Lipopheresis in the nephrotic syndrome

Kidney International, Vol. 56, Suppl. 71 (1999), pp. S-6–S-9

Lipopheresis in the nephrotic syndrome CAROL BRUNTON, ZACHARIAH VARGHESE, and JOHN F. MOORHEAD Centre for Nephrology, The Royal Free and UCL School of Medicine, London, England, United Kingdom

Lipopheresis in the nephrotic syndrome. Background. Experimental models have established a role for lipoproteins in the pathogenesis of progressive renal failure. However, conventional treatment rarely normalizes the high serum cholesterol of the nephrotic syndrome. The removal of low-density lipoprotein by lipopheresis is discussed. Methods. Lipopheresis may be beneficial in nephrotic patients with focal segmental glomerulosclerosis. The authors studied the long-term effects of low-density lipoprotein cholesterol (LDL-C) removal using the Kaneka Liposorber system, which binds LDL-C to dextran sulfate in a controlled trial in 20 nephrotic patients with different renal diseases. Results. A 21-month clamp of plasma total cholesterol at 6.0 mmol/liter or below was significantly lower than controls (x2 5 84.3, P , 0.001), followed 12 aphereses over 6 to 12 weeks in all but three apheresed patients. 1/Cr slopes were unchanged when the 50-day average period of lipopheresis treatments was excluded from analysis. Proteinuria was not reduced, but serum albumin tended to rise (NS). Fibrinogen fell by 29.8%; high-density lipoprotein, apoA1, and Lp(a) were unchanged. Two apheresed patients had a prolonged remission with a reduction of proteinuria to less than 250 mg/24 hr. The reasons for prolonged reduction of total cholesterol include depletion of tissue cholesterol, an improved fractional catabolic rate of very low density lipoprotein (VLDL), increased hepatocyte LDL turnover, and the maintenance of statin therapy. Conclusion. Lipopheresis is a safe and effective method for the control of LDL in nephrotic syndrome. Early clamping of total cholesterol in the normal range resulted in a prolonged and significant reduction of LDL compared with controls.

tients, multiple pharmacotherapy, and the presence of other risk factors such as hypertension and significant nephron loss. Because of the paucity of information on the use of lipopheresis in nephrotics, we used extracorporeal lowdensity lipoprotein (LDL) apheresis to achieve reductions in LDL of 60 to 70% in an 18-month controlled trial. LDL apheresis has been safely used in familial hypercholesterolemia [4], coronary artery disease heart disease [5], and peripheral vascular disease [6]. In nephrotics with steroid-resistant focal segmental glomerulosclerosis, there were short-term improvements in serum albumin, proteinuria, and creatinine clearance (CCr) in over 50% of those treated in several uncontrolled trials (Table 1) [7, 8]. Additional case reports also suggested that LDL apheresis was clinically valuable [7–9]. METHODS Patients Twenty nephrotic patients with a variety of biopsyproven glomerular diseases causing persistent hyperlipidemia and proteinuria were allocated to two groups by the minimization technique after informed consent. All had persistent proteinuria of more than 1.2 g/24 hr, and a CCr of more than 18 ml/min. Patients with amyloidosis, diabetes, endocrine or hepatic disease, minimal change disease, and/or a renal transplant were excluded.

The relationship between lipoproteins and renal disease has been summarized in numerous publications [1, 2] and conclusively demonstrated in animal experiments. Experimental evidence suggests that lipid-induced glomerular lesions have a pathogenesis analogous to atherosclerosis and that the lesions can be prevented or ameliorated by lipid-lowering therapy [1, 3]. In humans, however, the results of therapy using equivalent lipidlowering regimes have been disappointing, although this approach has not been fully tested in nephrotic patients. The reasons cited for this include small trials, the relative ineffectiveness of lipid-lowering therapy in nephrotic pa-

Randomization Patients were allocated to lipopheresis treatment (group A) or drugs (group B) in such a way as to minimize the difference in scores between the groups, using five prognostic factors: proteinuria of more than 3.0 g/24 hr, age of more than 45 years, male gender, glomerular filtration rate (GFR) of less than 65 ml/min, and concurrent immunosuppression. Group A had 11 patients, and group B had 9 patients. Apheresis group In group A, lipopheresis of 2.5 to 4.0 liters of plasma was performed 12 times over 6 to 12 weeks, and total cholesterol was clamped at 6.0 mmol/liter or less using

Key words: apheresis, statin, proteinuria, fibrinogen, cholesterol clamp, lipoprotein.

 1999 by the International Society of Nephrology

S-6

Brunton et al: Lipopheresis in the nephrotic syndrome

S-7

Table 1. Lipopheresis in two uncontrolled trials [7, 8] Reference author [7] Sakai et al [8] Muso et al

N of patients 16 8

Proteinuria Down by 84% Down 50% in 3 No change in 5

Creatinine clearance

Clinical notes

Up in 16 Up in 7

Useful in 56% Steroid resistant FSGS

the Kaneka Liposorber. LDL apheresis was followed by the administration of an 3-hydroxy-3-methylglutaryl co-enzyme A (HMG-CoA) reductase inhibitor (simvastatin) to maintain a significant and sustained reduction in serum total cholesterol, LDL cholesterol (Fig. 1), and triglyceride for a period of up to 18 months. Drugs group Group B patients were treated with drugs alone. Simvastatin at 40 mg/day was given to both groups. RESULTS Group A Low-density lipoprotein cholesterol (LDL-C) was #/5 4.0 mmol/liter for 113 months in apheresed patients versus only 19 months for drugs alone. In group A, the LDL-C was .4.0 mmol/liter for 74 months vs. 138 months for drugs alone (x2 5 84.3, P , 0.001). Serum apoB levels mirrored the reduction in LDL, whereas high-density lipoprotein cholesterol levels, apolipoprotein A1, and lipoprotein(a) [Lp(a)] remained unchanged. In Group A, the total cholesterol was clamped at or below 6.0 mmol/liter from a mean start level of 9.92 mmol/liter. Fibrinogen fell by 29.8% (P , 0.01) compared with preapheresis levels, and there was an upward but nonsignificant trend in serum albumin. Interestingly, two apheresis patients had a sustained remission with urine protein concentrations of less than 250 mg/liter and a rise in serum albumin to more than 40 g/liter. Because apheresis treatments lasted 50 days on average to complete, this treatment period was excluded from analysis of changes in renal function. When analyzed this way, there was no demonstrable difference between groups in urinary protein excretion, rate of progression of renal disease using 1/Cr plots, rate of change of GFR in ml/min/month, percentage change in GFR, change in effective renal plasma flow, or change in filtration fraction. The blood pressure was unaffected. DISCUSSION The technique demonstrated, to our knowledge for the first time, that the cholesterol level could be effectively and safely clamped at the arbitrarily chosen level of 6.0 mmol/liter or below in hyperlipidemic nephrotic

Fig. 1. Prolonged reduction in low-density lipoprotein (LDL) cholesterol in the lipopheresis group by comparison with the drugs group. (A) Total cholesterol in the lipopheresis group. (B) Total cholesterol in the drugs group.

patients. Interestingly, serum cholesterol control extended far beyond the 50-day mean duration of apheresis. In addition, the remission in two participants, one male and one female, both aged 70 years, with membranous and mesangiocapillary glomerulonephritis, respectively, occurred despite previous unresponsiveness to immunosuppressive drugs. Their age, nephrotic range proteinuria, and diagnoses put them in a poor prognostic group, and they both had extensive glomerular basement membrane apoE deposition in renal biopsies. Although there was no effect on proteinuria or progression of disease in the majority, all of the patients continue to be followed because many cardiovascular studies suggest that the benefit of lipid lowering cannot be clearly established until between two and five years after the treatment has been started [10]. The mechanism by which the “cholesterol clamp” was achieved can only be speculated upon, as the processes

S-8

Brunton et al: Lipopheresis in the nephrotic syndrome

involved in nephrotic hyperlipidemia are incompletely understood. However, the reduction of plasma LDL levels by up to 70% probably up-regulates LDL receptors, whereas the up-regulation of HMG-CoA reductase is prevented by the coadministration of a statin. Simvastatin, the most potent HMG-CoA reductase inhibitor available at the time of the study, was given at a dose of 40 mg/day. It is likely that aggressive reduction of plasma cholesterol led to both mobilization of tissue stores of cholesterol [11] and constitutive release of apoB lipoprotein from the liver [12]. Twice-weekly treatments may have depleted these stores, leading to a resetting of the steady state at a significantly lower level. The significant improvement in triglyceride in apheresed patients would suggest an improved fractional catabolic rate (FCR) of very low density lipoprotein (VLDL), which would also be reflected in lower plasma LDL. In addition, we suggest that the cholesterol ester-rich VLDL synthesized in the nephrotic syndrome is a less favored substrate of LPL and is less easily metabolized than the triglyceride-rich VLDL that replaces it after lipopheresis. The resulting improvement in FCR and lower VLDL has been shown to improve the apo-CIII fraction and thus shift the ratio of CII/CIII such that LPL activity is increased [13]. This may have occurred in our current study. At the beginning of a course of apheresis, total plasma cholesterol rose to pretreatment levels within two to four days, but as apheresis treatments incremented, the starting concentration of cholesterol was much lower than at the beginning of treatment. Cholesterol synthesis was probably supplemented by cholesterol mobilization from tissues or constitutive hepatic release of apoB with VLDL secretion. There was no significant reduction in proteinuria during the study. Renal clearance of albumin has been shown by Davies et al [14] and Kaysen et al [15] to be the major stimulus for lipid synthesis. Except in those who remitted, only mild reductions in proteinuria were noted during the twice-weekly apheresis treatment period. It is possible that this change, however small, may have been sufficient to reduce the “proteinuria stimulus” to LDL secretion allowing the latter to be reestablished at a lower rate. LDL apheresis may have removed or altered the flux of other contributors to hypercholesterolemia. Receptor associated protein (RAP) and a2-macroglobulin are two candidate proteins that may have been present in increased concentrations following nonspecific increased synthesis in the nephrotic state. Both inhibit low-density lipoprotein receptorrelated protein (LRP), a multifunctional receptor that is regulated by RAP and responsible for removal of apoE containing lipoprotein [16], namely VLDL remnants and intermediate density lipoprotein (IDL). Only a small proportion of synthesized apoB is secreted, while the rest is degraded intracellularly. Low

concentrations of oleate-saturated albumin inhibit apoB degradation and high free fatty acid (FFA) concentrations inhibit LPL activity. In addition, increased chylomicron remnants have been shown to stimulate apoB secretion threefold in HepG2 cells, an effect that is even more pronounced when the albumin concentration is low [17]. The trend toward an increase in serum albumin and possibly postapheresis changes in lipoprotein and fatty acid composition may have redressed the balance in favor of intracellular degradation of apoB and an improved lipoprotein FCR. The kidney’s high oxygen consumption depends on a high renal blood flow, and there were short-term postapheresis improvements in effective renal plasma flow that were not sustained when studied at six month intervals. Comparable studies of apheresis in cardiovascular and familial hypercholesterolemia (FH) patients have shown significant improvements in renal plasma flow following a single treatment [18]. It is therefore possible that temporary reductions occurred in circulating mevalonate, which is oxidized in the kidney to form sterols, nonsterols and carbon dioxide. A cholesterol “clamp” was not achieved in three lipopheresis patients. Although no specific apoE genotype was present in these three, one had an E3/E2 phenotype that unmasked a type III dysbetalipoproteinemia in the presence of renal disease. He nevertheless achieved a 70% reduction in triglyceride and a 57% reduction in total cholesterol. The other two patients had no unremarkable clinical features, but all three had heavy proteinuria, and it is possible that they had excessive urinary loss of an 39-59-adenosine monophosphate (AMP)-activated protein kinase that inactivates HMG-CoA reductase and acetyl-coA carboxylase, the rate-limiting enzyme of fatty acid synthesis [19]. The observed increase noted in serum albumin without a concomitant reduction in net urine protein loss underlines the poor correlation between these two variables, as previously noted in progressive renal failure patients whose urinary protein excretion did not change [20]. Urine protein excretion is the net loss of filtered protein after tubular reabsorption; it does not need to be altered by lipid-lowering strategies and is a poor guide to their efficacy. Its contribution will always remain unknown in the absence of the ability to measure the two unquantifiable major variables of filtered and reabsorbed protein. Reprint requests to Dr. John F. Moorhead, Centre for Nephrology, The Royal Free and UCL School of Medicine, Pond Street, Hampstead, London NW3 2QG, United Kingdom. E-mail: [email protected]

REFERENCES 1. Moorhead JF, Chan MK, El Nahas AM, Varghese Z: Lipid nephrotoxicity in chronic progressive glomerular and tubulo-interstitial disease. Lancet 2:1309–1312, 1982

Brunton et al: Lipopheresis in the nephrotic syndrome 2. Keane WF, Mulcahy WS, Kasiske BL, Kim Y, O’Donnell MP: Hyperlipidemia and progressive renal disease. Kidney Int 39(Suppl 31):S41–S48, 1991 3. Diamond JR: Analogous pathobiologic mechanisms in glomerulosclerosis and atherosclerosis. Kidney Int 39(Suppl 31):S29–S34, 1991 4. Tatami R, Inoue N, Itoh H, Kishino B, Koga N, Nakashima Y, Nishide T, Okamura K, Saito Y, Teramoto T: Regression of coronary atherosclerosis by combined LDL-apheresis and lipidlowering drug therapy in patients with familial hypercholesterolemia: A multicenter study. The LARS Investigators. Atherosclerosis 95:1–13, 1992 5. Kroon AA, Aengevaeren WR, van der Werf T, Uijen GJ, Reiber JH, Bruschke AV, Stalenhoef AF: LDL-Apheresis Atherosclerosis Regression Study (Laars): Effect of aggressive versus conventional lipid lowering treatment on coronary atherosclerosis. Circulation 93:1826–1835, 1996 6. Agishi T: Anion-blood contact reaction (ABC reaction) in patients treated by LDL apheresis with dextran sulfate-cellulose column while receiving ACE inhibitors. (letter) JAMA 271:195–196, 1994 7. Sakai S, Mune M, Iino Y, Ohtsubo O, Sugino N: Effect of LDL apheresis in patients with drug-resistant nephrotic syndrome: Multi-center co-operative study. Kidney Dial 33:321–328, 1992 8. Muso E, Yashiro M, Matsushima M, Yoshida H, Sawanishi K, Sasayama S: Does LDL-apheresis in steroid-resistant nephrotic syndrome affect prognosis? Nephrol Dial Transplant 9:257–264, 1994 9. Hattori M, Ito K, Kawaguchi H, Tanaka T, Kubota R, Khono M: Treatment with a combination of low-density lipoprotein apheresis and pravastatin of a patient with drug-resistant nephrotic syndrome due to focal segmental glomerulosclerosis. Pediatr Nephrol 7:196– 198, 1993 10. Law MR, Wald NJ, Thompson SG: By how much and how quickly

11. 12.

13.

14. 15. 16. 17. 18.

19. 20.

S-9

does reduction in serum cholesterol concentration lower risk of ischaemic heart disease? BMJ 308:367–372, 1994 Thompson GR: Mobilisation of tissue cholesterol by apheresis. (review) Prog Clin Biol Res 337:183–187, 1990 Pullinger CR, North JD, Teng BB, Rifici VA, De Brito AER, Scott J: The apolipoprotein B gene is constitutively expressed in HepG2 cells: Regulation of secretion by oleic acid, albumin, and insulin, and measurement of the mRNA half-life. J Lipid Res 30:1065–1077, 1989 Nishizawa Y, Shoji T, Nishitani H, Yamakawa M, Konishi T, Kawasaki K, Morii H: Hypertriglyceridemia and lowered apolipoprotein C-II/C-III ratio in uremia: Effect of a fibric acid, clinofibrate. Kidney Int 44:1352–1359, 1993 Davies WR, Staprans I, Hutchinson FN, Kaysen GA: Proteinuria, not altered albumin metabolism, affects hyperlipidemia in the nephrotic rat. J Clin Invest 86:600–605, 1990 Kaysen GA, Gambertoglio J, Felts J, Hutchison FN: Albumin synthesis, albuminuria and hyperlipemia in nephrotic patients. Kidney Int 31:1368–1376, 1987 Herz J: The LDL receptor related protein: Portrait of a multifunctional receptor. Curr Opin Lipidol 4:107–113, 1993 Cianflone K, Vu H, Zhang Z, Sniderman AD: Effects of albumin on lipid synthesis, apo B-100 secretion, and LDL catabolism in HepG2 cells. Atherosclerosis 107:125–135, 1994 Rubba P, Iannuzzi A, Postiglione A, Scarpato N, Montefusco S, Gnasso A, Nappi G, Cortese C, Mancini M: Hemodynamic changes in the peripheral circulation after repeat low density lipoprotein apheresis in familial hypercholesterolemia. Circulation 81:610–616, 1990 Henin N, Vincent MF, Gruber HE, van den Berghe G: Inhibition of fatty acid and cholesterol synthesis by stimulation of AMPactivated protein kinase. FASEB J 9:541–546, 1995 Moorhead JF, Persaud JW, Varghese Z, Sweny P: Serum cholesterol falls spontaneously in nephrotic patients with progressive renal disease. Renal Fail 15:389–393, 1993