Lipoprotein composition in insulin-dependent diabetes mellitus with chronic renal failure: Effect of kidney and pancreas transplantation

Lipoprotein Composition in Insulin-Dependent Diabetes Mellitus With Chronic Renal Failure: Effect of Kidney and Pancreas Transplantation Thomas A. Hughes, A. Osama Gaber, Hosein S. Amiri, Xiaohu Wang, Debra S. Elmer, Rebecca P. Winsett, Donna K. Hathaway, Suzanne M. Hughes, and Maher Ghawji Chronic renal failure (CRF) in nondiabetics is associated with a number of lipoprotein abnormalities

that place these patients at high risk for atherosclerosis. This study compared the lipoprotein composition of nondiabetic controls (n = 68) with that of patients with insulin-dependent diabetes mellitus ([IDDM] n = 13) and of patients with IDDM and CRF ([IDDM + CRF] n = 74). Six lipoprotein subfractions (very-low-density lipoprotein [VLDL], intermediate-density lipoprotein [IDL], low-density lipoprotein [LDL], high-density lipoprotein-light [HDL-L], HDL-medium [HDL-Ml, and HDL-dense [HDL-D]) were isolated by rapid gradient ultracentrifugation using a fixed-angle rotor. The apolipoprotein (by reverse-phase high-performance liquid chromatography [HPLC]) and lipid (by enzymatic assays) composition of each subfraction was determined. The only abnormalities found in IDDM patients were increases in IDL and HDL-L triglyceride (TG) levels and an increase in the HDL-L free cholesterol (FC) level. The IDDM + CRF group had multiple abnormalities including (1) elevated TG, apolipoprotein (apo) C-II, and apo C-III levels in all lipid subfractions; (2) elevated VLDL and IDL apo B, TG, FC, cholesterol ester (CE), and phospholipid (PL) levels (with an increased CE/TG ratio in VLDL only); (3) decreased HDL-M apo A-l, apo A-II, CE, and PL levels, but an increased HDL-D apo A-l level; and (4) decreased lecithin:cholesterol acyltransferase (LCAT) activity. Twenty-five of the IDDM + CRF patients underwent combined pancreas and kidney (P + K) transplantation, and 12 patients received only a kidney transplant. Lipoprotein composition was determined at 3, 6, and 12 months posttransplant. Both types of transplantation resulted in similar alterations in lipoprotein composition, even though there was essential normalization of blood glucose levels in most of the patients who received a pancreas transplant (hemoglobin Ale [HbAlc], 9.1% f 1.1% v 5.7% * 0.3% at 12 months, P < .Ol). These posttransplant changes included (1) no improvement in the elevated TG level in any lipid subfraction even though there was some reduction in apo C-III levels in VLDL; (2) reductions in levels of VLDL and IDL apo B but increases in LDL apo B; (3) increases in HDL apo C-III and FC concentrations despite an increase in LCAT activity; and (4) increases in apo A-l levels in HDL-L and HDL-M. The addition of a pancreas to a kidney transplant had no obvious impact on the lipoproteins. This is probably because the difference in glycemic control between the P + K group and the kidney-alone group is usually not associated with substantial abnormalities in lipoprotein composition. This study is too short to address the issue of whether pancreas transplantation can reduce the risk of atherosclerosis associated with IDDM and CRF. The immunosuppressive drugs and persistently abnormal renal function following transplantation probably adversely affected the lipoproteins. This is unfortunate because of the added urgency for lipoprotein normalization in these patients, since they are very likely to have advanced atherosclerotic disease, having typically experienced at least one cycle of renal failure and dialysis. Copyright 0 1994 by W.B. Saunders Company

ATIENTS WITH insulin-dependent diabetes mellitus (IDDM) typically do not have hyperlipidemia unless they are severely hyperglycemic or they have chronic renal insufficiency (CRI). However, they may have subtle lipoprotein compositional abnormalities. These include (1) cholesterol enrichment of very-low-density lipoprotein (VLDL),‘,? (2) an increase of free cholesterol (FC) levels and a reduction of phospholipid (PL) levels in high-density lipoprotein (HDL),x.4 and (3) triglyceride (TG) enrichment of intermediate-density lipoprotein (IDL), low-density lipoprotein (LDL), and HDL.2 In addition, Taskinen et al5 have demonstrated a reduced input into the plasma of HDL particles containing both apolipoprotein (apo) A-I and A-II with no change in the input of HDL containing only apo A-I. As renal insufficiency progresses in nondiabetics, there is an increase in TG,h-8 apo C-III,6.9Jo and IDL*l levels along with a reduction in HDL total cholesterol (HDL-C) and apo A-I levels.h~X.lo~lz-‘JHowever, LDL-C and total plasma apo B levels are usually normal or low.6J3J5.L6 The enzymes that control lipoprotein metabolism are also affected by CRI. Lipoprotein lipase (LPL),‘OJ’ hepatic lipase (HL),“‘,17 and Iccithin:cholesterol acyltransferase (LCAT)‘JX activities are usually reduced in CRI. Shoji et alI9 found that these changes in LPL, HL. and LCAT accounted for 5 1So/c of the variation in HDL-C in uremic patients. These studies clearly indicate that renal disease in nondiabetic patients

P

Metabolism, Vol43, No 3 (March), 1994: pp 333-347

causes major perturbations of lipoprotein metabolism that are likely to put these individuals at high risk for atherosclerosis. Attman et al?” recently reported that patients with IDDM and CRI (n = 24) had elevations in total plasma TGs. total cholesterol, apo C-II, and apo C-III but reductions in apo A-I and apo A-II, similar to CR1 patients without IDDM. Renal transplantation (TX) alone in nondiabetics decreases TG levels (although they generally still remain higher than normal).8.‘“,‘J but increases LDL-C and apo B levels,s~‘“~” as well as HDL-C and apo A-I levels.8~13~1s~~~ It

From the Departments of‘ Medicine and SurgeT. Universiy of Tennessee. Memphis, TN. Submitted December 19. 1992: accepted April 12. 1993. Supported by the General Clinical Research Center (MO1 RR0021 1) and the Diabetes Trust Fund in Medical Center East, Bitmingham, AL. Presented in part at the Third International Congress on Pancreatic and Islet Transplantation and $vmposium on Ar-tificialinsulin De& e’): Systems, Lyon. France. June 6-8, 1991. Published in abstract form in Transplant Proc 24:840-841. 1992. Address reprint requests to Thomas A. Hughes. MD, Division of Endoctinology and Metabolism. Department of Medicine, lJnh~ers$v of Tennessee, Memphis, 9.51 Court Al,e, Room 34OM, Memphis. TN 38163. Copyright (1 1994 by W.B. Saunders Cornpan! 00%0495194/4.?03-0012$03.00l0 333

334

HUGHES

Table 1. Patient Characteristics Before Transplant (mean + SEM) Kidney Controls

IDDM

68

”

Age W* % lBw*

48+11

IDDM + CRF

13

74

37?9

131 t 26 139 t 23

AlOW

12

PtK

25

39 t 8

38 t 6

37 + 7

144 ?C29

153 t 30

137 t 25

% Men

50

46

50

58

40

% White

82

92

85

83

88

Fasting glucose (mg/dL) Hb A,, (%)

-

233?30

179+

13

227 + 33

143 t 13$

-

8.4 ? 0.6

7.6 + 0.2

7.9 + 0.6

7.1 ? 0.2

29 + 2

30 IL 1

Hematocrit 38 ‘- 2

(Oh) Serum creat-

302

1t

inine (mg/dU Albumin

-

1.2 + 0.1

8.1 + 0.5t

8.6 2 1.1

8.5 ? 0.7

(g/dU

-

4.2iO.l

3.6kO.lt

3.4 + 0.2

3.7 * 0.1

*Mean + standard deviation. tf

< .Ol compared with IDDM.

SP < .05 compared with kidney alone.

has been proposed that the persistent elevations in TG levels and the increases in HDL-C and apo A-I levels are the result of prednisone therapy,X,‘J since these are the typical lipoprotein changes seen in normal subjects23,24 or patient9 treated with prednisone. Cyclosporin therapy has been shown to increase LDL-C and apo B levelsZ6 and therefore may be contributing to these particular abnormalities. However, the elevation of LDL-C levels has also been seen in post-TX patients who were not treated with cyclosporin X.77 Combined pancreas and kidney transplantation (P + K) decreases total cholestero128-3(’and TGs’“J” and increases HDL-C levels.3” Pancreas-alone transplantation in IDDM patients without chronic renal failure ([CRF] n = 6) results in little change in total, VLDL, or HDL cholesterol, or total TG levels, but does increase both LDL-C and TG levels.3’ We have been monitoring the lipoprotein composition of IDDM patients undergoing either kidney-alone or combined pancreas-renal transplantation. In this report, we will present the changes that occur in lipoprotein composition during the first year posttransplant. We will also further characterize the lipoprotein abnormalities in patients with both IDDM and CRF.

ET AL

Diabetic Controls (IDDM) Thirteen patients with IDDM but normal renal function served as diabetic controls (Table 1). These individuals were being evaluated for pancreas-alone transplantation and therefore underwent the same studies as the patients with CRF.

Diabetic Patients With CRF (IDDM + CRF) Seventy-four patients with IDDM + CRF were evaluated for P + K transplantation (Table 1). Fifty-eight percent of these patients where on peritoneal dialysis at the time of transplantation, and 21% were on hemodialysis; the remainder had not yet required dialysis. All blood samples obtained from the hemodialysis patients were drawn on the day following dialysis. All blood samples obtained from the peritoneal dialysis patients were drawn after a 14-hour fast but while they were receiving continuous dialysis. There were no differences in age, age at diagnosis of diabetes. %IBW, fasting plasma glucose, hemoglobin Ale (HbA,,), or blood pressure between the IDDM and IDDM + CRF groups. However, the IDDM + CRF group was more anemic and had lower serum albumin concentrations than the IDDM group (Table I). In addition, the IDDM + CRF group had more severe autonomic neuropathy, were more likely to have left ventricular hypertrophy by echocardiography (17%, 1’ 50%). and were treated with more antihypertensive medications (0.7 ? 0.2 I’ I .4 ? 0.1 drugs/d, P = ,014). Fifty percent of patients in each group had either diastolic or systolic dysfunction by echocardiography. but there was a relatively low incidence of clinical coronary artery disease (CAD) in both groups. Only 8% of IDDM patients had either a positive thallium-exercise test or significant CAD on cardiac catheterization. whereas 28% of the patients in the IDDM + CRF group were found to be positive for disease by one of these tests. Twenty-three percent of IDDM patients (primarily thiazides) and 31% of IDDM + CRF patients (primarily furosemide) were on diuretic therapy, whereas only 4 55 of the latter group and none of the IDDM patients were on P-blockers. Seven percent of the patients in each group were treated with either lovastatin or gemfibrozil. IDDM + CRF patients on hemodialysis were less obese and had lower HbAI, and higher albumin concentrations than the IDDM + CRF patients not on dialysis (Table 2). Hemodialysis patients also had lower total plasma FC, apo B, and apo C-II/C-III ratios than either the predialysis patients or the peritoneal dialysis patients (Table 2). The reduction in FC levels was due to reductions in IDL, LDL, and HDL levels. whereas the reduction in apo B levels was primarily due to a reduction in LDL apo B levels (Table 2). There were no significant differences between predialysis patients and those treated with peritoneal dialysis.

Table 2. Pretransplant Differences Between IDDM + CRF Patients: SUBJECTS

AND

METHODS

Comparison of Those Not Being Treated With Dialysis (none) Versus Those on Peritoneal Dialysis (CAPD) or Hemodialysis (hemo)

Nondiabetic Controls Thirty-four men and 34 women served as nondiabetic controls (Table 1). The men had a mean age of 49.9 years (range, 74 to 67 years) and a mean percent ideal body weight (%lBW) of 129% ? 26% (range, 89% to 223%). and were 88%’ white and 12% black. The women had a mean age of 46.2 years (range, 26 to 69 years) and a %IBWof 134% -t 25% (range, 89% to 187%). and were 74Cr, white and 26% black. All controls had a total plasma cholesterol level less than 220 mg/dL and fasting plasma TG level less than 180 mg/dL. They were taking no medications except for replacement thyroid hormone and postmenopausal estrogens.

NOW3

15

n

CAPD

43

Hemo

10

Age (vr)

39 + 2

39 % 1

40 -t 2

% IBW

156 2 8

141 * 5

136 2 5

(mg/dU

154 + 21

182 % 17

214 -+ 37

Hb A,, (%)

8.2 ? 0.4

7.7 * 0.3

6.9 ? 0.3

32 2 2

29 2 1

28 r 1

t.0531

Fasting glucose

Hematocrit 1%)

C.036)

LIPOPROTEINS

FOLLOWING

335

TRANSPLANTATION

Table 2. Pretransplant Differences Between IDDM + CRF Patients: Comparison of Those Not Being Treated With Dialysis (none) Versus Those on Peritoneal Dialysis (CAPD) or Hemodialysis (hemo) {Cont’d) NOW

CAPD

Hemo

4.4 + 0.5

9.3 ? 0.7 (.OOl)

Creatinina (mg/dL)

8.9 + 0.8

(.OOl)

3.5 2 0.1

3.9 * 0.1

f.052)

[.039]

(.Oll)

[.020]

1.008)

[.056]

t.008)

[.021]

Albumin 3.4 * 0.2

(g/dLl

Total plasma (mg/dL) Cholesterol

230 + 18

218 + 12

189 + 11

TG

146 + 15

184 + 23

176 5 27

FC

77 + 5

72 + 4

202217

PL

107 2 8

Apo B Apt

C-III

Apo C-II Apo C-II/C-III VLDL

2032

17.8 f 1.57 5.3 + 0.6 0.31 + 0.3

60 ? 4 193 2 13

11

69 2 8

91 +5 21.4?

1.98

20.4 + 2.02

5.1 + 0.4

4.1 -t 0.5

0.26 ‘- 0.02

0.20 ? 0.02

(mg/ dL)

CE

20 + 5

TG

85?

13

31 ? 6

27 + 5

112 -t 18

119 + 22 17 -c 4

FC

15 + 2

162

PL

32 t 4

42 ? 7

2

44+7

Apo B

11+2

10 + 2

10 + 2

CE

22 + 3

23 -t 3

19 + 3

TG

25 2 3

33 2 3

23 ? 3

FC

lo+

1

11 -c 1

PL

182

2

21 2 2

18 2 3

Apo B

13 * 2

12 2 1

9.3 2 1.4

Dietary Information

IDL (mg/dL)

821

[.023]

LDL(mg/dL) CE

114 r 15

97 + 7

85+

TG

24 + 2

25 + 2

20 + 3

10

FC

38 + 3

31 +- 2 (.08)

25 t 2

i.004)

[.08]

PL

75*

64 5 4

Apo B

83 2 7

52 + 6 50 2 6

l.09) 1.003)

[.033]

11

69 2 4 1.071

HDL-L (mg/dL) CE

20 * 3

18 + 2

19 * 3

TG

4.8 + 0.5

5.7 + 0.7

3.9 + 0.6

I.081

FC

6.9 2 0.9

6.8 * 0.5

L.081

PL

26 2 4

25 c 3

5.0 f 0.9 20 ? 5

Apo A-l

24 -t 5

20 * 2

23 + 6

Apo A-II

5.9 2 0.8

5.3 ? 0.5

6.3 f 1.6

CE

23 ? 2

TG

4.6 2 0.4

6.0 2 0.7 (.08)

21 t 1

5.1 + 0.6

FC

7.5 + 0.5

7.0 f 0.4

5.8 5 0.9

PL

39 2 2

39 2 2

41 * 7

Apo A-l

81 + 5

74 * 2

80 t 5

25 + 1

23 2 1

28 -+ 2

25 + 3 f.09)

L.081

Total chc7.3 +- 0.7

7.7 2 0.4

8.5 2 0.9

TG

1.8 2 0.2

2.6 + 0.4 (.OS)

2.3 + 0.3

PL

12 2 1

12 + 1

16 + 4

Apo A-l

40 % 4

45 + 2

46 + 4

lesterol

Apo A-II

8.6 i 0.9

7.9 + 0.5

NOTE.

are means

2 SEM.

values

Results

Y none and numbers

in brackets

10.3 & 1.2 Numbers

in parentheses

are P values

Control subjects and posttransplant P + K patients were on ad libitum diets. Pretransplant IDDM patients without CRF were instructed on a standard American Diabetes Association (ADA) diet. as were kidney-alone posttransplant patients. Patients with IDDM + CRF were on standard “renal” diets depending on the level of renal failure and the type of dialysis that they were receiving. There was very little effort to maintain these IDDM + CRF patients on an ADA diet because of the severe restrictions already imposed by their renal diet. However, most patients who had previously adhered to an ADA diet seemed to continue to follow the basic tenets of this diet after the onset of CRF. Unfortunately, no data are available to determine the compliance of each patient with their particular diet prescription over the course of this study.

Lipoprotein Composition Analysis

HDL-M (mg/dL)

Apo A-II HDL-D (mg/dL)

undergone successful transplantation and who have completed at least one posttransplant lipoprotein evaluation (Table 1). Twelve of these patients have been included in the kidney-alone group because they declined to have a pancreas transplant after their initial evaluation. they received a living-related donor kidney and therefore did not receive a pancreas, or they lost their pancreas transplant within the first week of surgery due to graft thrombosis. There were no differences in age, age at diagnosis of diabetes, %IBW, HbAl,, hematocrit, serum creatinine, or albumin concentrations between these groups. However, 91% of the kidney-alone group were treated with calcium-channel blockers before transplant, whereas only 27% of the P + K group were on this treatment. All other antihypertensive medications were used by similar percentages of patients in the two groups. Between 10% and 30% of the patients continued to use furosemide after transplantation at various times as symptoms (primarily edema) required. However, after transplantation, this diuretic was generally used in lower dosages than before transplantation. Two of the P + K and one of the kidney-alone patients were on either lovastatin or gemfibrozil therapy before surgery, and these medications were restarted after surgery. None of the P + K patients were on exogenous insulin therapy at the time of their posttransplant lipoprotein analyses.

I.0451 are P

Y CAPD.

Posttransplant Groups All patients with IDDM and CRF were evaluated for combined P + K transplantation. The current analysis includes the 37 patients in the IDDM + CRF group described above who have

A single blood sample was obtained from each patient at each time point. Plasma lipoproteins (VLDL. IDL, LDL, and three HDL subfractions) were isolated by gradient ultracentrifugation as previously described except for several alterations of the salt gradient (described below). which shortened the necessary spin time.” Briefly, 9 mL plasma was increased to a density of 1.27 g/mL with 4.5 g KBr and added to an ultracentrifuge tube. A second layer of NaCl-EDTA-Tris buffer increased to a density of 1.22 g/mL with KBr was added to the centrifuge tube, and finally the tube was filled with buffer (d = 1.006 g/mL) to a finalvolume of 40 mL. The tubes were sealed and spun to 6.0 x 10” rad’ s-l at 70,000 rpm (361.000 x g) at 15°C in a 70 Ti fixed-angle rotor (4 3 hours and 15 minutes). VLDL, IDL, LDL. and three HDL subfractions designated L. M, and D (“light,” “medium,” and lowest to highest density) were obtained. These HDL “dense,” subfractions correspond roughly to HDL:b, HDL2,+3,, and HDL3bc3c. respectively. Apolipoprotein levels were measured by reversed-phase highperformance liquid chromatography (HPLC).3? Total cholesterol and FC (Boehringer Mannheim Diagnostics, Indianapolis, IN), TG (Sigma, St Louis, MO), and PL (Wako, Tokyo, Japan) concentrations of each lipoprotein subfraction were determined with commercially available enzymatic assays. The apo B content of LDL was determined by the Lowry assay in 50 mmol/L sodium

HUGHES ET AL

336

dodecyl sulfate and O.lN NaOH using bovine serum albumin as a standard. Apo B concentrations in VLDL and IDL were determined in a similar manner after precipitating the apo B and resolubilizing in sodium dodecyl sulfate.?’

Plasma Cholesterol Esterificatiow Activity (LCAT actit@) Fifty mmol/L plasma sample aliquot exactly

microliters of concentrated buffer (.I5 mol/L NaCI. 100 Tris, and 10 mmol/L EDTA) was added to 500 FL fresh and kept on ice or in the refrigerator at all times. The was divided into two aliquots of 250 p,L each. One 250-FL was left on ice. and the other was incubated at 37°C for 6 hours. Each aliquot was assayed in triplicate for FC. The change in the FC level was used to calculate the absolute cholesterol esterification activity (mg/dL/h). Each sample was incubated twice on consecutive days, and all aliquots from the same patient were analyzed in the same assay within 1 week of incubation.

Statistical Analysis Data are presented as the mean ? standard error of the mean unless otherwise stated. Differences between groups were determined by ANOVA using the NPARIWAY procedure in SAS (Statistical Analysis System. SAS Institute, Cary, NC). If the variances of the two groups proved to be unequal, then the Satterthwaite approximation for reducing the degrees of freedom was used. Differences pretransplantation and posttransplantation were determined by paired differences also using SAS. Because there were three repeated measures of the same variables posttransplantation, the Bonferroni correction would suggest that the level of significance should be reduced from P less than .05 to P less than ,017. RESULTS

Gender Differences in Lipoprotein Composition The men in all five groups (controls, IDDM, IDDM + CRF, kidney-alone, and P + K) had higher VLDL TG and lower HDL-L concentrations than the women, as expected from our previous rep0rt.j: These gender differences persisted throughout all of the subsequent analyses, except that the men had a dramatic increase in HDL-L cholesterol ester (CE), PL, apo A-I, and apo A-II levels 3 months posttransplant, which temporarily made their levels comparable to those of the women (Table 3). Unfortunately, these changes were short-lived. Given the initial lipoprotein differences and this one posttransplant divergence, both men and women had essentially the same lipoprotein changes in response to transplantation (Table 3). Therefore, all of the analyses shown below will include both genders.

Pretransplant Lipoprotein Abnormalities There were no abnormalities in the typical lipoprotein parameters in IDDM patients (Fig 1). However, the IDDM + CRF patients had significant increases in total cholesterol, TG, and total plasma apo B levels, with reductions in HDL total cholesterol and apo A-I levels. There were increases in VLDL and IDL total mass and a reduction in HDL-M mass (Fig 2). There were no differences in LDL. HDL-L, or HDL-D mass. The mass differences were associated with similar differences in apo B and apo A-I levels in each of these subfractions (Fig 2),

suggesting that the mass changes were due to alterations in the number of lipoprotein particles. The increases in VLDL and IDL levels were due to increases in all the lipid components of these lipoproteins (Fig 3). However, there was relative CE enrichment of the VLDL core but a reduction of the FC to PL ratio (FCIPL) in the IDL surface lipid. The only LDL lipid component that was significantly altered was TG (Fig 3). However, there was a significant reduction in the FC/PL ratio in LDL, along with the reduction in the CE/TG ratio (Fig 4). All levels of HDL-L components tended to be high in the IDDM group, but only the increase in the FC lcvcl reached statistical significance (Fig 5). These difl’erences were probably due to the higher percentage of women in this group. The only abnormalities seen in HDL-L and HDL-D in the IDDM + CRF group were increases in TG levels and a small increase in apo A-I levels in HDL-D (Fig 5). HDL-M. on the other hand, was substantially altered in the IDDM + CRF group (Fig 5). There were very significant reductions in apo A-I, apo A-II, PL, and CE levels, but an increase in the TG level. Unlike IDL and LDL, thcrc was a tendency toward an increase in the FC/PL ratio in both HDL-L and HDL-M, although this increase was statistically significant only in HDL-L (Fig 4). Both HDL subfractions showed substantial triglyceride enrichment of their core lipid (TG + CE, Fig 4). Apo C-III and apo C-II levels were clcvatcd in the IDDM + CRF group in both VLDL and total HDL (Fig 6). However, they were also increased in HDL in the IDDM group. The apo C-I level was only increased in VLDL in the IDDM + CRF group. These changes led to reductions in the apo C-II/C-III ratios of VLDL in both IDDM groups. However, only the IDDM + CRF group had a reduction of this ratio in HDL.

Posttransplant Lipoprotein Changes Twenty-five patients from the IDDM + CRF group described above have undergone combined P + K transplantation. Twelve additional patients either received only a kidney or lost their pancreas within the first week after surgery. Presurgical evaluations were conducted as early as 1 year before transplantation. These patients have been evaluated for up to 1 year posttransplant. At 3. 6, and I2 months posttransplant, both groups had similar blood pressure, body weight (Table 4). and hematocrit, creatinine (Table 4), and albumin concentrations. Both groups had significant weight gains at 6 months (combined. +4.2 2 1.2 kg, P = ,003) and 12 months (combined, +J.O 2 1.4 kg. P = .007) compared with their pretransplant weights. There were no differences between the groups in the dosages of azathioprine or cyclosporin. Howcvcr, P + K patients were on higher doses of prednisone at 3 months but lower doses at 12 months posttransplant (Table 4) compared with the kidney-alone group. The average number of antihypcrtensive medications was not statistically different in the two groups at 3 and 6 months (although there were trends toward greater usage in the kidney-alone group at both time points). However, at 12 months, the kidney-alone patients

LIPOPROTEINS FOLLOWING TRANSPLANTATION

Table 3. Posttransplant

337

Changes in Lipoproteins in All Men (Kidney-Alone and P + K TX) and All Women (Mean -t SEM) Pre-TX

Post-3 Months

Post-6 Months

Post-12 Months

Total plasma (mg/dL)

Ctlolesterol Men

195

Women

219 k 14

k

12

260 k 13 t.004)

212 t 10

216 k 12

240 t 14

217 + 13

221 2 12

TG Men

165 + 21

149? 32

127 ? 19 t.057)

149 + 14

Women

142 k 22

150 k 26

132 + 18

138 + 21

FC: Men

63 k 4

89 + 6 t.054)

78 ? 5 t.082)

81 + 3 t.013)

Women

69 + 4

87 2 5 (.050)

81 2 5 (.051)

76 + 4

Pl. Men

187k 18

225 k 11

188 + 5

199 i 6

Women

201 2 14

220 t 12

202 t 10

210 2 7

lOlk8

101 -+6 (.008)

Ape 6 Men

76 2 7

118 + 11

Women

87 2

a

105_t 10

94 i 8

99 2 7 1.060)

Ape C-III Men

17 k 2

24 k 2

18 2 2

1821

Women

19 t 2

22 5 3

21 -+2

18 + 2

Ape C-II Men

3.4 2 0.4

7.8 ? 0.7 (.009)

6.6 ? 1.1(.023)

5.9 + 0.8 (.062)

Women

4.6 ? 0.4

7.5 k 0.7 f.002)

6.7 2 0.9 t.026)

6.3 + 0.3 (.027)

Ape C-II/C-III Men

0.21 t 0.02

0.34 k 0.03 i.002)

0.36 + 0.05 f.039)

0.32 t 0.04 t.036)

Women

0.26 i 0.03

0.37 -c0.03 (.007)

0.32 + 0.03 i.051)

0.41 i 0.06 (,Oll)

VLDL (mg/dL) Cli Men

25 + 4

13 + 3 (.095)

11 k 2 t.052)

16 k 3

Women

25 2 6

12 + 3

1254

17 2 5

108 k 18

84? 20

75t15

9OL13

83514

83 + 17

71 t 13

81 2 15

TG Men Women FC Men

17 k 3

12 + 3

1122

152%

Women

14 k 2

13 k 3

12 k 2

1252

Men

38 2 7

27 t 6 (.073)

23 -t4 t.027)

31*4

Women

34 k 6

28 -t6

24 -t5

27 i:5

PI.

ApoB Men Women

9.9 2 1.6

7.2 2 2.1

4.5 2 1.5

9.0 t-1.8

12.0 2 2.1

7.2 + 1.9

6.0 2 1.5t.052)

8.1 f 1.6

IDL (mg/dL) CIE Men

17 k 2

14 + 3

ll+-2

15 + 2

Women

20 -t3

15 t 3

13 f 2

17t 3

Men

23 + 2

30 k 10

23 r 3

26 -'2

Women

24 f 4

27 2 5

28 k 4

23 + 4

TG

FC Men

7.8 k 0.6

9.0 2 1.8

8.4 t 1.1

10.4k 1.0

Women

9.1 2 0.9

9.4 + 1.6

9.6 2 1.4

9.1 ? 1.4

PIMen

15 2 2

16 -t3

1421

17 t 2

Women

17 k 2

17 k 3

1522

16 k 2

ApoB Men Women

9.3 lr1.2

9.2 -t2.2

7.4 2 1.1

10.3 k 1.7

10.9i 1.4

9.7 i-2.0

8.3 _f1.4

9.5 i 1.7

338

HUGHES ET AL

Table 3. Posttransplant Changes in Lipoproteins in All Men (Kidney-Alone and P + K TX) and All Women (Mean f SEM) (Cont’d) Pre-TX

Post-3 Months

Post-6 Months

92 + 9

132 z! 9 (.OOl)

109 + 9

105 ?I 10

120 -c 12

100 f 12

105?

Post-12 Months

LDL (mg/dL) CE Man Women

100 + 11

8

TG Men

23 k 2

24 2 3

19+

2

22 k 2

Women

22 ? 3

27 ? 4

22 i 2

22 + 3

Men

29 k 2

45 + 3 (.052)

41 ? 3 (.063)

40 + 2 (.026)

Women

34 t 3

45 k 3 (.051)

40 + 4 (.044)

37 + 2

Men

67 k 7

ai + 2 (.089)

71 k4

70 + 5

Women

65 r 7

79 + 6 (.072)

69 ? 6

71 k4

Men

57 k 6

101 k 6 l.074)

89 k a (.013)

81 2 6 t.007)

Women

64 k 6

88 + a t.049)

79 + 7 (.005)

81 k 6 (.OlO)

Men

15 * 2

33 k 6 (.OOg)

162

2

16 k 2

Women

23 f 2

25 c 3

26 k 5

24 ? 4

Men

4.2 + 0.5

4.8 + 1.0

3.7 k 0.5

4.0 k 0.7

Women

5.2 f 1.2

5.1 * 1.5

5.0 k 0.5

4.5 t 0.6

FC

PL

Apo B

HDL-L (mg/dL) CE

TG

FC Men

4.4 ? 0.4

11.4 + 1.6 (.004)

7.6 2 1.1 (.004)

6.6 t 0.9 (.015)

Women

6.4 2 0.6

10.4 t 1.2 (.029)

10.1 + 1.2 (.027)

8.9 2 1.2 (.066)

PL Men

17 -t 3

39 + 7 (.015)

23 + 3

25 2 4 (.075)

Women

28 k 3

37 + 4

37 -t 6

35 k 5

Men

15 + 3

34 + a (.oio)

21 +4

23 + 4 (.009)

Women

27 2 4

37 f 6 (.oia)

35 2 7

37 i 7

Apo A-l

Apo A-it Men

3.6 + 0.6

7.8 f 0.9 (.033)

5.4 2 0.6

7.3 2 1.1 (.0?2)

Women

7.1 k 0.9

a.9 + 0.9

7.6 r 1.0

8.4 ? 1.0

HDL-M (mg/dL) CE Men

22 k 2

29 + 3

26 + 2

24 t 2

Women

26 ? 1

28 +- 3

27 + 2

27 f 1

TG Men

4.9 i 0.5

4.0 + 0.9 (.060)

4.1 f 0.5

5.1 + 0.5

Women

5.4 t 1.1

4.9 k 1.4

4.6 + 0.4

5.0 k 0.4

FC Men

5.3 + 0.4

10.8 2 0.8 (.OOl)

9.7 t 1.1 (.009)

8.7 k 1.2 (.026)

Women

6.9 k 0.4

9.0 + 0.8 (.004)

9.5 k 0.7 (.019)

9.4 + 0.7 (.029)

PL Men

36 it_ 5

50 + 3 (.089)

45 k 3

44 f 3

Women

45 ? 3

47 i 3

46 t 2

48 + 2

Men

70 + 4

95 z 5 (.020)

89 + 6

a3 k 5 t.035)

Women

a2 + 4

94 rt 6

90 2 5

87 k 5

Men

25 t 2

30 ? 3

30 r 3

31 k 2 (.016)

Women

27 2 1

30 + 1 (.043)

27 2 2

31 * 2

Apo A-l

Apo A-II

HDL-D (mg/dL) Cholesterol Men

7.8 + 0.5

8.8 ‘- 0.9

7.6 + 0.5

6.8 t 0.7

Women

7.9 f 0.8

7.1 2 0.5

6.9 k 0.6

7.3 t 0.6

339

LIPOPROTEINS FOLLOWING TRANSPLANTATION

Table 3. Posttransplant

Changes in Lipoproteins in All Men (Kidney-Alone and P + K TX) and All Women [Mean 2 SEM) (Cont’d) Post-3

Pre-TX

HDL-D (mg/dL)

Months

Post-6

Months

Post-12

Months

(Cont’d)

TG Men

2.1 t 0.3

1.5 * 0.4

1.6 t 0.2

1.6 + 0.2

Women

2.4 + 0.7

2.1

1.9

2.1 2 0.2

2 0.6

k

0.2

PL Men

13 -t 1

142

1

12k

1

12

2

1

Women

12 5 2

12-t

1

105

1

12

+

1

Apo A-l Men

44+2

48 2 3

42 k 3

41

-t3

Women

43 5 3

42 2 3

39 k 3

41

k 2

Apo A-II Men

9.2 * 0.9

8.6 * 1.1

8.1 + 0.8

8.3

+ 0.7

Women

7.8 t 0.7

7.9 + 0.9

7.0 i 1.0

8.2

+ 0.9

NOTE. Numbers in parentheses arePvalues

vpre-Tx (pretransplant).

were treated with an average of 1.9 2 0.4 medications, whereas the P + K patients were on 0.9 ? 0.2 medications (P = .007). Both of these values are slightly less than their respective pretransplant values (2.2 * 0.3 li 1.2 2 0.2, P = .004). There was a slightly higher usage of all of the antihypertensive medications in the kidney-alone group posttransplant; however, calcium-channel blockers were by far the most commonly used agents in both groups (89% of kidney-alone and 53% of P + K patients). None of the P + K patients were on insulin therapy when these studies were conducted. There were no significant changes in total plasma cholesterol or TG levels posttransplantation in either subgroup or when the two subgroups were combined (Fig 7). However, there *as a significant increase in total HDL-C, apo A-I, and apo B levels if all patients were analyzed as a single group. ‘This effect was greatest at the 3-month time point, but remained significant at the l-year determination. Both subgroups had similar changes in HDL-C levels. In fact, in none of the subsequent analyses were we able to demonstrate a difference between the two subgroups. There were significant reductions in VLDL and IDL apo

0

T. Chol

Trig

HDL-C

LDL-C

ApoB

ApoA-I

Fig 1. Comparison of plasma total cholesterol, TG, HDL-C, LDL-C, and apofipoprotein determinations. Nonbracketed numbers are the P values for the comparison of the underlying bar to the nondiabetic controls (0, n = 68). Bracketed numbers are the P values for the comparison of the IDDM + CRF (M. n = 74) group to the IDDM (0, n = B) group.

B and CE levels after 6 months, and these changes persisted for 1 year in VLDL (Fig 8). Conversely, there was a highly significant increase in LDL apo B levels at all time points (Fig 8). CE levels also increased transiently in LDL. There were no changes in TG levels in any of these lipoproteins following transplantation. These alterations led to substantial reductions in the CEiTG ratio in both VLDL and IDL (Fig 9). There was no change in the FCiPL ratio in VLDL, but there was a small, transient increase in this ratio in IDL. Even though there were no changes in the FC/PL or CE/TG ratios in LDL, there were significant reductions in the surface lipid to apo B ratio and the core lipid to apo B ratio, especially after 1 year (Fig 9). These data would suggest that there was a reduction in the number of VLDL and IDL particles, while there was an increase in LDL particles. Both the VLDL and IDL core lipid became less cholesterol-enriched while the LDL particles became smaller. The most striking change in HDL posttransplantation was the very significant increase in FC levels in HDL-L and HDL-M (Fig 10). These increases were highly significant in both subgroups at all time points. There were no changes in TG or CE concentrations in any of the HDL subfractions. Therefore, the increase in total HDL-C levels described above was entirely due to this increase in FC levels. The PL content of HDL-L and HDL-M increased only slightly. Therefore. the FC/PL ratio was significantly increased in both HDL subfractions (P < .002) at 3 and 6 months posttransplantation in the combined population. However. by 12 months posttransplant, the FCIPL ratio had returned to pretransplant levels despite a continued elevation of FC levels. Apo A-I levels (Fig 10) increased modestly after transplantation in both HDL-L and HDL-M (but not in HDL-D). Apo A-II levels also increased slightly in both lipid subfractions, so there was no change in the apo A-II/A-I ratio. Apo C-III levels decreased in VLDL posttransplantation, whereas apo C-II levels did not change (Fig 11). This led to a highly significant increase in the VLDL apo C-II/C-III ratio. HDL levels also showed a significant increase in the apo C-II/C-III ratio (Fig 11). However, the

HUGHES

LDL

HDL-L

HDL-M

VLDL

HDL-D

IDL ApoB

Fig 2. Comparison of (a) total mass and (b) apo B and apo A-l concentrations symbols.

increase in HDL levels was due to a disproportionate increase in apo C-II versus apo C-III levels, since both apolipoprotein levels increased substantially after surgery. Apo C-I levels also increased in HDL at all time points (P < .OOOl), but were only increased in VLDL after 12 months (P = .04, data not shown).

The patients with only IDDM (2.62 + 0.20 mg/dL/h) had an LCAT activity similar to that of the nondiabetic controls (2.72 i 0.18 mgldlih). However, the IDDM + CRF patients had a substantially reduced LCAT activity of 2.22 f 0.09 mg/dL/h (P = .07 1’ IDDM, P = ,007 1’ controls). Patients who underwent kidney-alone transplantation had essentially identical presurgical LCAT activity

0 PL

PL

FC

FC

CE

CE

TG

TG

LDL

I

HDL-L

I

IDL-M

ET AL

HDL-D

ApoA-I

in each lipoprotein. See Fig 1 for explanation of P values and

(2.08 ? 0.24 mg/dL/h) to those who received the dual transplant (2.04 + 0. I7 mg/dL/ h). Following transplantation, both groups continued to have similar activities (Table 4). However, at each time point the LCAT activity of the combined groups was significantly greater than their prctransplant value (3 months: +0.81 ? 0.28, P = .008: 6 months: +0.70 ? 0.30, P = .029: 12 months: +I.24 ? 0.31, P = .OOOS).

DISCUSSION

Pretranspiant Lipoprotein Abnonnulitirs It has been reported that patients with IDDM without CRF have CE-enriched VLDL’.’ and TG-enriched IDL. LDL. and HDL.’ Our IDDM patients without CRF had

J

Fig 3. Comparison of PL, FC, CE, and TG concentrations in (a) VLDL. (b) IDL. and (c) LDL. CE/TG ratios and FC/PL ratios are shown in the inserts. See Fig 1 for explanation of P values and symbols.

JPOPROTEINS

341

FOLLOWING TRANSPLANTATION

0.65

4a 0.60

0.25

CE/lG

FCIPL

a

Fig 4. Comparison of CE/TG ratios and FC/PL ratios in (a) LDL, (b) HDL-L, and (c) HDL-M. See Fig 1 for explanation of P values and

symbols.

normal VLDL lipid composition and had only TG enrichment of their IDL. In addition, we saw none of the previously reported abnormalitie9 of the FC/PL ratio in any of the lipid subfractions in these IDDM patients. The differences between our findings and those of some previous studies were probably due to sampling differences, since our IDDM group and the previously reported groups consisted of less than 20 patients each. Our findings were very similar to those of Taskinen et al,’ who only selected IDDM patients without hypertriglyceridemia and found no abnormalities of the cholesterol and TG composition in any of their patients’ lipoproteins. This is only the second report of the lipoprotein abnormalities in patients with both IDDM and CRF. Similar to Attman et al,?” we found that patients with IDDM had almost all of the lipoprotein abnormalities typically associated with CRF in nondiabetics. These include elevated TG levels in all lipoprotein subfractions.h-R elevated apo C-III levels,h~“~t” elevated IDL levels,” and a reduction in the total HDL-C and total plasma apo A-I levels.h~K~t”~“-‘J However. we also found an increase in total plasma apo B levels, instead of the usually reported reduction.h.‘~.li.lf~ Attman ct al”’ also found an increase in apo B levels in IDDM patients with a glomerular filtration rate less than 20

0

GE/l-G

FCIPL

mL/min. Our patients had a reduction in LCAT activity similar to that of nondiabetic patients with CRF.‘.‘* The hypertriglyceridemia clearly affected all lipoprotein subfractions, including all three HDL subfractions. This hypertriglyceridemia was associated with an increase in apo C-III levels in both VLDL and HDL. Despite an increase in apo C-II levels in both of these subfractions, there were reductions in the apo C-II/ape C-III ratios. These alterations in apolipoproteins could be expected to decrease LPL activity and/or inhibit clearance of VLDL and IDL from the plasma. The increase in plasma apo B levels was due to almost identical absolute increases in apo B levels ( _ 5 mg/dL) in VLDL, IDL, and LDL. These increases represented less than a 10% increase in LDL apo B levels, but 100% increases in VLDL and IDL apo B concentrations. Therefore, the number of LDL particles in these patients was relatively normal, as was their total LDL mass. However, their LDL core lipid was clearly TG-enriched (as was their HDL-L and HDL-M core lipid). The reduction in HDL concentrations in our patients was entirely due to a reduction in HDL-M levels. There were no reductions in any of the components of HDL,-L. and there was actually a small increase in the HDL-D apo A-I

342

HUGHES

ET At

w 80 70 s’w a Em ? I 30 20 10 0

ApoA-I

ApoA-l

ApoA-II

ApoA-II

PL

CE

FC

PL

d

TG

ApoA-I

concentrations. Attman et aI?” also found a reduction in both apo A-I and apo A-II levels in IDDM + CRF patients.

Posttransplant Lipoprotein Changes Diabetic patients who received a P + K transplant had the same lipoprotein changes as patients who received only a kidney. This is not surprising, since IDDM patients without CRF typically have very few lipoprotein abnormalities, whereas diabetics with CRF have the same severe lipoprotein abnormalities as nondiabetic patients with CRF.20 Therefore, improving their hyperglycemia to nearnormal levels (as evidenced by their average HbA,, being <6.0%) would not be expected to have as dramatic an impact on their lipids as would correcting their renal function. This postulate is supported by the lack of an effect 10

ApoA-II

PL

FC

CE

TG

Fig 5. Comparison of apo A-l, apo A-II, PL, FC, CE, total cholesterol (TC), and TG concentrations in (a) HDL-L, (b) HDL-M, and (c) HDL-D. See Fig 1 for explanation of P values and symbols.

TG

TC

400 1

,

on lipoproteins following pancreas-alone transplantation in IDDM patients without CRF.” The more important issue is whether the addition of the pancreas transplant prevents the atherosclerosis that is so common in IDDM even without major abnormalities in lipoprotein composition. Of course, this short-term study is unable to address this issue. It is important to note that neither group had a significant improvement in the most common lipoprotein abnormality associated with renal failure. Specifically. there was no reduction in the hypertriglyceridemia in any of the lipoprotein subfractions. Typically, nondiabetics have some improvement in TG levels following renal transplantation,x,i3,1J as do diabetics following P + K transplantation.‘“J”’ This persistent elevation in TG levels was seen despite some reduction in the VLDL apo C-III level (although it

, 0.4

4

6a

0.3

10.2

.a001

.007

t.mli

I.011

ApoC-Ill

ApoCll

Fig 6. Comparison of apo C concentrations values and symbols.

ApoCl

I

0

ApoClll

ApoCll

ApoC-l

in (a) VLDL and (b) total HDL. Apo C ratios are shown in the inserts. See Fig 1 for explanation of P

343

LIPOPROTEINS FOLLOWING TRANSPLANTATION

Table 4. Posttransplant Weight, Renal Function, Blood Glucose Control, Immunosuppression,

and LCAT Activity Posttransplant

3 Months

6 Months

12 Months

Kidney-alone

68 * 4

76 2 4

76 2 6

P+K

66 ? 4

72 2 3

72 5 4

Kidney-alone

1.5 2 0.2

1.5 ? 0.1

1.8 2 0.3

P+K

1.6 ? 0.2

1.6 2 0.2

1.8 2 0.2

Kidney-alone

8.3 + 0.9

8.6 t 0.8

9.1 -c 1.1

P+K

5.2 lr 0.3*

5.7 lr 0.3t

5.7 2 0.3t

Body weight (kg)

Serum creatinine

(mg/dL)

HbA,, (W,)

Cyclosporin

dose (mg/dL)

Kidney-alone

323 + 25

290 + 33

250 t 41

P+K

346 + 36

296 + 20

377 % 46

Prednisone

dose (mg/dL)

Kidney-alone

15i

P+K

19 -t I*

LCATactivity

1

142

1

13 _f 1

14-t

1

10 f 1*

(mg/dLlh)

Kidney-alone

2.8 & 0.5

2.3 -t 0.5

2.2 ? 0.2

P+K

2.6 + 0.3

2.4 + 0.2

2.6 5 0.2

lf < .05, tP < .Ol:compared with kidney-alone

at same time period.

remained substantially above normal) and an improvement in the apo C-IIiapo C-III ratio. However, both apo C-III and apo C-II levels increased in HDL. HDL acts as a reservoir for the apo Cs, which can then be rapidly transferred to postprandial chylomicrons and VLDL. Once attached to these TG-rich lipoproteins, apo C-II and apo C-III control TG lipolysis and hepatic clearance of these particles. It is possible that this increase in HDL apo C-III levels led to an even greater excess of apo C-III in postprandial TG-rich lipoproteins and thus inhibited lipolysis and particle clearance. Both groups of patients had significant increases in LDL apo B levels following transplantation even though their VLDL .and IDL apo B levels decreased. This increase in LDL apo B levels was not associated with increases in the lipid components of LDL, so there were resultant reductions in both the surface lipid to apo B ratio and the core ml

lipid to apo B ratio. This clearly indicates that there was an increase in the number of LDL particles and that these particles were generally smaller. Both increased LDL apo B concentrations’3 and small LDL particles34 are possible risk factors for atherosclerosis. especially in patients with hypertriglyceridemia. With the data available from this study, we are obviously unable to identify the specific mechanisms for the persistent hyperlipoproteinemia in these patients. However, three possible etiological factors are readily apparent. First, all patients received prednisone and cyclosporin as part of their immunosuppression. Prednisone elevates plasma TG levels acutely,25J5 induces weight gain (particularly visceral obesity) with long-term therapy. and elevates LDL cholesterol levels.“s We found substantial weight gain in our patients, and even though we did not measure their waist to hip ratios, it was readily apparent that this excess fat was distributed centrally. Both total obesity and visceral obesity are frequently associated with excess VLDL TG secretion (TG-rich VLDL).3h In addition, cyclosporin has been shown to increase apo B levels26 and is probably contributing to this change in our patients. Second. as is typically the case, our patients clearly did not have normal renal function following transplantation (serum creatinine levels averaged 1.8 mg/dL). Grutzmacher et aI6 have shown that pretransplant nondiabetic patients with creatinine clearances between 30 and 70 mL/min (similar to those of our posttransplant patients) have elevated TG (and apo C-III) concentrations in all lipoprotein subfractions. In fact, their patients with clearances of less than 30 mL/min had very little further worsening of any of their lipoprotein measurements. It is therefore likely that posttransplant patients with mild to moderate renal dysfunction would have similar lipoprotein abnormalities, Finally, both the P + K and kidney-alone transplantation patients were clearly on more liberal diets following surgery, and this may have led to greater intakes of saturated fats and sucrose, foods that could increase LDL and TG levels. Unfortunately, we do not have food records to document these changes. The most dramatic changes in HDL composition were 200

T 7a

.07

______i TC

Fig 7. Posttransplant changes in total plasma (a) cholesterol (TC), TG, HDL-C, and (b) apolipoproteins. Numbers outside of the black boxes represent the P values of the paired differences comparing the measurement taken at the time point next to the number with the pretransplant measurement for that particular group, either the P + K I-_) group or the kidney (---) group. Numbers inside the black boxes represent the P values of the paired differences comparing the measurements taken at the time points next to the number with the pretransplant measurement for the two groups combined, both P + K and kidney groups.

HUGHES

344

ET AL

70

8a 60

\

I

I -1

Fig 8. Posttransplant changes in TG, CE, and apo B levels in (a) VLDL, (b) IDL. and (c) LDL. The TG values in VLDL were divided by 2 in order to scale them for this figure. See Fig 7 for an explanation of the P values and symbols.

1.0 0.61

: 9a

o,9

9b

CEiTG

0.5

0.6 :: ‘G

m

2

0.7

:c i0 0.6 0.5 0.4 Pm-TX

1.6

13

Fig 9. Posttransplant changes in the FC/PL ratios and the CE/TG ratios in (a) VLDL and (b) IDL. (c) Ratio of the total surface lipid (PL + FC) to apo B and the total core lipid (CE + TG) to apo B in LDL. See Fig 7 for an explanation of the P values and symbols.

1.2 ,, POSL-3 mm

LIPOPROTEINS

12

345

FOLLOWING TRANSPLANTATION

126

I-10a

wa

q

: lob

HDL-M

so

E 2 .E

(C so-

cc0

.e*

k

t f? 8 40-

_

.P

B

m

lie

HDL-L

_--4 ml m

Fig 10. Posttransplant changes in (a) FC and (b) apo A-l in HDL-L and HDL-M. See Fig 7 for an explanation of the P values and symbols. The P values for the individual groups in Fig 10a are not shown, since they are highly significant for both groups at all time points and there is not adequate room to show them.

very significant increases in the FC concentrations and the FC/PL ratios in both groups. This increase occurred despite ,a significant increase in LCAT activity. In fact, the increase in LCAT activity may have been driven by this increase in substrate (there was also a small increase in HDL PL levels). Essentially the entire increase in HDL total cholesterol levels following transplantation was due to this increase in FC levels, since we found no change in CE concentrations. Pretransplant, these patients had mid- to high-normal levels of FC, so this increase clearly pushed them into the abnormal range. The FC/PL ratio of the other lipoproteins also tended to increase. It has been suggested that an increase in this ratio is strongly associated with atherosclerosis in both animalti’and humans.jx Therefore, the increase in HDL-C levels may not be a beneficial side effect of this therapy. Short-term prednisone therapy has been shown to increase HDL-C levels.z4.25 especially HDL?, in normal subjects, but it is not known whether this was due to an increase in FC or in CE levels. Cyclosporin appears to have little impact on HDL-C levels.2h We did tind significant increases in apo A-I levels in both HDL-L and HDL-M along with a late increase in apo A-II levels in

Fig 11. Posttransplant values and symbols.

HDL-M. These would usually be considered beneficial changes and are likely to be the result of the prednisone therapy.‘4.25 The addition of a pancreas to a kidney transplant did not have any obvious impact on the lipoproteins. This is not surprising, since the difference in glycemic control between these two groups of patients is not usually associated with major abnormalities in lipoprotein composition. Any subtle improvements in lipid concentrations probably would have been obscured by the persistent renal dysfunction, medication effects, and possibly by the dietary changes. The more important question, which we cannot answer at this time, is whether pancreas transplantation will reduce the risk of atherosclerosis associated with IDDM or prevent recurrence of diabetic nephropathy. These potential benefits will probably be due to a reduction in the hyperglycemia rather than to an alteration in the currently used lipoprotein measurements. It is clear that the management of the lipoprotein abnormalities in these patients is going to be complex. It would be beneficial if we had immunosuppressive drugs that did not adversely alter lipoproteins and if renal

changes in apo C-II and apo C-III and their ratios in (a) VLDL and [b) total HDL. See Fig 7 for an explanation of the P

346

HUGHES ET AL

function could be more regularly normalized by transplantation. There are no ready solutions to these problems. We are therefore left with treating these patients in the same way that we treat IDDM patients with mild to moderate renal dysfunction. Diet and exercise treatment modalities

likely to have advanced atherosclerotic disease because they have usually already gone through at least one cycle of renal failure and dialysis. Therefore, once we have identified particular posttransplant lipoprotein abnormalities, it will be important to develop drug regimens that can be used

remain very important. Drug therapy is complicated by the concurrent use of cyclosporin (as well as other potentially toxic medications), which brings an added risk of drug reactions (particularly myositis). There is an added urgency

successfully and safely in these complex patients.

for treatment

of these patients,

since they are much more

ACKNOWLEDGMENT would like to thank James Karas for his excellent statistical advice in the preparation of the manuscript. The authors

REFERENCES 1. James RW, Pometta D: Differences in lipoprotein tion composition and distribution between type I diabetic control subjects. Diabetes 39:1158-l 164, 1990

subfracmen and

2. Rivellese AA, Roman0 G, Patti L. et al: Lipoprotein composition abnormalities in long-term well controlled insulin-dependent diabetic patients. Diabetes Nutr Metab 5:99-105. 1992 3. Bagdade JD. Subbaiah PV: Abnormal high density lipoprotein composition in women with insulin-dependent diabetes. J Lab Clin Med 113:235-240,1989

hemodialysis 1983

and in renal

graft

recipients.

Nephron

35:171-174,

16. Gibson JC, Savdie E, Simons LA, et al: High density and low density lipoproteins in chronic renal failure before and after renal transplantation. Aust NZ J Med 9:136-142, 1979 17. Verschoor L, Baggen RGA, Jansen H, et al: Different causes for hypertriglyceridemia in male and female patients with chronic renal failure. J Clin Endocrinol Metab 56592-598, 1983

4. Bagdade JD, Subbaiah PV: Whole-plasma and high density lipoprotein subfraction surface lipid composition in IDDM men. Diabetes 38:1226-1230. 1989

18. Jung K, Neumann R, Precht K. et al: Lecithin:cholesterol acyltransferase activity, HDL cholesterol and apolipoprotein A in serum of patients undergoing chronic haemodialysis. Enzyme 25~273.275, 1980

5. Taskinen MR. Kahri J, Koivisto V. et al: Metabolism of HDL apolipoprotein A-I and apolipoprotein A-II in type I (insulindependent) diabetes mellitus. Diabetologia 35:347-356, 1992

19. Shoji T, Nishizama Y, Nishitani H. et al: Impaired metabolism of high density lipoprotein in uremic patients. Kidney Int 41:1653-1661.1992

6. Grutzmacher P, Marz W, Peschke apolipoproteins during the progression Nephron 50:103-111, 1988

20. Attman PO. Nyberg G, William-Olsson T, et al: Dyslipoproteinemia in diabetic renal failure. Kidney Int 42:1381-1389, 1992

B, et al: Lipoproteins and of chronic renal disease.

7. Wakabayashi Y. Okubo M, Shimada H, et al: Decreased VLDL apolipoprotein CIIiapolipoprotein CIII ratio may be seen in both normotriglyceridemic and hypertriglyceridemic patients on chronic hemodialysis treatment. Metabolism 36:815-820, 1987 8. Allegra V, Mengozzi G, Vasile A: Lipoprotein lipids and apolipoproteins in healthy renal transplant patients. Nephron 51:216-217, 1989 9. Averna MR, Barbagallo CM, Galione A. et al: Serum apolipoprotein profile of hypertriglyceridemic patients with chronic renal failure on hemodialysis: A comparison with type IV hyperlipoproteinemic patients. Metabolism 38:601-602. 1989 10. Ohta T, Hattori S, Nishiyama qualitative changes of apolipoprotein in patients on continuous ambulatory lism 38:843-849, 1989

S, et al: Quantitative and AI-containing lipoproteins peritoneal dialysis. Metabo-

11. Ron D, Oren 1. Aviram M, et al: Accumulation of lipoprotein remnants in patients with chronic renal failure. Atherosclerosis 46~67-75. 1983 12. Fuh MMT, Lee CM, Jeng CY, et al: Effect of chronic renal failure on high density lipoprotein kinetics. Kidney Int 37:1295l300.1990 13. Lacour B, Roullet JB. Beyne P, et al: atherogenicity indices by the analysis of patients with chronic renal failure with or and in renal transplant patients. J Clin 23:805-810. 1985

Comparison of several serum composition in without hemodialysis, Chem Clin Biochem

14. Ettinger WH, Bender WL. Goldberg AP, et al: Lipoprotein lipid abnormalities in healthy renal transplant recipients: Persistence of low HDLz cholesterol. Nephron 47:17-21, 1987 15. Rubies-Prat J, Romero R, Chacon P, et al: Apoprotein A and apoprotein B in patients with chronic renal failure undergoing

21. Kasiske BL, Umen AJ: Persistent transplant patients. Medicine 66:309-316,

hyperlipidemia 1987

in renal

22. Savdie E, Gibson JC, Stewart JH, et al: High density lipoprotein in chronic renal failure and after renal transplantation. Br Med 3 1:928-930, 1979 23. Taskinen MR, Kuusi T. Yki-Jarvinen H. et al: Short-term effects of prednisone on serum lipids and high density lipoprotein subfractions in normolipidemic healthy men. J Clin Endocrinol Metab 67:291-297, 1988 24. Ettinger WH. Hazzard WR: Prednisone increases very-lowdensity lipoprotein and high-density lipoprotein in healthy men. Metabolism 37:1055-1058, 1988 25. Ettinger WH. Klinefelter HF. Kwiterovitch PO: Effect of short-term. low dose corticosteroids on plasma lipoprotein lipids. Atherosclerosis 63:167-172, 1987 26. Ballantyne CM, Podet EJ, Patsch cyclosporine therapy on plasma lipoprotein 56,1989

WP, et al: Effects of levels. JAMA 262:53-

27. Kasiske BL: Risk factors for accelerated atherosclerosis renal transplant recipients. Am J Med 84:985-992, 1988

in

28. Georgi BA, Bowers VD, Smith JL, et al: Comparison of fasting serum cholesterol levels between diabetic recipients who undergo renal-pancreas transplants and renal transplantation only. Transplant Proc 21:2839-2840.1989 29. Dafoe DC. Campbell DA, Merion RM, et al: Pancreatic transplantation-University of Michigan. Transplant Proc 4:55-62, 1987 (suppl4) 30. Larsen JL, Stratta RJ, Ozaki CF, et al: Lipid status after pancreas-kidney transplantation. Diabetes Care 15:35-42, 1992 31. Bolinder J, Tyden G. Tibell A, et al: Long-term metabolic control after pancreas transplantation with enteric exocrine diversion. Diabetologia 34:S76-S80, 1991 (suppl 1)

LIPOPROTEINS

FOLLOWING TRANSPLANTATION

32. Hughes TA, Moore MA, Neame P, et al: Rapid quantitative apolipoprotein analysis by gradient ultracentrifugation and reversed-phase high performance liquid chromatography. J Lipid Res 29:363-376. 198X 33. Vega GL, Grundy SM: Comparison of apolipoprotein B to cholesterol in low density lipoproteins of patients with coronary artery disease. J Lipid Res 25:580-592, 1984 34. Campos H, Genest JJ Jr. Blijlevens E, et al: Low density lipoprotein particle size and coronary artery disease. Arterioscl Thromb 12:187-195, 1992 35. Brindley DN, Salter AM: Hormonal regulation of the hepatic low density lipoprotein receptor and the catabolism of low

347

density lipoproteins: Relationship with the secretion of very low density lipoproteins. Prog Lipid Res 30:349-360, 1991 36. Leenen R, van der Kooy K. Seidell JC. et al: Visceral fat accumulation measured by magnetic resonance imaging in relation to serum lipids in obese men and women. Atherosclerosis 94:171181.1992 37. Kuksis A, Roberts A, Thompson JS, et al: Plasma phosphatidylcholine/free cholesterol ratio as an indicator for atherosclerosis. Arteriosclerosis 3:389-397. 1983 38. Kuksis A, Myler JJ, Geher K, et al: Decreased plasma phosphatidylcholineifree cholesterol ratio as an indicator of risk for ischemicvascular disease. Arteriosclerosis 2:296-302,1982