The effects of plasma exchange on cholesterol metabolism

195

Atherosclerosis, 31 (1978) 195-204 0 Elsevier/North-Holland Scientific Publishers, Ltd.

THE EFFECTS METABOLISM

OF PLASMA EXCHANGE

ON CHOLESTEROL

L.A. SIMONS, J. COREY GIBSON, J.P. ISBISTER and J.C. BIGGS Lipid Clinic and Haematology Department, University of NSW, Sydney (Australia)

St. Vincent’s Hospital and School of Medicine,

(Received 3 April, 1978) (Revised, received 20 June, 1978) (Accepted 22 June, 1978)

Summary Four patients heterozygous for familial hypercholesterolaemia were treated by repeated plasma exchange with or without lipid-lowering drugs. Repeated plasma exchange without drug therapy in 3 patients was associated with a significant l&-28% decrement in plasma cholesterol level, comparing control with plateau values observed 3 weeks after exchange. Further decrements in plateau values followed the addition of lipid-lowering drugs used in combination, clofibrate-nicotinic acid or clofibrate-nicotinic acid-cholestyramine (range of total decrement 39-5076). Plasma exchange was associated with an increased excretion of endogenous faecal steroids, but this increase was completely abolished by the subsequent administration of clofibrate-nicotinic acid. This therapy prevented any increase in bile acid excretion with concomitant use of cholestyramine resin. Plasma exchange with drug therapy was associated with a sustained rise in plasma cholesterol specific radioactivity. In a fourth patient, clofibrate-nicotinic acid was administered prior to plasma exchange and led to a 24% fall in plasma cholesterol. Subsequent plasma exchange in this patient produced no sustained change in plasma cholesterol plateau level. In two patients, withdrawal of drugs allowed plasma cholesterol to return to preexchange control levels. These observations suggest that plasma exchange probably produced an increase in endogenous cholesterol synthesis and a mobilisation of tissue cholesterol. In relation to plateau cholesterol values 3 weeks after an exchange, the data suggested that the reduction in plasma cholesterol level with plasma exchange and drug therapy could have been achieved by intensive drug therapy alone. Address for correspondence: Dr. L.A. Simons, Medical Professiorial Unit, St. Vincent’s Hospital, Darliwhurst, NSW 2010. Australia. This study was supported in part by grants from the National Heart Foundation of Australia and the National Health and Medical Research Council of Australia.

Clofibrate -- Faecal steroids ucid --Plasma c.rchar~gc

Key words:

Familial

hypcrcholesterolacrrlia

.- ~l’icwtinic

Introduction Repeated plasma exchange using the continuous flow blood cell separator has been suggested as a form of management for severe and resistant hypercholesterolaemia, particularly familial hypercholesterolaemia (FH) in the homozygous form [ 11. The more commonly seen heterozygotes for FH frequently have marked degrees of hypercholesterolaemia (> 13 mmol/l) which may also be resistant to conventional diet and drug therapy. Any acute fall in plasma cholesterol concentration with plasma exchange would be followed by compensatory changes, such as an increase in cholesterol and lipoprotein synthesis. However, this is difficult to document in the unsteady state [2]. The concomitant use of lipid-lowering drugs might prove therapeutically desirable, and might provide further insight into the metabolic effects of plasma exchange. This paper describes a two year evaluation of the effects of repeated plasma exchange with or without lipid-lowering drugs on cholesterol metabolism in four patients with FH in the heterozygous form. Methods Patients Clinical, biochemical and angiographic presented in Table 1. All patients appeared erozygous form, as judged by the presence increased levels of low density lipoprotein mata, premature coronary atheroscIerosis TABLE

data for the 4 patients studied are to be suffering from FH in the hetof marked hypercholesterolaemia, (LDL) cholesterol, tendon xanthoand similarly affected first degree

1

SUMMARY

OF

PRE-TREATMENT

CLINICAL

DATA

a -

Patient

Age

Sex

wt

RBW

Chol.

Trig.

(kg)

(%)

(mmol/l)

(mmolll)

Plasma 35

M

70

93

C0r0n.

Clinical

Angio.

LDL

11.1

9.4

1.6

Tendon

X,

C.A.D.

(2)

X.

C.A.D.

(3)

X.

C.A.D.

(3)

X,

N.D.

asymptomatic 55

F

66

116

13.9

12.1

1.5

Tendon angina

57

F

52

92

11.0

9.5

1.1

Tendon angina,

M . infarct 30

M

79

108

10.3

8.8

1.2

Tendon

asymptomatic a Abbreviations: nary done.

angiography;

RBW.

relative

C.A.D.

(3).

body

weight;

coronary

Chol., artery

Cholesterol; disease

(3

Trig., major

Triglycerides;

vessels);

Coron.

X = xanthomata;

Angio., N.D.

core= not

relatives. All patients had been diagnosed some years previously and had been consuming typical cholesterol lowering diets (polyunsat/sat >1.5 and <300 mg cholesterol/day) plus, at various times, clofibrate, cholestyramine or nicotinic acid. These diets were continued during the present study but, with the exception of Patient 4, drug therapy was suspended 2-3 months before commencement of plasma exchange. Details of subsequent drug therapy are presented in the Results. All studies were performed on an outpatient basis. There were no significant changes in body weight throughout the studies. Plasma

exchange

At least 6 weeks before the commencement of plasma exchange, Patients 1-3 received an intravenous injection of 50 PCi of [4-‘4C]cholesterol (Day 0) 131. A follow-up injection of 250 ,uCi of [ la,2a(n)-3H]cholesterol was made 9-12 months later, at a time when plasma i4C activities had become quite low. Informed consent was obtained prior to the performance of these and other procedures. Plasma exchange was performed with the continuous flow blood cell separator (Aminco Celltrifuge@) as described by Thompson et al. [ 11, with minor modification. Patients were usually exchanged in the morning, following a full breakfast. 3-4 litres of plasma were removed from each patient and this was replaced progressively by an equivalent amount of Hartmann’s solution and 25% human albumin. Plasma exchange was performed at intervals of 3-4 weeks until introduction of combination drug therapy, after which time it was performed much less frequently. 400 ml of fresh frozen plasma was infused at the termination of the first 13 exchanges in Patients 1-3 to replace clotting factors, but this procedure was subsequently abandoned without untoward effects. Analytical

methods

Plasma samples were collected approximately weekly, largely in the nonfasting state. Faecal collections were obtained in Patients 1 and 2 over intervals of several weeks, as indicated in Figs. 1 and 2. They were analysed in 5-6 day pools. Plasma cholesterol and triglyceride concentrations were measured on the Autoanalyser@ after isopropanol extraction (Technicon methods N-24a and N-78 respectively). Plasma lipoproteins were isolated by sequential density adjustment [ 41. Specific radioactivity of total plasma cholesterol 14C and/or 3H was measured in a Folch extract, according to Simons and Myant [ 51. Faecal excretion of acidic steroids was measured by gas-liquid chromatography and endogenous neutral steroids by the isotopic balance method, as previously described [ 51. Chromic oxide was used as a metabolic marker for variations in faecal flow and recovery, and doses of [ 3H]beta-sitosterol were used at intervals to assess degradation of neutral steroids in the gut [5]. Materials

All radiochemicals were obtained from the Radiochemical Centre, Amersham, U.K. Lipid lowering drugs were generously donated as follows: clofibrate (Atromid-S@), ICI Australia Ltd; cholestyramine (Questran@), Mead Johnson Pty Ltd; and nicotinic acid (Nicangin@), Astra Chemicals Pty Ltd.

Fig.

1. Metabolic

responses

to plasma

exchange

(star)

and drug

therapy

in Patient

1.

Results During a single plasma exchange, cholesterol levels fall to quite low values (e.g. 5-6 mmol/l). This is followed by a rapid rise over the next few days until equilibrium or plateau values are reached, usually within 2-3 weeks. These changing levels complicate the evaluation of such therapy, particularly in comparison with chronic drug therapy where the effects are more stable. The most accurate expression of the phenomenon would be the integrated area under the curve of plasma cholesterol concentration versus time. For simplicity we have chosen to express equilibrium values reached 3 weeks after an exchange, often at the commencement of the next exchange. Repeated plasma exchange was a well tolerated clinical procedure and patients were able to return home 2 h after its conclusion. Hypolipidaemic drug therapy was introduced after 8 or 9 exchanges in Patients 1-3, when it appeared that even lower equilibrium plasma cholesterol levels were desirable. Plasma exchange was suspended after a further 7-10 visits and combination drug therapy was continued. In the case of Patient 4, combination therapy with

199

TABLE

2

EFFECTS ____~_

OF

TREATMENT

Patient

No.

of

Nrt

exchanges

chol.

Initial

Final level

chol.

level

a

b

removed

chol.

(R)

(mmol/l)

WmolP)

1

18

130

11.1

6.8-7.8

2

15

105

13.9

7.0-7.8

3

16

102

11.0

5.7-6.5

27

7.8

7.5-7.7

.5

4 a Before

plasma

exchange.

b Range

observed

after

plasma

exchange

and drug

therapy

(equilibrium

values).

nicotinic acid and clofibrate was fully explored before commencing plasma exchange. The overall effects of plasma exchange and combination drug therapy are summarised in Table 2. More than 100 g of cholesterol were removed from the plasma pool in Patients l-3 by 15-18 exchanges. Important reductions in plasma cholesterol concentration were achieved with exchange and combination drug therapy. After 5 exchanges in Patient 4 and removal of 27 g of cholesterol, no lowering of plasma cholesterol equilibrium value was found beyond that observed with drug therapy alone. Plasma exchange alone produced no sustained change in plasma triglyceride levels, although these were slightly lower with nicotinic acid and clofibrate therapy. A detailed consideration of the individual patient responses provides further metabolic information. Patient 1 Experimental findings in Patient 1 are presented in Fig. 1 and Table 3. Plasma exchange alone produced a fall in plasma cholesterol level from 11.1 to an equilibrium value of 8.6 mmol/l, but this full effect was not sustained. After

TABLE

3

ENDOGENOIJS THERAPY

FAECAL

STEROID

EXCRETION

FOLLOWING

PLASMA

EXCHANGE

AND

DRUG

a _

Patient

Control

Plasma

Clofibrate

Clofibrate

exchange

nicotinic

+

Clofibrate

+

acid

nicotinic

acid

cholestyramine ___ 1

Neutral

598

r 37

679

r 100

Acidic

320

f 22

469

+

Total

918

f 46

1148

(3.17)

2

*

293

+ 168

*

1083

(8.46)

Neutral

291

f 39

408i

Acidic

297

+ 32

480

588

f 51

Total

79Ok

89

(5,29)

1512

50

168

i

46

**

-

319

i

48

**

-

f 115

**

? 148

**

64)

98

24**

285

+

23

728

t

17

(4,

**

23)

Drug

5 **

+

dosage:

Difference

(Number clofibrate

between

of pools, 2 g/d:

duration

of collection

cholestyramine

16 g/d:

treatmentandcontrolmeanscompared

4**

i 125

247

k 46

522

+ 127

336

t 49

**

(3-11) -_____

___ f 1 SD

89+

424 (3,14)

___~__ a Mran

17**

(4.15)

443+

58** ? 128

-

95*

(3,14)

889 (11.

t

in days). nicotinic

acid

6 g/d

by Student’s

Pt.

t-test:

1, 9 g/d **

Pt. 2.

P < 0.01;

* P \ 0.05.

+

200

8 exchanges, clofibrate 2 g/day was introduced on Day 206, but plasma cholesterol values improved little further during the next, 147 days. Nicotinic acid 4-6 g/d by itself also produced no further decrement in plasma (,holesterol level, but the combination of nicotinic acid and clofibrate from Day 437 was associated with a significantly lower cholesterol range (6.8--7.8 mmoi/l). Exchanging was suspended from Day 540, and plasma cholesterol values did not change appreciably. Plasma exchange alone was associated with a significant increase in faecal excretion of acidic steroids (Fig. 1, Table 3). The addition of clofibrate significantly increased the output of endogenous neutral steroids while returning the bile acid output to control values. Combination of clofibrate and nicotinic acid significantly suppressed the output of neutral steroids and bile acids below control values. The response of total plasma cholesterol specific radioactivity to plasma exchange was variable (Fig. 1). In some instances an acute fall was observed, coinciding with the infusion of non-radioactive fresh frozen plasma. After several exchanges, a rise in specific activity was noted. After Day 437, in combination with clofibrate and nicotinic acid therapy and in the presence of a sustained fall in plasma cholesterol level, plasma exchange was associated with a prominent rise in specific radioactivity lasting for at least 3 weeks. This was again observed after an exchange on Day 540. Patient 2 Experimental findings in Patient 2 are presented in Fig. 2 and Table 3. The results were qualitatively similar to those observed in Patient 1 and will only be described briefly. Plasma exchange alone produced a significant fall in plasma cholesterol level (13.9 to 10 mmol, equilibrium value). Subsequent therapy with clofibrate 2 g/d or nicotinic acid up to 9 g/d produced only a marginal change in plasma cholesterol level. The combination of clofibrate and nicotinic acid, and the later addition of cholestyramine as well, produced lower cholesterol values (7.0-7.8 mmol/l). The plasma cholesterol level immediately rose to control values when drug therapy was stopped, but decreased again quite quickly when started again. In Patient 2, plasma exchange alone significantly increased the excretion of acidic and neutral steroids. As in Patient 1, clofibrate abolished the increased bile acid excretion, while sustaining an increased output of neutral steroids. Clofibrate and nicotinic acid were associated with a significant reduction in the output of neutral steroids and this pattern remained unchanged with the addition of cholestyramine. Total plasma cholesterol specific radioactivity behaviour was similar to Patient 1, with prolonged rises in activity being observed after plasma exchange, in the presence of nicotinic acid 6-9 g/d (Fig. 2). No diagnostic changes in specific radioactivity were observed when drug therapy was stopped or started again. Patient 3 Experimental findings in Patient 3 are presented in Fig. 3. As in the earlier patients, plasma exchange alone produced a significant fall in plasma choles-

201

+ 5%I I

+++ I%+

I -”

+

i

Fig.

2. Metabolic

responses

to plasma

exchange

(star

or plus

sign)

and drug

therapy

in Patient

2.

Fig.

3. Metabolic

responses

to plasma

exchange

(star

or plus

sign)

and drug

therapy

in Patient

3.

DATE

Fig. 4. Plasma cholesterol

values during drug therapy

DAIS

and plasma exchange

(plus sign) in Patient

4.

terol level (11-9 mmol/l, equilibrium value). The use of clofibrate (2 g/d) or nicotinic acid (3 g/d) failed to produce a further fall in the level of plasma cholesterol, but nicotinic acid (6 g/d) and, latterly, nicotinic acid plus clofibrate were associated with a lower plasma cholesterol level, in the range 5.7-6.5 mmol/l. Plasma cholesterol rapidly climbed to a level in excess of control values when drug therapy was stopped, but again decreased with the combination of clofibrate and nicotinic acid. Total plasma cholesterol specific radioactivity exhibited a prominent rise after plasma exchange, in the presence of clofibrate and nicotinic acid therapy and a lower plasma cholesterol concentration. Patient 4 Experimental findings in Patient 4 are presented in Fig. 4. An initial plasma cholesterol level of 10.3 mmol/l was uninfluenced by clofibrate alone (earlier observations not presented) nor by nicotinic acid (9 g/d). Combination of nicotinic acid and clofibrate suppressed the plasma cholesterol level to 7.8 mmol/l, at which point plasma exchanging was started. After 5 exchanges, no further fall in the plasma cholesterol equilibrium level was observed, despite continuation of drug therapy. Discussion Plasma exchange produces an impressive fall in plasma cholesterol concentration in homozygotes for FH [l]. Heterozygous subjects seem to behave in a similar fashion, although the plasma cholesterol appears to stabilise at higher than desirable values. The combination of plasma exchange and drug therapy can produce still lower plasma cholesterol levels, presumably through blockade of compensatory responses or through additional modes of action. Could the plasma cholesterol decrements observed after plasma exchange in Patients l-3, and expressed as equilibrium values, have been achieved by intensive drug therapy alone, without plasma exchange? The answer to this question is probably yes, as judged by the rise and fall in the plasma cholesterol level associated with stopping and restarting drug therapy (Figs. 2 and 3), and by the failure of plasma exchange to improve on drug therapy in Patient 4 (Fig. 4). It is not feasible to compare the acute effects of plasma exchange with the chronic effects of drug therapy, for reasons indicated in the Results. The acute

203

fall in plasma cholesterol level with plasma exchange is relatively short-lived, although it may be of biological importance in reducing tissue pools. We have made metabolic observations in the presence of an unsteady state, and this calls into question the validity of some of the methods, in particular faecal endogenous neutral steroids by the isotopic balance method. This method relies on the precursor-product relationship between plasma and bile steroids, and assumes approximately similar specific activities at a corresponding time point [6]. However, the isotopic balance method becomes unreliable when there are rapid shifts in plasma specific radioactivity, such as with plasma exchange. In the earlier exchanges, each patient received an infusion of fresh frozen plasma, containing up to 1 g of non-radioactive cholesterol. This usually resulted in a sharp fall in specific activity. However, after a few days subsequent changes in plasma specific activity had become more gradual. The faecal neutral steroid results were obtained from collection periods over several weeks and this probably would have allowed sufficient time for plasma and bile specific activities to equilibrate. If any error in neutral steroid estimation has resulted, it would be an overestimate, due to the rapid fall in specific activity. Although we did demonstrate increases in faecal steroid excretion after plasma exchange in Patient 2, these changes generally persisted for several weeks, which suggests that this was not a continuing overestimation. In Patients l-3 more than 100 g of cholesterol were removed from the exchangable pool. Following plasma exchange, but before drug therapy, there appears to have been an increase in the output of endogenous faecal steroids, despite the fact that plasma cholesterol concentrations were actually rising rather than falling. Although these studies were performed in an unsteady metabolic state, certain qualitative inferences can be drawn. The large amount of cholesterol so removed may well exceed the amount of exchangable cholesterol in the whole body of any of the patients before starting the study. This fact and the compensatory increase in faecal steroid excretion are consistent with both an increase in endogenous cholesterol synthesis and a mobilisation of stored tissue cholesterol. With current methodology, it is extremely difficult to separate these two phenomena, as each would have an opposing effect on the rate of decay of total plasma cholesterol specific radioactivity. Plasma exchange with concomitant drug therapy was associated with a rise in total plasma cholesterol specific radioactivity (Figs. l-3). This reflects entry into the plasma pool of “hot” cholesterol (e.g. tissue cholesterol more radioactive than plasma cholesterol) many months after pulse labelling [ 5,7]. Thus, tissue cholesterol may be mobilized after plasma exchange, but this has also been reported with [ 81 and nicotinic acid therapies [9]. Patients l-3 did report some clofibrate diminution in the size of tendon xanthomata. The pronounced plasma cholesterol decrement with clofibrate-nicotinic acid combinations has been reported previously [lo]. The observed effects of clofibrate on faecal steroid excretion are consistent with previously published observations [ 8,111. The effects of nicotinic acid on steroid excretion are unpredictable [9,12]. The combination of clofibrate or nicotinic acid with cholestyr-amine is theoretically desirable [13], because it may block the increase in cholesterol synthesis which accompanies cholestyramine therapy [ 141. The present study demonstrates how effective this block is. Thompson and Myant [15]

20-l

have previously shown that clofibrate and cholrstyraminr can decrease synthesis of apo-LDL. The reversibility of advanced coronary atherosclerosis remains in doubt, but. we have now created a circumstance to document regression. Patients l--3 have had coronary atherosclerosis documented by angiography. They have achieved substantial serum lipid lowering, although angina pectoris in Patients 2 and 3 has persisted. They will be re-submitted to coronary angiography after 2 years of this lipid-lowering regimen. It is unlikely, however, that plasma exchange will be utilised again in these or similar patients, Acknowledgements We are grateful to Miss J. Yip, Mrs. M. Rowe and Mr. J. Ruys for expert technical assistance, to Sister V. Farrell and Sister V. Clouton for assistance in patient management and operation of the Celltrifuge, and to Miss S. Maunder for secretarial assistance. References 1 Thompson, G.R., Lowt!nthal, R. and Myant, N.B., Plasma exchange in the management of homo1.y gous familial hypercholesterolaemia, Lancet, 1 (1975) 1208. 2 Thompson. G.R.. Spinks, T., Ranicar. A. and Myant. N.B., Non-st.eady state studies of low-densitylipoprotein turnover in familial hypercholesterolaemia, Clin. Sci. Mol. Med., 52 (1977) 361. 3 Lewis, B. and Myant. N.B., Studies in the metabolism of cholesterol in subjects with normal plasma cholesterol levels and in patients with essential hypercholesterolaemia. Clin. Sci.. 32 (1967) 201. 4 IIatch, F.T. and Lees, R.S., Practical methods for plasma lipoprotein analysis, Adv. Lipid Res., 6 (1968) 1. 5 Simons, L.A. and Myant, N.B., The effect of d-thyroxine on the metabolism of cholesterol in familial hyperbetalipoproteinaemia, Atherosclerosis, 19 (1974) 103. 6 Miettinen, T.A., Ahrens, E.H. and Grundy, S.M.. Quantitative isolation and gas--liquid chromat,ographic analysis of total dietary and fecal neutral steroids, J. Lipid Res.. 6 (1965) 41 I. 7 Moutafis, C.D. and Myant, N.B., The metabolism bf cholesterol in two hypercholesterolaennc patients treated with cholestyramine, Clin. Sci., 37 (1969) 443. 8 Grundy, S.M.. Ahrens, E.H., Salen, G., Schreibman, P.H. and Nestel, P.J., Mechanisms of action of clofibrate on cholesterol metabolism in patients with hyperlipidemia. J. Lipid Res.. 13 (1972) 531. 9 Miettinen, T.A., The effects of nicotinic acid on catabolism and synthesis of cholesterol in man. Clin. Chim. Acta, 20 (1968) 43. Carlson, L.A., Danielson, M., Ekberg, I., Klintemar, B. and Rosenhamer. G., Reduction of myocardial reinfarction by the combined treatment with clofibrate and nicotinic acid, Atherosclerosis, 28 (1977) 81. 11 Simons. L.A. and Myant, N.B., Cholesterol metabolism in hypertriglycerida~,lnia and the effects of treatment, Clin. Chim. Acta, 65 (1975) 117. 12 Miller. O.N., Hamilton, J.G. and Goldsmith, G.A., Investigation of the mechanism of action of nico10

tinic acid on serum lipid levels in man, Amer. J. Clin. Nutr., 9 (1960) 480. Moutafis, D.D.. Myant. N.B., Mancini. M. and Oriente, P., Cholestyramine and nicotinic acid in the treatment of familial hyperbetalipoproteinaemia in the homozygous form, Atherosclerosis, 14 (1971) 247. 14 Simons, L.A. and Myant. N.B.. A comparison of the effects of polidrxide and cholestyramine on thr metabolism of cholesterol in man, Artery, 2 (1976) 129. 15 Thompson, G.R. and Myant. N.B.. Low density lipoprotein turnover in familial hypcrcholestrrolaemia after plasma exchange, Atherosclerosis. 23, (1976) 371. 13