American Journal of Transplantation 2008; 8: 103–110 Blackwell Munksgaard
C 2008 The Authors C 2008 The American Society of Journal compilation Transplantation and the American Society of Transplant Surgeons
doi: 10.1111/j.1600-6143.2007.02035.x
Proinflammatory CD14+CD16+ Monocytes Are Associated With Subclinical Atherosclerosis in Renal Transplant Patients C. Ulrich∗ , G. H. Heine, M. K. Gerhart, H. Kohler ¨ and M. Girndt Department of Medicine IV, University of the Saarland, Homburg/Saar, Germany ∗ Corresponding author: Christof Ulrich,
[email protected] Atherosclerotic cardiovascular disease is a major cause of death in renal transplant (TX) recipients. Atherosclerotic lesions are characterized by monocytic infiltration. Circulating monocytes can be divided into functionally distinct subpopulations, among which CD14++CD16+ and CD14+CD16+ monocytes (summarized as CD16+ monocytes) are proinflammatory cells. We hypothesized that the frequency of circulating CD16+ monocytes is associated with subclinical atherosclerosis in TX patients. Monocyte subpopulations were quantified in 95 TX and 31 hemodialysis patients (HD). In TX patients, subclinical atherosclerosis was determined by carotid intima media thickness (IMT) measurement. TX patients had lower frequencies of CD16+ monocytes than HD patients. When stratifying by immunosuppressive treatment, patients on methylprednisolone (MP) therapy had fewer CD14+CD16+ monocytes than patients not receiving MP. CD14+CD16+ monocytes decrease very shortly after transplantation. CD14+CD16+ monocyte frequency correlated with IMT in TX recipients (r = 0.34, p < 0.001). This correlation was most pronounced among patients without MP treatment (r = 0.55, p = 0.02). In a multivariate regression analysis, the association of CD14+CD16+ monocytes with IMT was independent from traditional cardiovascular risk factors. The frequency of proinflammatory CD14+CD16+ monocytes is independently associated with subclinical atherosclerosis in transplant recipients. Further studies on the association between circulating leukocytes and atherosclerosis should take monocyte heterogeneity into account. Key words: Atherosclerosis, CD14+CD16+, intima media thickness, monocytes, renal transplant recipients Received 25 October 2006, revised 18 September 2007 and accepted for publication 25 September 2007
Introduction Cardiovascular disease (CVD) is a leading cause of death for adult renal allograft recipients (1) and fatal cardiovascular events are a major cause of late allograft loss (2). Recent extensive evidence supports inflammation as a key pathogenetic mechanism in the development and progression of atherosclerosis and in triggering clinical atherothrombotic CVD events (3). Monocytes play a key role in atherogenesis. They invade lesion-prone vessel areas, in which they differentiate into macrophages and foam cells. These are the predominant cell population in the early stage of atherosclerosis. Later, macrophages release lysosomal enzymes and may thus induce the fracture of the fibrous cap in the plaque shoulder of complex atheromatous lesions, which results in thrombocyte aggregation and acute ischemic syndromes. Monocyte heterogeneity is widely acknowledged (4). Monocytes may differentiate into phenotypically and functionally distinct subpopulations that can be distinguished by the expression of the LPS receptor CD14 and the Fcc receptor CD16. The majority of monocytes is CD14 strongly positive and CD16 negative (CD14++CD16−). In healthy subjects some 10–20% of all monocytes are CD16 positive (CD16+), which are further subdivided into CD14++CD16+ and CD14+CD16+ cells. These latter subsets are considered proinflammatory monocytes, as they efficiently produce proinflammatory cytokines (5), while they produce little of the anti-inflammatory cytokine IL-10 in comparison to CD14++CD16− monocytes (6). Increased numbers of CD14+CD16+ monocytes are found in inflammatory and infectious diseases. We recently reported increased frequencies of CD14+CD16+ monocytes in patients suffering from coronary heart disease (7). Similarly, patients suffering from end-stage renal disease have higher numbers of proinflammatory CD14+CD16+ monocytes compared to healthy controls (8–12). This shift in monocyte subpopulations may contribute to the accelerated atherosclerosis in dialysis patients. Renal transplantation reverses uremia, and immunosuppressive medication directly alters the distribution of 103
Ulrich et al.
monocyte subpopulations. We thus hypothesized that, compared to dialysis treatment, renal transplantation may partially normalize the frequency of proinflammatory monocytes. This hypothesis raised the question whether the frequency of proinflammatory monocytes is associated with subclinical atherosclerotic changes in renal transplant (TX) recipients.
Subjects and Methods
Table 2: Immunosuppressive regimen of 95 kidney transplant recipients, stratified according to methylprednisolone (MP) intake TX
Cyclosporine (CyA) Tacrolimus (FK) No calcineurin inhibitor Azathioprine (AZA) Mycophenolate mofetil (MMF)
−MP (N = 18)
+MP (N = 77)
17 1 – 6 1
40 35 2 24 4
Subjects A total of 95 TX patients were cross-sectionally studied. All patients had been transplanted for at least 6 months (mean 82 ± 63 months, range 6–237 months) and were being followed-up by our outpatient clinic (Table 1). These 95 patients represent a subgroup of a larger cohort of transplant recipients who participated in an ultrasound study, on which we recently reported (13). There were 93 deceased donors and 2 living kidney donations within 95 TX recipients. Mean serum creatinine was 1.7 ± 0.7 mg/dL. Patients with rapid deterioration of renal function (increase in serum creatinine >0.8 mg/dL within 28 days) were excluded. All but two patients received the calcineurin inhibitors cyclosporine A (CyA) or tacrolimus (FK) for immunosuppression. Coimmunosuppressive medication comprised azathioprine (AZA), mycophenolate mofetil (MMF) and/or methylprednisolone (MP) (Table 2). As a reference group, we investigated a cohort of 31 hemodialysis patients (HD) who had no systemic immunosuppressive therapy and who were matched for age (± 5 years), gender, the presence of diabetes mellitus
Table 1: Characteristics of hemodialysis (HD) and renal transplant patients without (−MP) and with methylprednisolone medication (+MP)
(+MP, N = 77)
Age (years) 54.8 ± 15.0 52.1 ± 14.1 Gender (male, %) 44.4 57.1 Time on dialysis (years) 4.9 ± 4.8 4.5 ± 4.4 Time since TX (years) 11.8 ± 4.2 5.8 ± 5.0 Creatinine (mg/dL) 1.6 ± 0.5 1.7 ± 0.7 GFR (mL/min/1.73 m2 ) 55.6 ± 24.7 53.1 ± 21.6 CRP (mg/L) 5.5 ± 5.5 8.6 ± 13.0 Mean blood 106.9 ± 10.3 109.0 ± 12.2 pressure (mmHg) Pulse pressure (mmHg) 73.9 ± 18.4 65.5 ± 14.3 Cholesterol (mg/dL) 199.7 ± 45.9 207.4 ± 42.9 Triglycerides (mg/dL) 170.6 ± 85.3 177.8 ± 107.0 DM (%) 27.8 27.3 Current smoker (%) 22.2 20.8 CVD (%) 33.3 20.8 Diuretics (%) 38.9 46.8 ARB (%) 38.9 33.8 ACE I (%) 33.3 20.8 CCB (%) 38.9 57.1
Reevaluation of intima media thickness (IMT) and the frequency of monocyte subgroups were determined in a subgroup of these 95 TX patients (N = 41). The follow-up examination of these patients was 2.6 ± 0.3 years after the first analysis.
p-Values 0.482 0.238 0.847 0.001 0.504 0.671 0.343 0.499 0.036 0.499 0.792 0.966 0.893 0.256 0.546 0.681 0.256 0.162
CVD = cardiovascular disease; ARB = angiotensin receptor blocker; ACE I = ACE inhibitor; CCB = calcium channel blocker; GFR = glomerular filtration rate; DM = diabetes mellitus.
104
In six TX recipients (50.4 ± 9.1 years, one female/five male), CD14+CD16+ monocytes were analyzed longitudinally before transplantation and at four time points (14 days, 30 days, 90 days and >540 days (mean 641), [range 546–860]) after transplantation. All six patients received FK, AZA and MP as maintenance immunosuppressive drugs and daclizumab as an induction agent. MP treatment was started with i.v. boli (day 0: 750 mg, day 1: 100 mg), and was gradually tapered to 20 mg MP on day 14, 4 mg on day 90 and 2–4 mg, thereafter.
Informed consent was obtained from all study subjects and the study protocol was approved by the local ethics committee.
TX (−MP, N = 18)
and HMG-CoA-reductase (statin) as well as angiotensin receptor blocker (ARB) medication to 31 TX patients (age: 55.9 ± 12.9 years). The 31 control subjects (age: 57.9 ± 12.9 years) were recruited from a cohort of 64 patients undergoing hemodialysis treatment at our institution. Blood of these 31 patients was flow-cytometrically analyzed between November 2003 and April 2004, the time period in which the transplant patients have been analyzed. The HD patients were treated with synthetic membranes (polyamide or polysulfone) thrice weekly in our dialysis unit.
Methods Blood was drawn from all subjects under standardized conditions. Plasma glucose, creatinine and total cholesterol were obtained using standard techniques. Adjusted estimated glomerular filtration rate (GFR) was calculated using the modification of diet in renal disease (MDRD) study equation 3 (14). Patients with self-reported diabetes mellitus (DM), with a nonfasting blood sugar level of >200 mg/dL, with a fasting blood sugar lever of >126 mg/dL or with current use of hypoglycemic medication were categorized as diabetic (TX: N = 26, HD: N = 28). Patients were categorized as active smokers if they were current smokers or had stopped smoking 5
40
da
da 90
30
da
ys
ys da 14
Tx e be fo r
%CD14+CD16+
2006/2007
25 14 15 2 31
17 12 4 9 28
20
0
B
2003/2004 Cyclosporine (CyA) Tacrolimus (FK) Azathioprine (AZA) Mycophenolate mofetil (MMF) Steroid
**
TX+MP (2003/2004) TX-MP (2006/2007) TX-MP (2003/2004) TX+MP (2006/2007)
10 5 0
Figure 2: (A) Time course of the frequency of CD16+ monocytes in six kidney transplant recipients. All patients were treated with tacrolimus, azathioprine, methylprednisolone and induction therapy with daclizumab. Indicated is the dosage of methylprednisolone (MP) for each time point of measurement. Data were compared by one-way ANOVA for matched observations with Newman–Keuls as posttest (∗∗∗ = p < 0.001). (B) Change in the methylprednisolone therapy scheme in five TX patients and changes in the frequency of CD14+CD16+ monocytes after 2.5 years of follow-up (white squares = %CD14+CD16+ of patients with methylprednisolone medication (+MP; N = 4); grey squares the same patients after drug removal (−MP); grey triangle = one patient without MP treatment in 2003; white triangle = the same patient with methylprednisolone therapy in 2006. Data were compared by t-test for paired samples (∗∗ = p < 0.01).
withdrawn. The frequency of CD14+CD16+ in these patients significantly increased after removal of steroids (left hand side, Figure 2B) and vice versa, one patient who was switched from glucocorticoid-free immunosuppression to glucocorticoid-based immunosuppression had a reduced American Journal of Transplantation 2008; 8: 103–110
frequency of CD14+CD16+ monocytes (right hand side, Figure 2B). CD14+CD16+ monocytes, cardiovascular risk factors and subclinical atherosclerosis In TX patients, IMT significantly correlated with PP (PP: r = 0.39, p < 0.001) and with age (r = 0.49, p < 0.001). There was no significant association between IMT and other classical cardiovascular risk factors (cholesterol, triglycerides, presence of DM and nicotine abusus). The inflammatory marker CRP did not correlate with the frequency of CD16+ monocytes (r = −0.0503, p = 0.609) and statin therapy did apparently not affect the frequency of the CD16+ monocyte subpopulations (%CD14+CD16+: no statin: 5.5 ± 2.4 vs. statin: 6.2 ± 2.7; p = 0.275; %CD14 + 16+: no statin: 8.9 ± 7.0 vs. statin: 9.4 ± 6.1, p = 0.750). TX patients who had suffered cardiovascular events (CVD+: N = 22) had an increased IMT in comparison to patients without CVD in their history (CVD−; N = 73; CVD+: 0.82 ± 0.16 mm vs. CVD−: 0.68 ± 0.10 mm, p = 0.001). IMT significantly correlated with the frequency of CD14+CD16+ monocytes (r = 0.34; p < 0.001), but not with CD14++CD16+ monocytes. These correlations remained significant when absolute cells numbers of CD14+CD16+ monocytes instead of the relative percentage of this subset among all monocytes were considered. Table 5: Comparison of baseline characteristics of 41 kidney transplant recipients evaluated in 2003/2004 and 2006/2007
CRP (mg/L) Cholesterol (mg/dL) Triglycerides (mg/dL) Leucocyte/lL ×103 Monocyte/lL × 103 %CD14+CD16+ IMT (mm) Diuretics (%) ARB (%) ACE I (%) CCB (%)
TX 2003/2004
TX 2006/2007
p-Value
8.6 ± 12.5 206.6 ± 39.0 167.8 ± 121.6 7.7 ± 1.7 0.6 ± 0.2 9.8 ± 6.3 0.71 ± 0.12 41.5 34.1 22.0 46.3
4.6 ± 6.8 207.5 ± 46.3 178.4 ± 101.1 7.8 ± 2.1 0.7 ± 0.3 10.1 ± 6.1 0.80 ± 0.38 53.7 48.8 22.0 41.5
0.031 0.925 0.739 0.820 0.780 0.805 0.160 0.507 0.368 1.000 0.824
ARB = angiotensin receptor blocker; ACE I = ACE inhibitor; CCB = calcium channel blocker.
107
Ulrich et al.
Frequency of CD14+CD16+ (%)
A 35 30 25 20
- MP r= 0.55 P= 0.02 + MP r= 0.19 P= 0.10
15 10 5 0 0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
IMT (mm)
35
35
30
30
25
25
20 15
-MP r=0.593 P=0.071
10 5
+MP r=0.101 P=0.589
0 0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
IMT [mm]
%CD14+CD16+
%CD14+CD16+
B
20 15
-MP r=0.405 P=0.164
10 5
+MP r=0.005 P=0.978
0 0.4
0.8
1.2
1.6
2.0
2.4
IMT [mm]
Figure 3: (A) Correlation of intima-media thickness (IMT) and CD14+CD16+ monocytes in 18 renal transplant recipients without methylprednisolone (−MP; grey squares) and 77 renal transplant recipients with methylprednisolone (+MP; white squares). (B) Correlation of intima-media thickness (IMT) and CD14+CD16+ monocytes in 41 renal transplant recipients. Left hand side = data of patients in 2003/2004 (10 patients had no methylprednisolone therapy) (−MP; grey squares); MP was part of the drug regime in 31 patient (+MP; white squares). Right hand side = data of the same patients in 2006/2007 (13 patients without/28 patients with MP).
The frequency of CD14+CD16+ monocytes significantly correlated with PP (r = 0.24, p = 0.017); there were no other associations between CD14+CD16+ monocytes and classical cardiovascular risk factors. TX patients with CVD tended to have a higher frequency of CD14+CD16+ monocytes than those without CVD, but this difference did not reach statistical significance (CVD+: 11.1 ± 8.4% vs. CVD−: 8.4 ± 6.0%, p = 0.109). When stratifying transplant recipients by MP intake, the correlation between the frequency of CD14+CD16+ monocytes and IMT was restricted to patients without MP intake, whereas monocyte subsets and IMT did not correlate in patients receiving MP treatment (Figure 3A). After 2.6 years of the first analysis, IMT and the frequency of monocyte subsets were again determined in 41 of the 108
95 TX patients. The intra-assay variability between the two IMT measurements was 9.0 ± 12.2%. The frequency of CD14+CD16+ monocytes of these patients was not significantly different (%CD14+CD16+: 9.7 ± 6.8 [2003/2004] vs. 10.1 ± 6.1 [2006/2007], p = 0.805). Interestingly, we could confirm the association of IMT and the frequency of CD14+CD16+ monocyte (r = 0.31; p = 0.049). As depicted in Figure 3B, the association between IMT and the frequency of CD14+CD16+ is biased by steroid treatment, a finding also seen in the first analysis. Interestingly, the inflammatory marker CRP was reduced by half in TX patients analyzed in 2006/2007 in comparison to the data of patients measured in 2003/2004 (Table 5). Again, there were no correlation between CRP and proinflammatory CD14+CD16+ monocytes (r = −0.07, p = 0.653). American Journal of Transplantation 2008; 8: 103–110
Proinflammatory CD14+CD16+ Monocytes in Renal Transplant Patients Table 6: Linear regression model evaluating the association of IMT, classical cardiovascular risk factors and the frequency of CD14+CD16+ monocytes Dependent variable: IMT (mm)
Constant Age (years) CD14+CD16+ (%) PP (mmHg)
Coefficient B
T-statistics
p-Value
0.0365 0.0004 0.0005 0.0002
6.253 4.299 2.687 2.170
