g of circulating cholesterol); the plasma concentration decreased from 350 to 242 mg/ 100 ml. Return to pre-treatment levels occurred over the subsequent 7 days, The second course of plasmapheresis was performed after the patient had been taking clofibrate 2 g/day for one month; the drug was continued during the study. She was maintained on a low cholesterol solid diet during the second study. Removal of an average of 521 ml of plasma for 3 days resulted in the removal of 0.97, 1.35, and 1.21 g of cholesterol each day (total of 3.53 g of circulating cholesterol) and reduced the plasma concentration from 286 to 218 mg/lOO ml; return to pre-treatment levels occurred after approximately 3 days following cessation of plasmapheresis. A sustained course of plasmapheresis was performed in a patient with heterozygous type II hyperlipoproteinemia (A.G.) with removal of progressively smaller amounts of plasma in order to maintain the patient in a relative “steady-state” at a lower level of plasma cholesterol (Fig. 3). On an ad lib diet his plasma cholesterol concentration was 349 mg/lOO ml; in the month prior to plasmapheresis on the corn-oil formula diet the plasma cholesterol concentration decreased to 227 mg/lOO ml. First, a 3-day course of plasmapheresis was performed with an average removal of 517 ml of plasma each day, for a total removal of 3.54 g of cholesterol; the plasma cholesterol concentration decreased from 227 to 167 mg/lOO ml. Plasmapheresis was then performed on a daily basis for 6 more days in an attempt to keep the patient at this lower cholesterol concentration. Plasma was removed at an average rate of 294 ml/day with removal of approximately 600 mg of cholesterol per day. During this period the plasma cholesterol concentration remained between 176 and 199 mg/lOO ml. With cessation of plasmapheresis, the plasma cholesterol returned to the pre-treatment level in approximately 2 days. The plasma volume was relatively unaffected by the plasmapheresis process. In patient O.C. the pre-plasmapheresis plasma volume was 2565 ml and decreased to 2500 ml after the initial 3-day course of plasmapheresis. In patient J.H. the plasma volume decreased from 2075 to 1794 ml after the initial 3-day course of plasmapheresis. These relatively small changes in plasma volume would have the effect of increasing the measured plasma cholesterol concentration for a given circulating cholesterol mass; thus our observed decreases in plasma cholesterol concentration occurred despite a small contraction of the plasma compartment. The lack of a significant fluid flux into or out of the plasma space was also substantiated by the stability of the pre- and post-plasmapheresis hematocrit and serum albumin concentrations. Patient O.C.‘s pre-plasmapheresis hematocrit was 33. After the first day of plasmapheresis it was 33, after day 2 it was 34 and after day 3 it was 32.5. During patient O.C.‘s second plasmapheresis the hematocrit was 34 prior to plasmapheresis and was 31, 34 and 30 after the first, second and third days of plasmapheresis respectively. O.C.‘s serum albumin was not measured during her first course of plasmapheresis; the serum albumin concentration was 4.8 g/100 ml prior to the second course of plasmapheresis and was 4.6 g/100 ml after the 3 days of plasmapheresis; during this second course of plasmapheresis albumin was added to the packed red cellsaline reinfusion solution.
Patient J.H.‘s hematocrit was 40 prior to her initial course of plasmapheresis and increased to 43 after two days of plasmapheresis. It was not re-checked until 48 h after cessation of plasmapheresis when it had dropped to 32. The reason for the drop in hematocrit was not obvious, but appeared to be due in part at least to frequent blood sampling rather than to the plasmapheresis per se. During the second course of plasmapheresis, patient J.H.‘s pre-plasmapheresis hematocrit was 38; it was 41 after day 1, and 36 on the day after the plasmapheresis was completed. Prior to the first course of plasmapheresis J.H.‘s serum albumin concentration was 4.0 g/100 ml; on the day after the plasmapheresis was completed (the time when the plasma cholesterol concentration was at its minimum) the serum albumin concentration was 3.7 g/l00 ml. Prior to the second course of plasmapheresis J.H.‘s serum albumin was 4.6 g/100 ml, after the second course it was 4.3 g/100 ml. The plasma which was removed during the plasmapheresis process was replaced with normal saline without albumin in patient J.H. Patient A.G. did not have his plasma volume measurement performed. The pre-plasmapheresis hematocrit was 39 and serum albumin concentration was 4.5 g/100 ml. After the initial 3 days of intensive plasmapheresis the hematocrit was 36 and the serum albumin concentration was 4.1 g/100 ml. (The plasma removed by plasmapheresis was replaced by saline without albumin.) Over the subsequent 6 days of daily plasmapheresis the hematocrit progressively decreased to 32 at the time the plasmapheresis course was completed. There was no significant change in heart rate or blood pressure during or after the course of plasmapheresis in any of the patients. Discussion Our data indicate that intensive plasmapheresis can transiently, but substantially decrease the plasma cholesterol concentration. Cessation of plasmapheresis was associated with a prompt increase in the serum cholesterol concentration; however, when a 3-day course of intensive plasmapheresis was followed by a B-day course of daily plasmapheresis with removal of lesser amounts of plasma, a new and lowered “steady-state” level of the serum cholesterol concentration was achieved in patient A.G. We applied the kinetic model of Berlin and Schimke [lo] to our data. These authors studied the changes in tissue enzyme content in response to hormonal stimuli. They assumed that the transition from one steady state to another was accomplished by a process in which enzyme synthesis remained constant whereas the rate of enzyme degradation was governed by a first order catabolic reaction sequence. In analogy we assumed that plasmapheresis would not alter the production rate of plasma lipoprotein cholesterol but that the disappearance of plasma cholesterol at each instant would be proportional to the plasma concentration at that instant. The quantitative relationships are expressed as follows:
(1) where: C = concentration
of plasma cholesterol
(mg/ml);
S = production
rate of
plasma lipoprotein cholesterol (mg/ml/day); k = first order disappearance constant for plasma lipoprotein cholesterol. Equation 1 is compatible with the conclusions of Langer et al [ 31 namely that the synthetic rate for LDI, is constant, and that its removal rate is proportional to its concentration. The decrease in plasma cholesterol concentration with plasmapheresis can be predicted from the above equation. If v = fraction of the extracellular cholesterol pool * removed by plasmapheresis per day, then:
where C, is the initial plasma cholesterol concentration (prior to plasmapheresis) and C is the cholesterol concentration at time = t in days. The rate of return to pre-plasmapheresis baseline cholesterol concentration, after cessation of plasmapheresis, is predicted from the equation:
c/c, = g-)
-
[(&J--11’-“
or -kt
w&I
= (G&I)
-
I(G/C,)
-
11’
(3)
where CM is the minimum level of the plasma cholesterol concentration induced by plasmapheresis. From inspection of equations 2 and 3, it is apparent that, for a given value of v, the time-course of C/C, is determined by k. One purpose of this study was to compare the value of k calculated from the non-steady state response to plasmapheresis to the value of k independ@ntly determined by others using steady state measurements of LDL turnover. Previous studies of LDL kinetics in man have usually required radioactive isotope infusions and require that the patient be maintained in a steady state. The response to plasmapheresis has the potential of providing the same kinetic information and does not require the infusion of isotopes into the patient. We calculated the first order disappearance constant (k) from the response to plasmapheresis by putting equation 3 into logarithmic form: ln Co - ’ = -kt Co - CM A least squares regression analysis of the data was then performed. For patient O.C., during the return to baseline after the first plasmapheresis, k = 0.061 (r = 0.900). During the return to baseline after the second plasmapheresis, k = 0.11 (r = 0.857).
* The extracellular cholesterol pool confined to the intravascular compartment is probably close to the proportion of the extracellular LDL pool which is intravascular in type II individuals. This was found to be 68.4 + 4.3% in heterozygous type II patients 131. Therefore we calculated v in equation 2 as 110.684 X plasma volumes for our patients.
The reason for the difference in the rate of return to baseline after each course of plasmapheresis was not apparent. This patient’s plasma was replaced with normal saline after the first plasmapheresis and with saline and albumin after the second course of plasmapheresis. However, any decrease in serum albumin which might have occurred after the first course of plasmapheresis would have been expected to increase the synthesis of LDL, if one can extrapolate from experiments performed in rats where plasmapheresis extensive enough to decrease the plasma proteins was associated with an increase rate of lipoprotein synthesis [ll]. Thus a possible decrease in serum albumin concentration after the first course of plasmapheresis cannot account for a slower rate of post-plasmapheresis increase in the plasma cholesterol concentration relative to the second plasmapheresis course. We compared our values of k, derived from the response to plasmapheresis, to other reported values for k. Patient O.C. was also a subject in a study on the effect of hyperalimentation on lipid metabolism; using isotopic techniques described in that study [12], patient O.C.‘s steady-state k value was 0.110 f 0.009 [ 131. Simons et al. [ 41 have reported a similar value for k in homozygous type II patients. Thus our values of k derived from the response to plasmapheresis in patient O.C. (particularly after the second course of plasmapheresis) were similar to independently determined values for homozygotes. A least squares regression analysis of the rate of return to baseline in non-homozygous patient J.H. after plasmapheresis course no. 1, when the patient was taking no medications, resulted in a calculated k of 0.203 (r = 0.977); a group of 7 heterozygous type II patients were studied in the steady state using the methods noted above [12]. The mean value of k in these patients was 0.177 ? 0.050 [13]. Thus the value of k derived from the response to plasmapheresis in our heterozygote or mixed lipoprotein phenotype patient was comparable to the k value derived completely independently in a group of heterozygous in the steady state. A more graphic way of comparing the response to plasmapheresis and predictions from equations l-3 is presented in Figs. 1 and 2. In these figures we have plotted C/C, during and after plasmapheresis as determined in our patients. We have also plotted a theoretical C/C, curve using equations 2 and 3 and values of k as determined by a least-squares fit of the plasma cholesterol values during each post-plasmapheresis return to baseline as described above, utilizing the logarithmic form of equation 3. Figures 1 and 2 show relatively good agreement between the predicted decrease in the plasma cholesterol concentration with plasmapheresis (as predicted from equation 2 and the k values noted above) and the values observed in our patients. However, patient J.H. (Fig. 2) demonstrated a consistent oneday delay in the response of the plasma cholesterol to plasmapheresis. A contraction of the plasma volume as a result of plasmapheresis could have increased the plasma cholesterol concentration despite the removal of a significant amount of cholesterol. In patient J.H. the plasma volume decreased by approximately 10% over the first course of 3 days of plasmapheresis. Patient J.H.‘s hematocrit increased by 10% after the first day of each course of plasmapheresis. Thus the one day delay in the fall of the plasma cholesterol concentration may have been due to a transient decrease in the size of the plasma compartment in this patient.
The predicted rate of return to baseline after cessation of plasmapheresis (equation 3) was also in relatively good agreement with our observed curves in homozygous patient O.C. and in patient J.H. (Fig. 2) when she was taking no medications (upper panel); when she was taking clofibrate, the return to baseline was more rapid than predicted (lower panel), and the calculated value of k was 0.582 during the return to baseline. Thus, our data from these familial hypercholesterolemic patients in the nonsteady post-plasmapheresis state, in the absence of lipid-lowering drugs, correlated relatively closely with plasma cholesterol concentrations which were predicted by applying non-steady-state equations. Short-term plasmapheresis appears to be a useful tool for perturbing the steady-state plasma cholesterol concentration and may have applicability in the study of cholesterol and lipoprotein metabolism in a variety of disease states. The effect of sustained plasmapheresis was studied in patient A.G. This patient’s improvement on a controlled dietary regimen made him less than an ideal study candidate since his control, pre-plasmapheresis plasma cholesterol concentration was in the normal range at 227 mg/lOO ml. Nonetheless, an intensive 3-day course of plasmapheresis caused a further reduction to 167 mg/ 100 ml. Continued plasmapheresis maintained the cholesterol concentration at a lower level for 6 days. Thus, sustained plasmapheresis has the potential to lower the circulating cholesterol level on a more chronic basis, and can thereby create a new “steady-state” at a lower circulating cholesterol concentration. The effect of a lowered circulating lipoprotein level on lipoprotein kinetics could then be studied. If the lowered circulating cholesterol and lipoprotein level could be maintained on a long-term basis chronic plasma removal might have a therapeutic application. Experimental plasmapheresis in rats has been shown to lower the plasma cholesterol concentration 1111; however, cessation of plasmapheresis was associated with an increased rate of cholesterol synthesis and the post-plasmapheresis cholesterol concentration was 20% higher than the control level. In those experiments a relatively large amount of plasma was removed so that the rat’s total plasma protein concentration was reduced by approximately 50%. The post-plasmapheresis increase of the serum cholesterol concentration above baseline appeared to be related to an increased rate of protein and lipoprotein synthesis in response to the protein depletion caused by the plasmapheresis. In our patient study, plasmapheresis was not intensive enough to decrease the plasma albumin level significantly, and post-plasmapheresis circulating cholesterol levels did not exceed those of the control, pre-plasmapheresis state, except for patient J.H. when she was taking clofibrate. Plasmapheresis has previously been performed in patients with biliary cirrhosis in order to lower the plasma lipid levels [14]. Long-term plasmapheresis was associated with a significant reduction of the plasma cholesterol concentrations, regression of skin xanthomas, and relief of the painful neuropathy associated with this condition. In our study, there was no noticeable change in the xanthomas of the homozygous patient, nor in the auscultatory characteristics of the femoral artery bruits. However, it would be unlikely for such a change to occur over the short time span of our study. Plasmapheresis ]15] and plasma exchange [16] have also been performed in
patients with familial hypercholesterolemia. A transient reduction of the plasma cholesterol concentration was achieved, and relief of angina was claimed in one study [16]. In these studies patients were maintained on hypolipidemic medication during the plasmapheresis, and a kinetic analysis of the rise in serum cholesterol concentration after cessation of plasmapheresis was not performed. Recently, a technique of plasmapheresis has been combined with specific precipitation of LDL from the patient’s plasma, and the LDL-poor plasma has been returned to the patient [17]. This approach may obviate the problem of plasma protein depletion with sustained plasmapheresis and allow chronic plasmapheresis to become a possible mode of therapy for familial hypercholesterolemia. References 1 Brown, M.S. and Goldstein, J.L., Familial hypercholesterolernia ~~ A genetic defect in the low-density lipoprotein receptor, New Engl. J. Med., 294 (1976) 1386-1390. 2 Rilheimer. D.W., Goldstein, J.L., Grundy, S.M. et al.. Reduction in cholesterol and low density lipoprotein synthesis after portacaval shunt surgery in a patient with homozygous familial hypercholesterolemia, J. Clin. Invest., 56 (1975) 1420-1430. 3 Langer. T.. Strober, W. and Levy, R.I.. The metabolism of low density lipoprotein in familial Type II hyperlipoproteinemia, J. Clin. Invest.. 51 (1972) 1528-1536. 4 Simons, L.A.D., Reichl. D., Myant, N.B. et al.. The metabolism of the apoprotein of low density lipoprotein in familial hyperbetalipoproteinemia in the homozvgous form. Atherosclerosis, 21 (1975) 283--298. ~ A genetic regulatory defect in cho5 Goldstein, J.L. and Brown. MS., Familial hypercholesterolernia lesterol metabolism, Amer. J. Med., 58 (1975) 147-150. 6 Apstein, C.S., Zilversmit, D.B., George, P.K. et al., Reduction plasmapheresis. Clin. Res., 22 (1974) 311a. 7 Brown. M.S. and Goldstein, J.L., Receptor-mediated
control
of serum of cholesterol
cholesterol metabolism,
with intensive Science,
191
(1976) 7 50--154. 8 Ahrens, Jr., E.H., Dole, V.P. and Blankenhorn. D.H.. The use of orally-fed liquid formulas in mrlabolic studies, Amer. J. Clin. Nutr.. 2 (1954) 336. methods for plasma lipoprotein analysis, Adv. Lipid. Res., 6 9 Hatch, F.T. and Lees, R.S., Practical (1968) l-68. 10 Berlin. C.M. and Schmike, R.T., Influence of turnover rates on the response of enzymes to cortisonr. Mol. Pharmacol., 1 (1965) 149-156. 11 Takeuchi, N. and Yamamura, Y., The effect of plasmapheresis on cholesterol synthesis in the rat -12 13 14
15
16 17
Relationship to protein synthesis, J. Lab. Clin. Med., 79 (1972) 801-813. Torsvik, II., Feldman, H.A., Fischer, J.F. et al., Effect of intravenous hyperalimentation on plasma lipoproteins in severe familial hypercholesterolemia, Lancet, 1 (19’75) 601404. Torsvik, H., Feldman, H.A., Brook, J.G. et al., Kinetics of low-density lipoprotein metabolism in Tvpe II hyperlipoproteinemia. Submitted for publication. M.H. et al.. Plasmapheresis and plasma exchange in the Turnherg, L.A., Mahoney, M.P.. Gleeson. treatment of hyperlipemia and xanthomatous neuropathy in patients with primary biliary cirrhosis, Gut, 13 (1972) 976-981. DeGennes, J.L.. Touraine, R., Maunand, B. et al., Formes homozygotes cutanCo-tendineuses de xanthomatose hypercholest&olt%niquc dans une observation familiale exemplaire - Essai de plasrnapher&e a: titre de traitement heroique, Bull. M&n. Sot. M&d. HBp. Paris, 118 (1967) 1377-1402. Thompson, G.R., Lowenthal. R. and Myant, N.B., Plasma exchange in the management of homo~ygous familial hypercholesterolemia. Lancet, 1 (1975) 1208-1211. Lupirn, P.J., Moorjani, S. and Awad, .J., A new approach to the management of familial hypercholestcroicmia - Removal of plasma cholesterol based on the principle of affinity chromatography, Lancct. 1 (1976)
1261-1265.
31 (1978) 105-115 0 Elsevier/North-Holland Scientific Publishers, Ltd.
Atherosclerosis,
EFFECT OF INTENSIVE PLASMAPHERESIS ON THE PLASMA CHOLESTEROL CONCENTRATION WITH FAMILIAL HYPERCHOLESTEROLEMIA
CARL S. APSTEIN *, DONALD B. ZILVERSMIT P.K. GEORGE Atherosclerosis and Clinical Research Cambridge, MA 02118 (U.S.A.)
Center,
**, ROBERT S. LEES and
Massachusetts
Institute
of Technology,
(Received 25 August, 1977) (Revised, received 24 July, 1978) (Accepted 31 July, 1978)
Summary Plasmapheresis was studied as a means of reducing the serum cholesterol concentration in 3 hypercholesterolemic patients who each underwent courses of intensive plasmapheresis with removal of 250-500 ml of plasma each day for 5-9 days. In one homozygous Type II patient, the serum cholesterol concentration decreased from 609 +_ 45 mg/lOO ml (mean + SEM) to 365 + 17 mg/lOO ml (40% decrease, P < 0.05) with two different courses of plasmapheresis. In the two other patients with non-homozygous hyperbetalipoproteinemia the serum cholesterol concentration decreased from 289 it 27 mg/ 100 ml to 205 + 19 mg/lOO ml (29% decrease, p < 0.05). After cessation of treatment, the cholesterol concentration returned to pre-treatment levels in IO-13 days in the homozygous patient and 7 days in one non-homozygous hyperbetalipoproteinemic patient; clofibrate (2 g/day) in this patient was associated with a smaller reduction of the cholesterol concentration with plasmapheresis and an increased rate of return of pre-treatment levels after plasmapheresis was stopped. Sustained plasmapheresis for 6 days in the other nonhomozygous hyperbetalipoproteinemic patient resulted in a new approximate “steady state” with a serum cholesterol concentration of 176-199 mg/lOO ml
Supported in part by NIH Grant HL-14209 for a Specialized Center of Research in Arteriosclerosis, and in part by General Clinical Research Center Grant RR88. * Dr. Apstein is a recipient of a Research Career Development Award from the N.I.H. * * Dr. Zilversmit is a Career Investigator of the American Heart Association. Present address: Cornell University, Division of Nutritional Sciences, 202 Savage Hill. Ithaca, NY 14850, U.S.A. Reprint requests should be sent to: Carl S. Apstein, M.D., Director, Cardiac Muscle Research Boston University School of Medicine, SO East Concord Street. Boston, MA 02118, U.S.A.
Lab,
compared with a pre-plasmapheresis value of 227 mg/l.OO ml. The response of the plasma cholesterol levels to plasmapheresis was subjected to kinetic analysis based on a current model of the regulation of lipoprotein metabolism. Key
words:
Cholesterol
~~ Hyperlipoproteinemia
Lipoproteins
Plasmapheresis
Introduction This study was undertaken to evaluate intensive plasmapheresis as a method of reducing the plasma cholesterol concentration and perturbing the steadystate cholesterol and lipoprotein levels and metabolism in order to study their kinetics. Type II hyperbetalipoproteinemia (familial hypercholesterolemia) is characterized by a high plasma concentration of cholestero1 and low density lipoprotein (LDL) and by premature atherosclerosis. Patients with homozygous familial hypercholesterolemia appear to have, at the same time, a reduction in the fractional catabolic rate of LDL and a 2-3-fold overproduction of LDL [l-4]. In heterozygous or mixed lipoprotein phenotypes of familial hypercholesterolemia normal rates of cholesterol and of LDL metabolism may occur, but at elevated plasma concentration [ 1,5]. The genetic mechanisms which lead to familial hypercholesterolemia have recently been reviewed [ 11; several sites appear to be involved in the ultimate determination of the circulating cholesterol and LDL level. In patients with familial hypercholesterolemia changes in plasma cholesterol concentration primarily reflect changes in LDL, as this protein carries most of the plasma cholesterol. Our data show that intensive plasmapheresis can lower the plasma cholesterol concentration; during the return of these concentrations to baseline levels after plasmapheresis is stopped the system is not in a “steady state” and the kinetics of cholesterol metabolism can be studied without the use of radioactive isotopes. A preliminary report of this study has appeared in abstract form [ 61. Methods Three patients with familial hypercholesterolemia were studied. One was a homozygous type IIa by tissue culture criteria [7] for a single dominant gene for hypercholesterolemia. The other two were presumably heterozygous or of a mixed lipoprotein phenotype. The patients were hospitalized as in-patients on the metabolic ward of the Massachusetts Institute of Technology Clinical Research Center. All hypolipidemic drugs were discontinued for 4 weeks prior to the start of plasmapheresis except in one patient (J.H.) in whom a second course of plasmapheresis was performed while the patient was taking clofibrate, 2 g/day. The patients were maintained on a 40% corn oil formula diet [8] for at least 5 weeks prior to plasmapheresis. Caloric consumption was adjusted to maintain the patients at constant weight during the period of study. Plasmapheresis was performed by collecting 500-600 ml of whole blood in a Fenwal Plasmapheresis Bag (Fenwal Laboratories, Division of Travenol Labora-
tories, Inc., Morton Grove, IL 60053) which contained 50 ml of ACD solution as anticoagulant. The citrate-whole blood mixture was centrifuged and about 350 ml of plasma was removed. The packed red cells were resuspended in a volume of saline equal to the volume of plasma which was removed and reinfused into the patient. Patient O.C. (see below) received human albumin in addition to saline when the packed red cells were reinfused during her second course of plasmapheresis. After the reinfusion of the packed red cells, a second 500--600 ml of blood was removed and the process repeated. The duration and periodicity of the course of intensive plasmapheresis varied slightly from patient to patient and is described for each case with the results of each course. Daily post-absorptive serum cholesterol and triglyceride determinations were performed as well as other laboratory tests on an individual basis. Cholesterol and triglyceride concentrations and lipoprotein phenotype were determined as described previously [ 91. A separate cholesterol determination was performed on each unit of plasma removed, as there was variable dilution by the ACD solution used as anticoagulant. The amount of cholesterol removed by plasmapheresis was calculated from the volume of plasma--citrate mixture removed, multiplied by its cholesterol concentration. Plasma volume was determined by the Evans’ blue dye dilution technique (Warner Lambert Corp., Morris Plains, NJ). Plasma volume was measured immediately prior to the start of the first 3-day course of plasmapheresis and after the completion of the 3-day course; the post-plasmapheresis plasma volume measurement was made at the time that the blood sample was drawn for the determination of the plasma cholesterol concentration. Because this was the time at which the plasma cholesterol concentration was the lowest, it was the most appropriate time to determine any dilution of plasma constituents due to an increase in plasma volume as a result of fluid shifts. The plasma volume determinations were performed on patients O.C. and J.H. before and after their first course of plasmapheresis. Subjects studied (1) O.C. was 16 years old at the time of study. She was documented to have receptor-negative homozygous familial hypercholesterolemia [ ‘71 by tissue culture of her skin fibroblasts, through the kindness of Drs. Michael Brown and Joseph Goldstein. She was known to have xanthomas since age 2, but had no symptoms of atherosclerosis at time of study. On examination, multiple tuberous xanthomas were seen on the dorsum of the hands, the finger folds, the elbows, knees and buttocks. Tendinous xanthomas were visible in the Achilles tendons and planar xanthomas on one palm. An early systolic murmur of aortic sclerosis was heard. All peripheral pulses were normal to palpation; bilateral systolic femoral bruits (2+/4+) were present. Prior to plasmapheresis the plasma cholesterol concentration was 689 mg/lOO ml, LDL cholesterol was 551 mg/lOO ml, HDL cholesterol was 26 mg/lOO ml, VLDL cholesterol was 112 mg/lOO ml and the triglyceride concentration was 83 mg/lOO ml. Lipoprotein electrophoresis demonstrated a type IIa pattern. (2) J.H. was a 40-year-old white female with documented hypercholesterolemia from the age of 28 when her serum cholesterol level was said to be 400 mg/lOO ml. At age 32, her serum cholesterol concentration was docu-
mented to be 500 mg/lOO ml. In the ensuing years, she was treated with thyroxine, caloric restriction, clofibrate, and colestipol, with variable reduction in the plasma cholesterol concentration. She gave a history of angina pectoris for approximately 6 months prior to the plasmapheresis study. Her ‘mother died of myocardial infarction at the age of 53 and a brother died at the age of 41 of his second myocardial infarction, a first such having occurred at the age of 29. One of the patient’s daughters also had hypercholesterolemia and a type II lipoprotein electrophoretic pattern. On examination, the patient had bilateral peri-orbital xanthelasmas but no tendon xanthomas. Prior to plasmapheresis and off drug therapy her plasma cholesterol concentration was 337 mg/lOO ml, LDL cholesterol was 247 mg/lOO ml, HDL cholesterol was 35 mg/lOO ml, VLDL cholesterol was 55 mg/lOO ml and her triglyceride concentration was 235 mg/lOO ml; she had a type II pattern on lipoprotein electrophoresis. (3) A.G., a 25-year-old black male had a history of xanthomas on both elbows and on the dorsum of both hands since adolescence, but no cardiovascular symptoms. Both parents were known to be hypercholesterolemic. On examination, tendon xanthomas were palpable in the fingers of both hands and in the Achilles tendons, with scars from xanthoma removal present bilaterally on the patellae and elbows. A faint arcus cornea was present bilaterally. His plasma cholesterol concentration on ad lib diet was 349 mg/lOO ml, LDL cholesterol was 286 mg/lOO ml, HDL cholesterol was 35 mg/lOO ml, and VLDL cholesterol was 28 mg/lOO ml and plasma triglyceride was 115 mg/lOO ml. Lipoprotein electrophoresis demonstrated a type II pattern. Patients J.H. and A.G. were considered to have familial hypercholesterolemia and hyperbetalipoproteinemia. They were clearly not homozygous and were presumably heterozygous or of a mixed phenotype lipoprotein family. This study was approved by the Committee on the Use of Humans as Experimental Subjects of the Massachusetts Institute of Technology Clinical Research Center. Results The results are summarized in Figures l-3. With homozygous patient O.C., plasmapheresis for 3 successive days caused a sharp drop in the plasma cholesterol concentration. During the first course of plasmapheresis, an average daily plasma removal of 642 ml resulted in the removal of 3.57, 2.93, and 2.59 g of cholesterol on 3 successive days so that 9.09 g cholesterol was removed from the circulating pool with a resultant decrease in plasma cholesterol concentration from 639 to 382 mg/lOO ml; with the second course of plasmapheresis an average daily plasma removal of 580 ml resulted in the removal of 3.80, 2.92, and 1.31 g of cholesterol on 3 successive days (total of 8.03 g of cholesterol) and reduced the plasma concentration from 578 to 348 mg/lOO ml. With cessation of plasmapheresis, the plasma cholesterol concentration returned towards baseline over the subsequent lo-13 days. Patient J.H. underwent two separate courses of plasmapheresis. No lipidlowering drugs were taken during the first course of plasmapheresis while the patient was maintained on a 40% corn oil formula diet [8]. An average of 550 ml of plasma were removed on each of 3 consecutive days with a resultant
P
1
‘ii
/
PLASMA COURSE
,).
f i
f
WERE% No
2
1 2
9
6
8
‘0
12
14
DAY
Fig. 1. Effect of plasmapheresis on plasma cholesterol in a homozygous patient decrease and return of the plasma cholesterol concentration after plasmapheresis with theoretical predictions baaed on equations 2 and 3. The arrows indicate the plasmapheresis. Abbreviations are those of equations 2 and 3 in text. The upper for the first course of plasmapheresis (Co = 639 mg/lOO ml. average Y = 0.159, panel shows the results of the second course of plaamapheresis for the first two plasmaphereses and v = 0.065 for plasmapheresis
OE.
(O.C.). The fractional is shown and compared time of each episode of panel showsthe results k = 0.061). The lower (Co = 578 mg/lOO ml, average v = 0.174 No. 3 of Course No. 2, k = 0.11).
PLAShfAPHERESfS
I
DAY
DAY
Fig. 2. Effect of plasmapheresis on plasma cholesterol in non-homozygous patient (J.H.). The format is the same as in Fig. 1. The upper panel shows the results for the first course of plasmapheresis (Co = 350 mg/lOO ml, average v = 0.144, k = 0.203) when the patient took no lipid-lowering drugs; the lower panel shows the results of the second course of plasmapheresis (Co = 286 mg/lOO ml. average v = 0.157. k = 0.582) when the patient was taking elofibrate, 2 g/day. Fig. 3. Effect of sustained plasmapheresis on the plasma cholesterol. Patient A.G.‘s response to sustained plasmapheresis is shown. The arrows indicate the time of each episode of plasmapheresis. On days 1, 2 and 3 an average of 517 ml of plasma was removed each day; on subsequent days an average daily plasma removal of 294 ml occurred.
removal of 1.60, 1.78 and 1.40 g of cholesterol each day (total of 4.78 g of circulating cholesterol); the plasma concentration decreased from 350 to 242 mg/ 100 ml. Return to pre-treatment levels occurred over the subsequent 7 days, The second course of plasmapheresis was performed after the patient had been taking clofibrate 2 g/day for one month; the drug was continued during the study. She was maintained on a low cholesterol solid diet during the second study. Removal of an average of 521 ml of plasma for 3 days resulted in the removal of 0.97, 1.35, and 1.21 g of cholesterol each day (total of 3.53 g of circulating cholesterol) and reduced the plasma concentration from 286 to 218 mg/lOO ml; return to pre-treatment levels occurred after approximately 3 days following cessation of plasmapheresis. A sustained course of plasmapheresis was performed in a patient with heterozygous type II hyperlipoproteinemia (A.G.) with removal of progressively smaller amounts of plasma in order to maintain the patient in a relative “steady-state” at a lower level of plasma cholesterol (Fig. 3). On an ad lib diet his plasma cholesterol concentration was 349 mg/lOO ml; in the month prior to plasmapheresis on the corn-oil formula diet the plasma cholesterol concentration decreased to 227 mg/lOO ml. First, a 3-day course of plasmapheresis was performed with an average removal of 517 ml of plasma each day, for a total removal of 3.54 g of cholesterol; the plasma cholesterol concentration decreased from 227 to 167 mg/lOO ml. Plasmapheresis was then performed on a daily basis for 6 more days in an attempt to keep the patient at this lower cholesterol concentration. Plasma was removed at an average rate of 294 ml/day with removal of approximately 600 mg of cholesterol per day. During this period the plasma cholesterol concentration remained between 176 and 199 mg/lOO ml. With cessation of plasmapheresis, the plasma cholesterol returned to the pre-treatment level in approximately 2 days. The plasma volume was relatively unaffected by the plasmapheresis process. In patient O.C. the pre-plasmapheresis plasma volume was 2565 ml and decreased to 2500 ml after the initial 3-day course of plasmapheresis. In patient J.H. the plasma volume decreased from 2075 to 1794 ml after the initial 3-day course of plasmapheresis. These relatively small changes in plasma volume would have the effect of increasing the measured plasma cholesterol concentration for a given circulating cholesterol mass; thus our observed decreases in plasma cholesterol concentration occurred despite a small contraction of the plasma compartment. The lack of a significant fluid flux into or out of the plasma space was also substantiated by the stability of the pre- and post-plasmapheresis hematocrit and serum albumin concentrations. Patient O.C.‘s pre-plasmapheresis hematocrit was 33. After the first day of plasmapheresis it was 33, after day 2 it was 34 and after day 3 it was 32.5. During patient O.C.‘s second plasmapheresis the hematocrit was 34 prior to plasmapheresis and was 31, 34 and 30 after the first, second and third days of plasmapheresis respectively. O.C.‘s serum albumin was not measured during her first course of plasmapheresis; the serum albumin concentration was 4.8 g/100 ml prior to the second course of plasmapheresis and was 4.6 g/100 ml after the 3 days of plasmapheresis; during this second course of plasmapheresis albumin was added to the packed red cellsaline reinfusion solution.
Patient J.H.‘s hematocrit was 40 prior to her initial course of plasmapheresis and increased to 43 after two days of plasmapheresis. It was not re-checked until 48 h after cessation of plasmapheresis when it had dropped to 32. The reason for the drop in hematocrit was not obvious, but appeared to be due in part at least to frequent blood sampling rather than to the plasmapheresis per se. During the second course of plasmapheresis, patient J.H.‘s pre-plasmapheresis hematocrit was 38; it was 41 after day 1, and 36 on the day after the plasmapheresis was completed. Prior to the first course of plasmapheresis J.H.‘s serum albumin concentration was 4.0 g/100 ml; on the day after the plasmapheresis was completed (the time when the plasma cholesterol concentration was at its minimum) the serum albumin concentration was 3.7 g/l00 ml. Prior to the second course of plasmapheresis J.H.‘s serum albumin was 4.6 g/100 ml, after the second course it was 4.3 g/100 ml. The plasma which was removed during the plasmapheresis process was replaced with normal saline without albumin in patient J.H. Patient A.G. did not have his plasma volume measurement performed. The pre-plasmapheresis hematocrit was 39 and serum albumin concentration was 4.5 g/100 ml. After the initial 3 days of intensive plasmapheresis the hematocrit was 36 and the serum albumin concentration was 4.1 g/100 ml. (The plasma removed by plasmapheresis was replaced by saline without albumin.) Over the subsequent 6 days of daily plasmapheresis the hematocrit progressively decreased to 32 at the time the plasmapheresis course was completed. There was no significant change in heart rate or blood pressure during or after the course of plasmapheresis in any of the patients. Discussion Our data indicate that intensive plasmapheresis can transiently, but substantially decrease the plasma cholesterol concentration. Cessation of plasmapheresis was associated with a prompt increase in the serum cholesterol concentration; however, when a 3-day course of intensive plasmapheresis was followed by a B-day course of daily plasmapheresis with removal of lesser amounts of plasma, a new and lowered “steady-state” level of the serum cholesterol concentration was achieved in patient A.G. We applied the kinetic model of Berlin and Schimke [lo] to our data. These authors studied the changes in tissue enzyme content in response to hormonal stimuli. They assumed that the transition from one steady state to another was accomplished by a process in which enzyme synthesis remained constant whereas the rate of enzyme degradation was governed by a first order catabolic reaction sequence. In analogy we assumed that plasmapheresis would not alter the production rate of plasma lipoprotein cholesterol but that the disappearance of plasma cholesterol at each instant would be proportional to the plasma concentration at that instant. The quantitative relationships are expressed as follows:
(1) where: C = concentration
of plasma cholesterol
(mg/ml);
S = production
rate of
plasma lipoprotein cholesterol (mg/ml/day); k = first order disappearance constant for plasma lipoprotein cholesterol. Equation 1 is compatible with the conclusions of Langer et al [ 31 namely that the synthetic rate for LDI, is constant, and that its removal rate is proportional to its concentration. The decrease in plasma cholesterol concentration with plasmapheresis can be predicted from the above equation. If v = fraction of the extracellular cholesterol pool * removed by plasmapheresis per day, then:
where C, is the initial plasma cholesterol concentration (prior to plasmapheresis) and C is the cholesterol concentration at time = t in days. The rate of return to pre-plasmapheresis baseline cholesterol concentration, after cessation of plasmapheresis, is predicted from the equation:
c/c, = g-)
-
[(&J--11’-“
or -kt
w&I
= (G&I)
-
I(G/C,)
-
11’
(3)
where CM is the minimum level of the plasma cholesterol concentration induced by plasmapheresis. From inspection of equations 2 and 3, it is apparent that, for a given value of v, the time-course of C/C, is determined by k. One purpose of this study was to compare the value of k calculated from the non-steady state response to plasmapheresis to the value of k independ@ntly determined by others using steady state measurements of LDL turnover. Previous studies of LDL kinetics in man have usually required radioactive isotope infusions and require that the patient be maintained in a steady state. The response to plasmapheresis has the potential of providing the same kinetic information and does not require the infusion of isotopes into the patient. We calculated the first order disappearance constant (k) from the response to plasmapheresis by putting equation 3 into logarithmic form: ln Co - ’ = -kt Co - CM A least squares regression analysis of the data was then performed. For patient O.C., during the return to baseline after the first plasmapheresis, k = 0.061 (r = 0.900). During the return to baseline after the second plasmapheresis, k = 0.11 (r = 0.857).
* The extracellular cholesterol pool confined to the intravascular compartment is probably close to the proportion of the extracellular LDL pool which is intravascular in type II individuals. This was found to be 68.4 + 4.3% in heterozygous type II patients 131. Therefore we calculated v in equation 2 as 110.684 X plasma volumes for our patients.
The reason for the difference in the rate of return to baseline after each course of plasmapheresis was not apparent. This patient’s plasma was replaced with normal saline after the first plasmapheresis and with saline and albumin after the second course of plasmapheresis. However, any decrease in serum albumin which might have occurred after the first course of plasmapheresis would have been expected to increase the synthesis of LDL, if one can extrapolate from experiments performed in rats where plasmapheresis extensive enough to decrease the plasma proteins was associated with an increase rate of lipoprotein synthesis [ll]. Thus a possible decrease in serum albumin concentration after the first course of plasmapheresis cannot account for a slower rate of post-plasmapheresis increase in the plasma cholesterol concentration relative to the second plasmapheresis course. We compared our values of k, derived from the response to plasmapheresis, to other reported values for k. Patient O.C. was also a subject in a study on the effect of hyperalimentation on lipid metabolism; using isotopic techniques described in that study [12], patient O.C.‘s steady-state k value was 0.110 f 0.009 [ 131. Simons et al. [ 41 have reported a similar value for k in homozygous type II patients. Thus our values of k derived from the response to plasmapheresis in patient O.C. (particularly after the second course of plasmapheresis) were similar to independently determined values for homozygotes. A least squares regression analysis of the rate of return to baseline in non-homozygous patient J.H. after plasmapheresis course no. 1, when the patient was taking no medications, resulted in a calculated k of 0.203 (r = 0.977); a group of 7 heterozygous type II patients were studied in the steady state using the methods noted above [12]. The mean value of k in these patients was 0.177 ? 0.050 [13]. Thus the value of k derived from the response to plasmapheresis in our heterozygote or mixed lipoprotein phenotype patient was comparable to the k value derived completely independently in a group of heterozygous in the steady state. A more graphic way of comparing the response to plasmapheresis and predictions from equations l-3 is presented in Figs. 1 and 2. In these figures we have plotted C/C, during and after plasmapheresis as determined in our patients. We have also plotted a theoretical C/C, curve using equations 2 and 3 and values of k as determined by a least-squares fit of the plasma cholesterol values during each post-plasmapheresis return to baseline as described above, utilizing the logarithmic form of equation 3. Figures 1 and 2 show relatively good agreement between the predicted decrease in the plasma cholesterol concentration with plasmapheresis (as predicted from equation 2 and the k values noted above) and the values observed in our patients. However, patient J.H. (Fig. 2) demonstrated a consistent oneday delay in the response of the plasma cholesterol to plasmapheresis. A contraction of the plasma volume as a result of plasmapheresis could have increased the plasma cholesterol concentration despite the removal of a significant amount of cholesterol. In patient J.H. the plasma volume decreased by approximately 10% over the first course of 3 days of plasmapheresis. Patient J.H.‘s hematocrit increased by 10% after the first day of each course of plasmapheresis. Thus the one day delay in the fall of the plasma cholesterol concentration may have been due to a transient decrease in the size of the plasma compartment in this patient.
The predicted rate of return to baseline after cessation of plasmapheresis (equation 3) was also in relatively good agreement with our observed curves in homozygous patient O.C. and in patient J.H. (Fig. 2) when she was taking no medications (upper panel); when she was taking clofibrate, the return to baseline was more rapid than predicted (lower panel), and the calculated value of k was 0.582 during the return to baseline. Thus, our data from these familial hypercholesterolemic patients in the nonsteady post-plasmapheresis state, in the absence of lipid-lowering drugs, correlated relatively closely with plasma cholesterol concentrations which were predicted by applying non-steady-state equations. Short-term plasmapheresis appears to be a useful tool for perturbing the steady-state plasma cholesterol concentration and may have applicability in the study of cholesterol and lipoprotein metabolism in a variety of disease states. The effect of sustained plasmapheresis was studied in patient A.G. This patient’s improvement on a controlled dietary regimen made him less than an ideal study candidate since his control, pre-plasmapheresis plasma cholesterol concentration was in the normal range at 227 mg/lOO ml. Nonetheless, an intensive 3-day course of plasmapheresis caused a further reduction to 167 mg/ 100 ml. Continued plasmapheresis maintained the cholesterol concentration at a lower level for 6 days. Thus, sustained plasmapheresis has the potential to lower the circulating cholesterol level on a more chronic basis, and can thereby create a new “steady-state” at a lower circulating cholesterol concentration. The effect of a lowered circulating lipoprotein level on lipoprotein kinetics could then be studied. If the lowered circulating cholesterol and lipoprotein level could be maintained on a long-term basis chronic plasma removal might have a therapeutic application. Experimental plasmapheresis in rats has been shown to lower the plasma cholesterol concentration 1111; however, cessation of plasmapheresis was associated with an increased rate of cholesterol synthesis and the post-plasmapheresis cholesterol concentration was 20% higher than the control level. In those experiments a relatively large amount of plasma was removed so that the rat’s total plasma protein concentration was reduced by approximately 50%. The post-plasmapheresis increase of the serum cholesterol concentration above baseline appeared to be related to an increased rate of protein and lipoprotein synthesis in response to the protein depletion caused by the plasmapheresis. In our patient study, plasmapheresis was not intensive enough to decrease the plasma albumin level significantly, and post-plasmapheresis circulating cholesterol levels did not exceed those of the control, pre-plasmapheresis state, except for patient J.H. when she was taking clofibrate. Plasmapheresis has previously been performed in patients with biliary cirrhosis in order to lower the plasma lipid levels [14]. Long-term plasmapheresis was associated with a significant reduction of the plasma cholesterol concentrations, regression of skin xanthomas, and relief of the painful neuropathy associated with this condition. In our study, there was no noticeable change in the xanthomas of the homozygous patient, nor in the auscultatory characteristics of the femoral artery bruits. However, it would be unlikely for such a change to occur over the short time span of our study. Plasmapheresis ]15] and plasma exchange [16] have also been performed in
patients with familial hypercholesterolemia. A transient reduction of the plasma cholesterol concentration was achieved, and relief of angina was claimed in one study [16]. In these studies patients were maintained on hypolipidemic medication during the plasmapheresis, and a kinetic analysis of the rise in serum cholesterol concentration after cessation of plasmapheresis was not performed. Recently, a technique of plasmapheresis has been combined with specific precipitation of LDL from the patient’s plasma, and the LDL-poor plasma has been returned to the patient [17]. This approach may obviate the problem of plasma protein depletion with sustained plasmapheresis and allow chronic plasmapheresis to become a possible mode of therapy for familial hypercholesterolemia. References 1 Brown, M.S. and Goldstein, J.L., Familial hypercholesterolernia ~~ A genetic defect in the low-density lipoprotein receptor, New Engl. J. Med., 294 (1976) 1386-1390. 2 Rilheimer. D.W., Goldstein, J.L., Grundy, S.M. et al.. Reduction in cholesterol and low density lipoprotein synthesis after portacaval shunt surgery in a patient with homozygous familial hypercholesterolemia, J. Clin. Invest., 56 (1975) 1420-1430. 3 Langer. T.. Strober, W. and Levy, R.I.. The metabolism of low density lipoprotein in familial Type II hyperlipoproteinemia, J. Clin. Invest.. 51 (1972) 1528-1536. 4 Simons, L.A.D., Reichl. D., Myant, N.B. et al.. The metabolism of the apoprotein of low density lipoprotein in familial hyperbetalipoproteinemia in the homozvgous form. Atherosclerosis, 21 (1975) 283--298. ~ A genetic regulatory defect in cho5 Goldstein, J.L. and Brown. MS., Familial hypercholesterolernia lesterol metabolism, Amer. J. Med., 58 (1975) 147-150. 6 Apstein, C.S., Zilversmit, D.B., George, P.K. et al., Reduction plasmapheresis. Clin. Res., 22 (1974) 311a. 7 Brown. M.S. and Goldstein, J.L., Receptor-mediated
control
of serum of cholesterol
cholesterol metabolism,
with intensive Science,
191
(1976) 7 50--154. 8 Ahrens, Jr., E.H., Dole, V.P. and Blankenhorn. D.H.. The use of orally-fed liquid formulas in mrlabolic studies, Amer. J. Clin. Nutr.. 2 (1954) 336. methods for plasma lipoprotein analysis, Adv. Lipid. Res., 6 9 Hatch, F.T. and Lees, R.S., Practical (1968) l-68. 10 Berlin. C.M. and Schmike, R.T., Influence of turnover rates on the response of enzymes to cortisonr. Mol. Pharmacol., 1 (1965) 149-156. 11 Takeuchi, N. and Yamamura, Y., The effect of plasmapheresis on cholesterol synthesis in the rat -12 13 14
15
16 17
Relationship to protein synthesis, J. Lab. Clin. Med., 79 (1972) 801-813. Torsvik, II., Feldman, H.A., Fischer, J.F. et al., Effect of intravenous hyperalimentation on plasma lipoproteins in severe familial hypercholesterolemia, Lancet, 1 (19’75) 601404. Torsvik, H., Feldman, H.A., Brook, J.G. et al., Kinetics of low-density lipoprotein metabolism in Tvpe II hyperlipoproteinemia. Submitted for publication. M.H. et al.. Plasmapheresis and plasma exchange in the Turnherg, L.A., Mahoney, M.P.. Gleeson. treatment of hyperlipemia and xanthomatous neuropathy in patients with primary biliary cirrhosis, Gut, 13 (1972) 976-981. DeGennes, J.L.. Touraine, R., Maunand, B. et al., Formes homozygotes cutanCo-tendineuses de xanthomatose hypercholest&olt%niquc dans une observation familiale exemplaire - Essai de plasrnapher&e a: titre de traitement heroique, Bull. M&n. Sot. M&d. HBp. Paris, 118 (1967) 1377-1402. Thompson, G.R., Lowenthal. R. and Myant, N.B., Plasma exchange in the management of homo~ygous familial hypercholesterolemia. Lancet, 1 (1975) 1208-1211. Lupirn, P.J., Moorjani, S. and Awad, .J., A new approach to the management of familial hypercholestcroicmia - Removal of plasma cholesterol based on the principle of affinity chromatography, Lancct. 1 (1976)
1261-1265.