- Email: [email protected]
Kinetics of Lipoprotein(a) in patients undergoing weekly lipoprotein apheresis for Lp(a) hyperlipoproteinemia
Atherosclerosis Supplements 30 (2017) 209e216 www.elsevier.com/locate/atherosclerosis
Kinetics of Lipoprotein(a) in patients undergoing weekly lipoprotein apheresis for Lp(a) hyperlipoproteinemia S. Tselmin a,*, G. Mu¨ller b, U. Schatz a, U. Julius a, S.R. Bornstein a, B. Hohenstein
a
a
Extracorporeal Treatment and Lipoprotein Apheresis Center, Department of Internal Medicine III, University Hospital Carl Gustav Carus at the Technische Universita¨t Dresden, Germany b Center for Evidence-Based Healthcare, Medizinische Fakulta¨t Carl Gustav Carus at the Technische Universita¨t Dresden, Fetscherstr. 74, 01307 Dresden, Germany
Abstract Introduction: Lipoprotein apheresis (LA) represents the only effective therapeutic option for patients with elevated Lipoprotein(a) (Lp(a)) levels. We aimed at analyzing the Lp(a) reduction, rebound rates as well as mean interval values between two weekly apheresis sessions, since this might be important for the prediction of the residual cardiovascular risk and development of individualized approaches for this special therapeutic strategy. Materials and methods: 20 patients under weekly and 2 patients under twice weekly apheresis were included. We measured serum concentrations of Lp(a), total, LDL-, HDL - cholesterol and triglycerides daily over 7 days after single LA sessions. Results: Mean Lp(a) levels was 158.1 69.82 nmol/l before the LA session, decreased acutely by 76 7% and increased to 97 13% of the baseline value within 7 days in patients under weekly treatment. By mathematical modeling, the acute Lp(a) reduction can be calculated from the function: y (nmol/l) ¼ 3.415 þ 0.738 * x (R2 ¼ 0.970), where x is the baseline Lp(a) value. The recovery rate can be predicted from the equation: y (%) ¼ 22.49 þ 18.64 * x e 1.14 * x2 (R2 ¼ 0.874), where x is the day after apheresis. The empirical formula for the mean interval value is: y (nmol/l) ¼ x 12, where x is the absolute reduction in nmol/l. Conclusion: We modeled for the first time equations to predict the course of Lp(a) serum levels under weekly LA which are simple, reliable and enable the development of optimal individualized protocols of this costly lipid lowering therapy. Ó 2017 Elsevier B.V. All rights reserved.
Keywords: Lipoprotein(a); Lipoprotein apheresis; Kinetics
1. Introduction LA represents the only effective therapy option for severe dyslipidemias that are resistant to life-style intervention and potent lipid lowering medication [1,2]. LA is of special importance for patients with elevated serum levels of Lp(a) [3], a well-known cardiovascular risk factor [4] associated not only with a high incidence of coronary
* Corresponding author. E-mail address: [email protected] (S. Tselmin). http://dx.doi.org/10.1016/j.atherosclerosissup.2017.05.033 1567-5688/Ó 2017 Elsevier B.V. All rights reserved.
heart disease (CHD) [5e7], but also of stroke [8], peripheral artery disease [9] and aortic valve stenosis [10]. The establishment of regular reimbursement for extracorporeal elimination of atherogenic lipoproteins and particularly the decision of the German Federal Joint Committee to accept Lp(a)-hyperlipoproteinemia as separate indication for LA [11] gave a great impulse to conducting trials on this topic that resulted in convincing findings. Observations using intravascular ultrasound in 15 coronary patients showed not only a halted progression, but a regression of atherosclerotic lesions within 18 months of specific apolipoprotein(a) apheresis [12]. A retrospective analysis in 120 patients demonstrated a decrease in CHD
210
S. Tselmin et al. / Atherosclerosis Supplements 30 (2017) 209e216
incidence from 1.056 to 0.144 events per patient [13] and a larger multicenter prospective study reported the drastic drop of annual cardiovascular incidence from 0.58 to 0.11 within five years during the LA in patients with isolated elevated Lp(a) [14]. Intriguingly, the decline of cardiovascular event rates during regular LA has been shown to being significantly greater in patients with elevated Lp(a) (with and without increased LDL-C) than in those with isolated LDLhypercholesterinemia (83.5 and 83% vs. 54% respectively) [15]. Evidence of these studies resulted in increasing numbers of patients with Lp(a)-hyperlipidemia in our and other apheresis centers. Data on optimal treatment modalities in terms of optimized reduction rates, absolute Lp(a) values that might be targeted within a single apheresis session or treatment frequency are poor, while we are facing a large number of patients with a spectrum of lipidologic disturbances and overall risk. From this point, the application of so far uniform treatment protocols might be further developed towards individual therapeutic strategies for these patients. One of the major issues under apheresis treatment is the rebound of Lp(a) serum concentrations during treatment intervals. While pretreatment values will depend on their therapeutic decrease following LA [16], other factors [17,18] including stimulated synthesis, kidney disease [19], postmenopausal age [20] might have an influence. Genetic studies reveal that the individual lifetime exposure towards LDL-C plays a relevant role and it is likely that the same is true for Lp(a). Therefore, it would be of high importance to evaluate changes to LDL-C and Lp(a) exposure under apheresis conditions. In this study we aimed at investigating detailed kinetics of the two cardiovascular risk factors LDL-C and Lp(a) in patients under regular and weekly apheresis treatment to describe a rebound curve and to model the formulas for prediction of the relevant LDL-C and Lp(a) kinetics. 2. Materials and methods 2.1. Participants and data collection The data collection and evaluation was approved by Dresden Ethics Committee and all study participants gave their written informed consent. Twenty-two patients (8 females, mean age 60.2 13.4 years) with indications to extracorporeal elimination of the elevated Lp(a) [11] were treated regularly with LA at the Extracorporeal Treatment and Apheresis Center at the University Hospital Carl Gustav Carus in Dresden. All of them received the maximum tolerated lipid lowering medication to achieve the LDL-C goal values for high risk persons according to the guidelines [21]. In 20 patients under weekly apheresis the serum concentrations of Lp(a), total, LDL (LDL-C), HDL (HDL-
C) cholesterol and triglycerides (TG) and hematocrit values were measured daily over a one week interval covering the time between the start of one LA sessions and the start of the next session 7 days later. Two persons with a rapid progressive course of severe coronary and peripheral atherosclerotic complications received apheresis twice weekly and their lipids levels were assessed before and just after each of two successively performed treatments within 7 days. Blood samples were drawn from peripheral veins, collected in standard vacutainers (Sarstedt AG, Nu¨rnbrecht, Germany) and analyzed in routine fashion at the Clinical Laboratory of the University Hospital in Dresden. 2.2. LA methods We performed LA with six different methods that have been described before [3]: TheraSorbÒ LDL (Miltenyi Biotec GmbH, Bergisch-Gladbach, Germany) (n ¼ 2), DALI (Fresenius Medical Care GmbH, Bad Homburg, Germany) (n ¼ 2), MONET (Fresenius Medical Care GmbH, Bad Homburg, Germany) (n ¼ 4), HELP (B Braun Avitum AG, Melsungen, Germany) (n ¼ 4), Lipidfiltration (Diamed Medizintechnik GmbH, Cologne, Germany) (n ¼ 4) and Liposorber D (Kaneka Corporation, Japan) (n ¼ 6). 2.3. Biochemical analysis Serum levels of Lp(a) were measured with an assay of the second generation: Immunoassay for Lipoprotein(a) (WHO/IFCC International Reference Reagent) e SRM 2B using the device Roche/Hitachi Cobas c 701/702. The antibodies identifying the not repetitive Kringle IV e domains enabled a more precise measurement of Lp(a) particles concentrations. Cholesterol Gen.2, LDL-Cholesterol plus 2nd generation, HDL-Cholesterol plus 3rd generation and Triglycerides tests (Roche/Hitachi Cobas c 701/702) were used to assess TC, LDL-C, HDL-C and TG serum levels. Hematocrit values from EDTA samples were measured using a hematology analyzer Sysmex XE-5000. Our initial approach aimed to use values following hematocrit correction. However, initial data analysis revealed that the influence of blood dilution or hematocrit correction respectively was very low (below 3e5%) and therefore neglectable. This led to the decision to perform all further evaluations on basis of the primary values. 2.4. Data analysis and statistics Continuous variables are expressed as means SD. To identify the difference between the mean values we have used Wilcoxon Signed Ranks Test for paired samples and the results were Bonferroni-adjusted corresponding to the spaces of time that were assessed. The interactions between the Lp(a) and LDL-C concentration over the time have been assessed with a General Linear Model with
S. Tselmin et al. / Atherosclerosis Supplements 30 (2017) 209e216
211
repeated measures. The Lp(a) and LDL-C reduction and rebound curves have been modeled by means of correlation analysis, which were performed with SPSS 23 software package.
value can be predicted from the following quadratic equation as depicted in Fig. 2:
3. Results
where the independent variable (x) is the day after apheresis (R2 ¼ 0.874, p < 0.001).
3.1. Course of Lp(a) concentrations between two LA sessions The initial mean Lp(a) serum level was 158.1 69.82 nmol/l and acutely decreased by 76 7% to 38.1 20.39 nmol/l after a single apheresis session. After 7 days, Lp(a) concentrations reached on average 97 13% of the baseline. In 6 patients, Lp(a) concentration reached the initial value on the sixth day and one of them showed 102% of the pretreatment Lp(a) level on the fourth day after the therapy. Daily measured mean Lp(a) values and the mean percentage rates of deviation from the initial level differed from one day to another up to the day 6 (p < 0.006, Bonferroni adjusted). The values of days 6 and 7 were not statistically different (p ¼ 1.00, Bonferroni adjusted) (Fig. 1). Based on this time course, the acute Lp(a) reduction to the end of a single apheresis session can be calculated from the linear function: y (nmol/l) ¼ 3.415 þ 0.738 * x, where x is the baseline Lp(a) value (R2 ¼ 0.970, p < 0.001). The rebound rate of Lp(a) serum levels following apheresis during a 7 days interval with regard to the initial
yð%Þ ¼ 22:49 þ 18:64 x 1:14 x2 ;
3.2. Modeling of a formula for the interval mean Lp(a) value between two consecutive LA sessions Regression analysis showed an almost linear correlation between the absolute acute Lp(a) reduction to the end of a single LA session and the interval mean Lp(a) value over the time to the next treatment. The interval mean Lp(a) value between two consecutive LA can be predicted from the following equation as depicted in Fig. 3: yðnmol=lÞ ¼ 5:1 þ 0:941 x;
R2 ¼ 0:907; p < 0:001 ;
where x is the absolute acute Lp(a) reduction after a single LA session. Since the complexity of this formula has drawbacks in daily routine use, we next aimed to develop a rule-of-thumb equation. Therefore, we rounded 0.941 up to 1.000. The mean absolute Lp (a) reduction in 20 patients was 120 nmol/l. So we multiplied the difference
Fig. 1. The changes of Lp(a) concentrations between two LA sessions. The initial mean Lp(a) serum level was acutely decreased by 76 7% after a single apheresis session. Under weekly LA treatment Lp(a) concentrations achieved on average 97 13% of the baseline before the next session within one week. The daily measured mean percentage rates of deviation from the initial level differed from one another up to the sixth day (p < 0.006, Bonferroni adjusted). The values of the sixth and seventh days were not statistically different (p ¼ 1.00, Bonferroni adjusted).
212
S. Tselmin et al. / Atherosclerosis Supplements 30 (2017) 209e216
Fig. 2. Graphic model of the Lp(a) serum concentrations recovery after a single LA session. The recovery rate of Lp(a) serum levels within one week after a single apheresis session with regard to initial value can be predicted from the following quadratic equation: yð%Þ ¼ 22:49 þ 18:64x 1:14x2 ; where the independent variable (x) is the day after apheresis. The quadratic regression equation showed p < 0.001. Coefficient of determination R2 ¼ 0.874 demonstrated the good model fit for a prediction of percentage deviation from the pretreatment Lp(a) concentration dependent on the day after the treatment.
Fig. 3. Mean Lp(a) value between two consecutive lipoprotein apheresis sessions. The interval mean Lp(a) value between two consecutive lipoprotein apheresis can be predicted from the following equation: yðnmol=lÞ ¼ 5:1 þ 0:941 x; ðR2 ¼ 0:907; p < 0:001Þ; where x is the absolute acute Lp(a) reduction after a single LA session.
S. Tselmin et al. / Atherosclerosis Supplements 30 (2017) 209e216
0.941e1 ¼ 0.059 * 120 ¼ 7.08 and added it to 5.1 from the above mentioned equation. That resulted in the constant value of 12.18, giving the empirical formula: yðnmol=lÞ ¼ x 12; where x is the absolute acute Lp(a) reduction after a single LA session. The testing of this equation in all 20 patients demonstrated very small deviation from the real mean values as well the values predicted from the initial linear formula. We have also tested the formula: y ¼ (x þ z)/2, where x is the Lp(a) level to the end of apheresis session and z is the Lp(a) value after 7 days (before the next session). However, the values predicted from this equation were significantly lower compared to the real Lp(a) levels (Fig. 4). In both patients who had a severe progress of atherosclerotic complications and received extracorporeal therapy twice a week, the Lp(a) level exceeded the initial value by 7% and 44% respectively already on the fourth day after a single treatment. 3.3. Course of LDL-C concentrations between two LA sessions The initial mean LDL-C serum level was 2.2 0.72 mmol/l and was decreased by 70 9% (to 0.61 0.19 mmol/l) following a single apheresis session. The LDL-C concentration achieved 98 15% of the initial value on day 7 after apheresis (before the next session). The mean LDL-C values or the mean percentage rates of
213
deviation from the initial level assessed on the fourth and fifth as well as on the sixth and seventh days did not differ significantly from each other (p ¼ 0.09 and 1.00 respectively, Bonferroni adjusted). The recovery rate of LDL-C serum levels within one week following the initial apheresis session with regard to initial value can be predicted from the following quadratic equation: yð%Þ ¼ 27:16 þ 21:17 x 1:58 x2 ; where the independent variable (x) is the day after apheresis (p < 0.001, R2 ¼ 0.771). 3.4. Comparison of LDL-C and Lp(a) kinetics Under a single apheresis session Lp(a) serum concentrations decreased farther than those of LDL-C (76 7% vs. 70 9%, p ¼ 0.0036). The levels of both lipoproteins recovered virtually in parallel up to the fifth day and then their values converged to the baseline (Fig. 5). Testing of interaction between the two rebound curves using General Linear Model with repeated measures revealed no significant difference with respect to the factors time and both lipoprotein values (Hotelling’s Trace: p ¼ 0.225). The Lp(a) regression curve lay within the 95% e confidential interval of the LDL-C regression curve so that these increases were statistically similar. 4. Discussion We studied the course of serum levels of the atherogenic lipoproteins between two consecutive apheresis sessions
Fig. 4. The optimized equation compared to exact regression and mean values. The testing of the optimized equation ðyðnmol=lÞ ¼ x 12Þ in all 20 patients demonstrated very small deviation from the real mean values as well the values predicted from the regression equation: yðnmol=lÞ ¼ 5:1 þ 0:941 x; where x is the absolute acute Lp(a) reduction after a single LA session. We have also tested the formula: y ¼ ðx þ zÞ=2; where x is the Lp(a) serum level to the end of apheresis session and z is the level in 7 days before the next session. The values predicted from this equation were significantly lower in comparison to the real Lp(a) levels.
214
S. Tselmin et al. / Atherosclerosis Supplements 30 (2017) 209e216
Fig. 5. Comparison of the Lp(a) and LDL-C concentrations during time course. The levels of both lipoproteins recovered in parallel up to day 5 before converging towards their maximum. The confidential interval for LDL-C is slightly wider than that of Lp(a). The testing of interaction between the two rebound curves using General Linear Model with repeated measures revealed no significant difference with respect to the factors time and both lipoprotein values (Hotelling’s Trace: p ¼ 0.225).
focusing on the rebound curve of Lp(a). Most patients had isolated Lp(a)-hyperlipoproteinemia with LDL-C lower than 2.6 mmol/l (primarily or due to lipid-lowering drugs and/or regular extracorporeal therapy). Nevertheless LDLC changes were also analyzed to compare them with that of Lp(a). The serum levels of TG, TC and HDL-C were before apheresis in the normal range and their reductions were seen as expected. A single LA treatment caused a sharp drop in Lp(a) serum levels (between 58 and 88%) confirming previous findings [22]. The rebound rate could be best predicted by the quadratic equation: y (%) ¼ 22.49 þ 18.64 * x 1.14 * x2, demonstrating the gradual increase up to 97% of the initial value before the next apheresis session within 7 days. The coefficient of determination R2 ¼ 0.874 demonstrated a good model fit for a prediction of percentage deviation from the pretreatment Lp(a) concentration dependent from the day after the treatment (x). The time course of LDL-C levels was similar to that of Lp(a). This may be of importance for the design of Lp(a) targeting apheresis end point studies, because most systems effectively remove both atherogenic lipoproteins. The first analysis of the serum levels under extracorporeal methods was conducted in 1977 in 3 hyperbetalipoproteinemic patients, who received daily plasmapheresis for 5e9 days [23]. Interestingly, despite an entirely different study design the cholesterol concentrations in a non-homozygous patient returned to the pretreatment values within 7 days after cessation of treatment course which is in accordance with our data. On the other hand, only cholesterol, but not Lp(a) levels were analyzed and in contrast to lipoprotein apheresis, plasma exchange resulted in contraction of plasma volume that had to be considered in the analysis. Our findings also confirmed the results of a Japanese group on Lp(a) kinetics performed in 13 patients who were
treated with dextran sulfate columns [24]. They demonstrated similar rebound curves with an increase nearly to the baseline within 7 days after apheresis. They showed a significant slowdown of LDL-C recovery and at the same time an increase of Lp(a) levels in patients undergoing pravastatin treatment. The curves for both lipoproteins in our study did not differ, while the measured daily LDL-C values were elevated despite the fact that all patients received the maximum tolerated lipid lowering therapy. It can be partly attributed to lower pretreatments LDL-C values in our patients (2.3 mmol/l vs. 3.7 mmol/l in patients of the aforementioned study) resulting in only slight fluctuations within the normal range. In contrast to the Japanese trial which was conducted in 1993, we would not pause statins in persons with high cardiovascular risk for testing effects of a lipid-altering medication on rebound rates. Patients from the most recent study on this topic also took statins and underwent a biweekly LA [25]. Acute reduction rates of Lp(a) and LDL-C were similar to our data and also the rebound curves of both lipoproteins were not significantly different, but the rebound to baseline took longer than that in our patients. Factors influencing the rebound of lipoproteins after LA are of importance. Basing on the results of the Kroon’s study [25] and our findings it is prudent to note the following evidence: the longer the time between two sessions is, the longer the recovery time for the serum concentrations of removed lipoproteins will be. It is important to mention that some research groups demonstrated no increase in the cholesterol synthesis after LA. In rats it was shown that the elevated cholesterol production was associated with albumin removal, caused by plasmapheresis [26]. Since a previous analysis from our group demonstrated a maximum 15% protein and albumin reduction, this might lead to a limited induction of
S. Tselmin et al. / Atherosclerosis Supplements 30 (2017) 209e216
cholesterol synthesis [27]. On the other hand Pfohl reported on transient stimulating of cholesterol biosynthesis after LDL-apheresis despite concomitant statin-therapy and the degree of stimulation was inversely related to the level to which the LDL-C was reduced [28]. The frequency of LA therapy might play a relevant role. Our analysis included two patients receiving apheresis twice weekly. In both patients we noted a rapid rebound of Lp(a) over baseline during the 4 days following apheresis, i.e. to the next treatment. Most likely this finding relates to the much higher Lp(a) levels these subjects had without LA, but it is also related to the high risk both subjects derived from their very high Lp(a) levels. Without doubt, it will be of interest to analyze more individuals with a necessity for a higher treatment frequency. Of note, our model to predict the Lp(a) rebound considered only the day following LA, not including the values before and at the end of a single session or the reduction rate that makes the calculation easier than in the papers mentioned above. In the same time the “R” parameter representing the impact of the factor time on lipids concentration was high (R2 ¼ 0.874, therefore R ¼ 0.935) for our formula, comparable with Apstein’s equation (0.900), though it was modeled for plasmapheresis and not for Lp(a) [23] and exceeded that from Kroon’s formula (0.74) derived from an analysis having a similar design to our study [25]. The linear correlation between the acute Lp(a) reduction and interval mean Lp(a) level revealed by the regression analysis enabled us to develop a very simply and reliable equation to predict this value without integration of the area under the rebound curve. This empirical formula is much more practicable for daily clinical routine, nevertheless enabling optimized and individualized protocols for this costly extracorporeal therapy. As a major drawback, this study had a monocentric design and therefore a limited number of participants. On the other hand, this enabled a very homogeneous preanalytic processing of samples. 5. Conclusions 1. A higher Lp(a) level before LA apheresis leads to a greater absolute decrease at the end of the treatment; 2. the rebound of Lp(a) after LA apheresis is best described by a square function: y (%) ¼ 27.16 þ 21.17 * x e 1.58 * x2, where the independent variable (x) is the day after apheresis; 3. the interval mean Lp(a) value between two weekly apheresis sessions can be predicted from an easy and reliable empiric formula: y (nmol/l) ¼ x 12, where x is the absolute acute Lp(a) reduction after a single LA session; 4. the rebound courses of Lp(a) and LDL-C following LA are similar; 5. patients with high cardiovascular risk and higher LA intensity demonstrate a rapid rebound of serum Lp(a)
215
concentrations; the causality of this association should be investigated in further studies and could contribute to carrying out the individual strategies of the costly LA therapy. Conflict of interest S. Tselmin: no financial relationships or conflicts of interest regarding the content herein. G. Mu¨ller declared no conflict of interest. U. Schatz travel expenses by Sanofi-Aventis, Amgen, honoraria from Amgen, Sanofi-Aventis, Abbot, Fresenius. U. Julius travel expenses by Aegerion, Diamed, Fresenius Medical Care, Kaneka; honoraria from Aegerion, Amgen, Chiesi, Sanofi, Kaneka, Diamed, Fresenius Medical Care. SR. Bornstein declared no conflict of interest. B. Hohenstein received honoraria, travel support and project funding from Miltenyi Biotec, B Braun Avitum, Fresenius Medical Care and Kaneka. Author contributions S. Tselmin: data gathering and analyzing, writing the draft of publication. G. Mu¨ller: statistical data processing. U. Schatz: data gathering, data interpretation, draft editing. U. Julius: data interpretation and editing. SR. Bornstein, B. Hohenstein: idea of the study, data interpretation and editing of the manuscript. Acknowledgements The authors are grateful for the excellent work of the staff of the Extracorporeal Treatment and Apheresis Center at the University Hospital Carl Gustav Carus Dresden. The skilled help of Kerstin Haaser is gratefully acknowledged. This study was supported by a research grant from B. Braun Avitum to BH. References [1] Julius U, Tselmin S, Fischer S, Passauer J, Bornstein SR. The Dresden Apheresis Center e experience with LDL apheresis and immunoadsorption. Atheroscler Suppl 2009;10(5):12e6. [2] Julius U, Frind A, Tselmin S, Kopprasch S, Poberschin I, Siegert G. Comparison of different LDL apheresis methods. Expert Rev Cardiovasc Ther 2008;6(5):629e39. [3] Waldmann E, Parhofer KG. Lipoprotein apheresis to treat elevated lipoprotein(a). J Lipid Res 2016 Oct;57(10):1751e7. [4] Nordestgaard BG, Chapman MJ, Ray K. Lipoprotein(a) as a cardiovascular risk factor: current status. Eur Heart J 2010;31(23): 2844e53. [5] Tselmin S, Julius U, Mueller G, Fischer S, Bornstein SR. Cardiovascular events in patients with increased lipoprotein (a) e retrospective data analysis in an outpatient department of lipid disorders. Atheroscler Suppl 2009;10(5):79e84.
216
S. Tselmin et al. / Atherosclerosis Supplements 30 (2017) 209e216
[6] Kamstrup PR, Benn M, Tybjærg-Hansen A, Nordestgaard BG. Extreme lipoprotein(a) levels and risk of myocardial infarction in the general population: the Copenhagen City Heart Study. Circulation 2008;117(2):176e84. [7] Danesh J, Collins R, Peto R. Lipoprotein(a) and coronary heart disease. Meta-analysis of prospective studies. Circulation 2000; 102(10):1082e5. [8] Stra¨ter R, Becker S, von Eckardstein A, Heinecke A, Gutsche S, Junker R, et al. Prospective assessment of risk factors for recurrent stroke during childhood - a 5-year follow-up study. Lancet 2002; 360(9345):1540e5. [9] Dieplinger B, Lingenhel A, Baumgartner N, Poelz W, Dieplinger H, Haltmayer M, et al. Increased serum lipoprotein(a) concentrations and low molecular weight phenotypes of apolipoprotein(a) are associated with symptomatic peripheral arterial disease. Clin Chem 2007;53(7):1298e305. [10] Kamstrup PR, Tybjærg-Hansen A, Nordestgaard BG. Elevated lipoprotein(a) and risk of aortic valve stenosis in the general population. J Am Coll Cardiol 2014;63(5):470e7. [11] Bundesministerium fu¨r Gesundheit. Bekanntmachung eines Bes¨ nderung chlusses des Gemeinsamen Bundesausschusses u¨ber eine A der Richtlinie Methoden vertragsa¨rztliche Versorgung: Apherese bei isolierter Lp(a)-Erho¨hung. BAnz, 2008;138:3321. [12] Safarova MS, Ezhov MV, Afanasieva OI, Matchin YG, Atanesyan RV, Adamova IY, et al. Effect of specific lipoprotein(a) apheresis on coronary atherosclerosis regression assessed by quantitative coronary angiography. Atheroscler Suppl. 2013 Jan;14(1): 93e9. [13] Jaeger BR, Richter Y, Nagel D, Heigl F, Vogt A, Roeseler E, et al. Longitudinal cohort study on the effectiveness of lipid apheresis treatment to reduce high lipoprotein(a) levels and prevent major adverse coronary events. Nat Clin Pract Cardiovasc Med 2009;6(3): 229e39. [14] Roeseler E, Julius U, Heigl F, Spitthoever R, Heutling D, Breitenberger P, et al. Pro(a)LiFe-Study Group. Lipoprotein apheresis for lipoprotein(a)-associated cardiovascular disease: prospective 5 years of follow-up and apo(a) characterization. Arterioscler Thromb Vasc Biol. 2016 Sep;36(9):2019e27. [15] von Dryander M, Fischer S, Passauer J, Mu¨ller G, Bornstein SR, Julius U. Differences in the atherogenic risk of patients treated by lipoprotein apheresis according to their lipid pattern. Atheroscler Suppl 2013;14:39e44. [16] Kassner U, Vogt A, Rosada A, Barz F, Giannakidou-Jodan E, Berthold HK, et al. Designing a study to evaluate the effect of apheresis in patients with elevated lipoprotein(a). Atheroscler Suppl 2009;10(5):85e8.
[17] Armstrong VW, Schleef J, Thiery J, Muche R, Schuff-Werner P, Eisenhauer T, et al. Effect of HELP-LDL-apheresis on serum concentrations of human lipoprotein(a): kinetic analysis of the posttreatment return to baseline levels. Eur J Clin Invest 1989;19(3): 235e40. [18] Groß E, Hohenstein B, Julius U. Effects of Lipoprotein apheresis on the Lipoprotein(a) levels in the long run. Atheroscler Suppl 2015 May;18:226e32. [19] Kronenberg F. Causes and consequences of lipoprotein(a) abnormalities in kidney disease. Clin Exp Nephrol 2014;18(2):234e7. [20] Julius U, Fritsch H, Fritsch W, Rehak E, Fu¨cker K, Leonhardt W, et al. Impact of hormone replacement therapy on postprandial lipoproteins and lipoprotein(a) in normolipidemic postmenopausal women. Clin Investig 1994;72(7):502e7. [21] Perk J, De Backer G, Gohlke H, Graham I, Reiner Z, Verschuren WM, et al. European Guidelines on cardiovascular disease prevention in clinical practice (version 2012): The Fifth Joint Task Force of the European Society of Cardiology and Other Societies on Cardiovascular Disease Prevention in Clinical Practice (constituted by representatives of nine societies and by invited experts). Atherosclerosis 2012 Jul;223(1):1e68. [22] Leebmann J, Roeseler E, Julius U, Heigl F, Spitthoever R, Heutling Dfor the Pro(a)Life-Study Group. Lipoprotein apheresis in patients with maximally tolerated lipid lowering therapy, Lp(a)hyperlipoproteinemia and progressive cardiovascular disease. Circulation 2013;128(24):2567e76. [23] Apstein CS, Zilversmit DB, Lees RS, George PK. Effect of intensive plasmapheresis on the plasma cholesterol concentration with familial hypercholesterolemia. Atherosclerosis 1978 Oct;31(2):105e15. [24] Koizumi J, Koizumi I, Uno Y, Inazu A, Kajinami K, Haraki T, et al. Reduction of lipoprotein(a) by LDL-apheresis using a dextran sulfate cellulose column in patients with familial hypercholesterolemia. Atherosclerosis 1993 Apr;100(1):65e74. [25] Kroon AA, Van’t Hof MA, Demacker PNM, Stalenhoef AFH. The rebound of lipoproteins after LDL-apheresis. Kinetics and estimation of mean lipoprotein levels. Atherosclerosis 2000;152:519e26. [26] Takeuchi N, Yamamura Y. The effect of plasmapheresis on cholesterol synthesis in the rat: relationship to protein synthesis. J Lab Clin Med 1972 May;79(5):801e13. [27] Julius U, Siegert G, Kostka H, Schatz U, Hohenstein B. Effect of different lipoprotein apheresis methods on serum protein levels. Atheroscler Suppl 2015 May;18:95e102. [28] Pfohl M, Naoumova RP, Klass C, Knisel W, Jakober B, Risler T, et al. Acute and chronic effects on cholesterol biosynthesis of LDLapheresis with or without concomitant HMG-CoA reductase inhibitor therapy. J Lipid Res 1994 Nov;35(11):1946e55.
Kinetics of Lipoprotein(a) in patients undergoing weekly lipoprotein apheresis for Lp(a) hyperlipoproteinemia S. Tselmin a,*, G. Mu¨ller b, U. Schatz a, U. Julius a, S.R. Bornstein a, B. Hohenstein
a
a
Extracorporeal Treatment and Lipoprotein Apheresis Center, Department of Internal Medicine III, University Hospital Carl Gustav Carus at the Technische Universita¨t Dresden, Germany b Center for Evidence-Based Healthcare, Medizinische Fakulta¨t Carl Gustav Carus at the Technische Universita¨t Dresden, Fetscherstr. 74, 01307 Dresden, Germany
Abstract Introduction: Lipoprotein apheresis (LA) represents the only effective therapeutic option for patients with elevated Lipoprotein(a) (Lp(a)) levels. We aimed at analyzing the Lp(a) reduction, rebound rates as well as mean interval values between two weekly apheresis sessions, since this might be important for the prediction of the residual cardiovascular risk and development of individualized approaches for this special therapeutic strategy. Materials and methods: 20 patients under weekly and 2 patients under twice weekly apheresis were included. We measured serum concentrations of Lp(a), total, LDL-, HDL - cholesterol and triglycerides daily over 7 days after single LA sessions. Results: Mean Lp(a) levels was 158.1 69.82 nmol/l before the LA session, decreased acutely by 76 7% and increased to 97 13% of the baseline value within 7 days in patients under weekly treatment. By mathematical modeling, the acute Lp(a) reduction can be calculated from the function: y (nmol/l) ¼ 3.415 þ 0.738 * x (R2 ¼ 0.970), where x is the baseline Lp(a) value. The recovery rate can be predicted from the equation: y (%) ¼ 22.49 þ 18.64 * x e 1.14 * x2 (R2 ¼ 0.874), where x is the day after apheresis. The empirical formula for the mean interval value is: y (nmol/l) ¼ x 12, where x is the absolute reduction in nmol/l. Conclusion: We modeled for the first time equations to predict the course of Lp(a) serum levels under weekly LA which are simple, reliable and enable the development of optimal individualized protocols of this costly lipid lowering therapy. Ó 2017 Elsevier B.V. All rights reserved.
Keywords: Lipoprotein(a); Lipoprotein apheresis; Kinetics
1. Introduction LA represents the only effective therapy option for severe dyslipidemias that are resistant to life-style intervention and potent lipid lowering medication [1,2]. LA is of special importance for patients with elevated serum levels of Lp(a) [3], a well-known cardiovascular risk factor [4] associated not only with a high incidence of coronary
* Corresponding author. E-mail address: [email protected] (S. Tselmin). http://dx.doi.org/10.1016/j.atherosclerosissup.2017.05.033 1567-5688/Ó 2017 Elsevier B.V. All rights reserved.
heart disease (CHD) [5e7], but also of stroke [8], peripheral artery disease [9] and aortic valve stenosis [10]. The establishment of regular reimbursement for extracorporeal elimination of atherogenic lipoproteins and particularly the decision of the German Federal Joint Committee to accept Lp(a)-hyperlipoproteinemia as separate indication for LA [11] gave a great impulse to conducting trials on this topic that resulted in convincing findings. Observations using intravascular ultrasound in 15 coronary patients showed not only a halted progression, but a regression of atherosclerotic lesions within 18 months of specific apolipoprotein(a) apheresis [12]. A retrospective analysis in 120 patients demonstrated a decrease in CHD
210
S. Tselmin et al. / Atherosclerosis Supplements 30 (2017) 209e216
incidence from 1.056 to 0.144 events per patient [13] and a larger multicenter prospective study reported the drastic drop of annual cardiovascular incidence from 0.58 to 0.11 within five years during the LA in patients with isolated elevated Lp(a) [14]. Intriguingly, the decline of cardiovascular event rates during regular LA has been shown to being significantly greater in patients with elevated Lp(a) (with and without increased LDL-C) than in those with isolated LDLhypercholesterinemia (83.5 and 83% vs. 54% respectively) [15]. Evidence of these studies resulted in increasing numbers of patients with Lp(a)-hyperlipidemia in our and other apheresis centers. Data on optimal treatment modalities in terms of optimized reduction rates, absolute Lp(a) values that might be targeted within a single apheresis session or treatment frequency are poor, while we are facing a large number of patients with a spectrum of lipidologic disturbances and overall risk. From this point, the application of so far uniform treatment protocols might be further developed towards individual therapeutic strategies for these patients. One of the major issues under apheresis treatment is the rebound of Lp(a) serum concentrations during treatment intervals. While pretreatment values will depend on their therapeutic decrease following LA [16], other factors [17,18] including stimulated synthesis, kidney disease [19], postmenopausal age [20] might have an influence. Genetic studies reveal that the individual lifetime exposure towards LDL-C plays a relevant role and it is likely that the same is true for Lp(a). Therefore, it would be of high importance to evaluate changes to LDL-C and Lp(a) exposure under apheresis conditions. In this study we aimed at investigating detailed kinetics of the two cardiovascular risk factors LDL-C and Lp(a) in patients under regular and weekly apheresis treatment to describe a rebound curve and to model the formulas for prediction of the relevant LDL-C and Lp(a) kinetics. 2. Materials and methods 2.1. Participants and data collection The data collection and evaluation was approved by Dresden Ethics Committee and all study participants gave their written informed consent. Twenty-two patients (8 females, mean age 60.2 13.4 years) with indications to extracorporeal elimination of the elevated Lp(a) [11] were treated regularly with LA at the Extracorporeal Treatment and Apheresis Center at the University Hospital Carl Gustav Carus in Dresden. All of them received the maximum tolerated lipid lowering medication to achieve the LDL-C goal values for high risk persons according to the guidelines [21]. In 20 patients under weekly apheresis the serum concentrations of Lp(a), total, LDL (LDL-C), HDL (HDL-
C) cholesterol and triglycerides (TG) and hematocrit values were measured daily over a one week interval covering the time between the start of one LA sessions and the start of the next session 7 days later. Two persons with a rapid progressive course of severe coronary and peripheral atherosclerotic complications received apheresis twice weekly and their lipids levels were assessed before and just after each of two successively performed treatments within 7 days. Blood samples were drawn from peripheral veins, collected in standard vacutainers (Sarstedt AG, Nu¨rnbrecht, Germany) and analyzed in routine fashion at the Clinical Laboratory of the University Hospital in Dresden. 2.2. LA methods We performed LA with six different methods that have been described before [3]: TheraSorbÒ LDL (Miltenyi Biotec GmbH, Bergisch-Gladbach, Germany) (n ¼ 2), DALI (Fresenius Medical Care GmbH, Bad Homburg, Germany) (n ¼ 2), MONET (Fresenius Medical Care GmbH, Bad Homburg, Germany) (n ¼ 4), HELP (B Braun Avitum AG, Melsungen, Germany) (n ¼ 4), Lipidfiltration (Diamed Medizintechnik GmbH, Cologne, Germany) (n ¼ 4) and Liposorber D (Kaneka Corporation, Japan) (n ¼ 6). 2.3. Biochemical analysis Serum levels of Lp(a) were measured with an assay of the second generation: Immunoassay for Lipoprotein(a) (WHO/IFCC International Reference Reagent) e SRM 2B using the device Roche/Hitachi Cobas c 701/702. The antibodies identifying the not repetitive Kringle IV e domains enabled a more precise measurement of Lp(a) particles concentrations. Cholesterol Gen.2, LDL-Cholesterol plus 2nd generation, HDL-Cholesterol plus 3rd generation and Triglycerides tests (Roche/Hitachi Cobas c 701/702) were used to assess TC, LDL-C, HDL-C and TG serum levels. Hematocrit values from EDTA samples were measured using a hematology analyzer Sysmex XE-5000. Our initial approach aimed to use values following hematocrit correction. However, initial data analysis revealed that the influence of blood dilution or hematocrit correction respectively was very low (below 3e5%) and therefore neglectable. This led to the decision to perform all further evaluations on basis of the primary values. 2.4. Data analysis and statistics Continuous variables are expressed as means SD. To identify the difference between the mean values we have used Wilcoxon Signed Ranks Test for paired samples and the results were Bonferroni-adjusted corresponding to the spaces of time that were assessed. The interactions between the Lp(a) and LDL-C concentration over the time have been assessed with a General Linear Model with
S. Tselmin et al. / Atherosclerosis Supplements 30 (2017) 209e216
211
repeated measures. The Lp(a) and LDL-C reduction and rebound curves have been modeled by means of correlation analysis, which were performed with SPSS 23 software package.
value can be predicted from the following quadratic equation as depicted in Fig. 2:
3. Results
where the independent variable (x) is the day after apheresis (R2 ¼ 0.874, p < 0.001).
3.1. Course of Lp(a) concentrations between two LA sessions The initial mean Lp(a) serum level was 158.1 69.82 nmol/l and acutely decreased by 76 7% to 38.1 20.39 nmol/l after a single apheresis session. After 7 days, Lp(a) concentrations reached on average 97 13% of the baseline. In 6 patients, Lp(a) concentration reached the initial value on the sixth day and one of them showed 102% of the pretreatment Lp(a) level on the fourth day after the therapy. Daily measured mean Lp(a) values and the mean percentage rates of deviation from the initial level differed from one day to another up to the day 6 (p < 0.006, Bonferroni adjusted). The values of days 6 and 7 were not statistically different (p ¼ 1.00, Bonferroni adjusted) (Fig. 1). Based on this time course, the acute Lp(a) reduction to the end of a single apheresis session can be calculated from the linear function: y (nmol/l) ¼ 3.415 þ 0.738 * x, where x is the baseline Lp(a) value (R2 ¼ 0.970, p < 0.001). The rebound rate of Lp(a) serum levels following apheresis during a 7 days interval with regard to the initial
yð%Þ ¼ 22:49 þ 18:64 x 1:14 x2 ;
3.2. Modeling of a formula for the interval mean Lp(a) value between two consecutive LA sessions Regression analysis showed an almost linear correlation between the absolute acute Lp(a) reduction to the end of a single LA session and the interval mean Lp(a) value over the time to the next treatment. The interval mean Lp(a) value between two consecutive LA can be predicted from the following equation as depicted in Fig. 3: yðnmol=lÞ ¼ 5:1 þ 0:941 x;
R2 ¼ 0:907; p < 0:001 ;
where x is the absolute acute Lp(a) reduction after a single LA session. Since the complexity of this formula has drawbacks in daily routine use, we next aimed to develop a rule-of-thumb equation. Therefore, we rounded 0.941 up to 1.000. The mean absolute Lp (a) reduction in 20 patients was 120 nmol/l. So we multiplied the difference
Fig. 1. The changes of Lp(a) concentrations between two LA sessions. The initial mean Lp(a) serum level was acutely decreased by 76 7% after a single apheresis session. Under weekly LA treatment Lp(a) concentrations achieved on average 97 13% of the baseline before the next session within one week. The daily measured mean percentage rates of deviation from the initial level differed from one another up to the sixth day (p < 0.006, Bonferroni adjusted). The values of the sixth and seventh days were not statistically different (p ¼ 1.00, Bonferroni adjusted).
212
S. Tselmin et al. / Atherosclerosis Supplements 30 (2017) 209e216
Fig. 2. Graphic model of the Lp(a) serum concentrations recovery after a single LA session. The recovery rate of Lp(a) serum levels within one week after a single apheresis session with regard to initial value can be predicted from the following quadratic equation: yð%Þ ¼ 22:49 þ 18:64x 1:14x2 ; where the independent variable (x) is the day after apheresis. The quadratic regression equation showed p < 0.001. Coefficient of determination R2 ¼ 0.874 demonstrated the good model fit for a prediction of percentage deviation from the pretreatment Lp(a) concentration dependent on the day after the treatment.
Fig. 3. Mean Lp(a) value between two consecutive lipoprotein apheresis sessions. The interval mean Lp(a) value between two consecutive lipoprotein apheresis can be predicted from the following equation: yðnmol=lÞ ¼ 5:1 þ 0:941 x; ðR2 ¼ 0:907; p < 0:001Þ; where x is the absolute acute Lp(a) reduction after a single LA session.
S. Tselmin et al. / Atherosclerosis Supplements 30 (2017) 209e216
0.941e1 ¼ 0.059 * 120 ¼ 7.08 and added it to 5.1 from the above mentioned equation. That resulted in the constant value of 12.18, giving the empirical formula: yðnmol=lÞ ¼ x 12; where x is the absolute acute Lp(a) reduction after a single LA session. The testing of this equation in all 20 patients demonstrated very small deviation from the real mean values as well the values predicted from the initial linear formula. We have also tested the formula: y ¼ (x þ z)/2, where x is the Lp(a) level to the end of apheresis session and z is the Lp(a) value after 7 days (before the next session). However, the values predicted from this equation were significantly lower compared to the real Lp(a) levels (Fig. 4). In both patients who had a severe progress of atherosclerotic complications and received extracorporeal therapy twice a week, the Lp(a) level exceeded the initial value by 7% and 44% respectively already on the fourth day after a single treatment. 3.3. Course of LDL-C concentrations between two LA sessions The initial mean LDL-C serum level was 2.2 0.72 mmol/l and was decreased by 70 9% (to 0.61 0.19 mmol/l) following a single apheresis session. The LDL-C concentration achieved 98 15% of the initial value on day 7 after apheresis (before the next session). The mean LDL-C values or the mean percentage rates of
213
deviation from the initial level assessed on the fourth and fifth as well as on the sixth and seventh days did not differ significantly from each other (p ¼ 0.09 and 1.00 respectively, Bonferroni adjusted). The recovery rate of LDL-C serum levels within one week following the initial apheresis session with regard to initial value can be predicted from the following quadratic equation: yð%Þ ¼ 27:16 þ 21:17 x 1:58 x2 ; where the independent variable (x) is the day after apheresis (p < 0.001, R2 ¼ 0.771). 3.4. Comparison of LDL-C and Lp(a) kinetics Under a single apheresis session Lp(a) serum concentrations decreased farther than those of LDL-C (76 7% vs. 70 9%, p ¼ 0.0036). The levels of both lipoproteins recovered virtually in parallel up to the fifth day and then their values converged to the baseline (Fig. 5). Testing of interaction between the two rebound curves using General Linear Model with repeated measures revealed no significant difference with respect to the factors time and both lipoprotein values (Hotelling’s Trace: p ¼ 0.225). The Lp(a) regression curve lay within the 95% e confidential interval of the LDL-C regression curve so that these increases were statistically similar. 4. Discussion We studied the course of serum levels of the atherogenic lipoproteins between two consecutive apheresis sessions
Fig. 4. The optimized equation compared to exact regression and mean values. The testing of the optimized equation ðyðnmol=lÞ ¼ x 12Þ in all 20 patients demonstrated very small deviation from the real mean values as well the values predicted from the regression equation: yðnmol=lÞ ¼ 5:1 þ 0:941 x; where x is the absolute acute Lp(a) reduction after a single LA session. We have also tested the formula: y ¼ ðx þ zÞ=2; where x is the Lp(a) serum level to the end of apheresis session and z is the level in 7 days before the next session. The values predicted from this equation were significantly lower in comparison to the real Lp(a) levels.
214
S. Tselmin et al. / Atherosclerosis Supplements 30 (2017) 209e216
Fig. 5. Comparison of the Lp(a) and LDL-C concentrations during time course. The levels of both lipoproteins recovered in parallel up to day 5 before converging towards their maximum. The confidential interval for LDL-C is slightly wider than that of Lp(a). The testing of interaction between the two rebound curves using General Linear Model with repeated measures revealed no significant difference with respect to the factors time and both lipoprotein values (Hotelling’s Trace: p ¼ 0.225).
focusing on the rebound curve of Lp(a). Most patients had isolated Lp(a)-hyperlipoproteinemia with LDL-C lower than 2.6 mmol/l (primarily or due to lipid-lowering drugs and/or regular extracorporeal therapy). Nevertheless LDLC changes were also analyzed to compare them with that of Lp(a). The serum levels of TG, TC and HDL-C were before apheresis in the normal range and their reductions were seen as expected. A single LA treatment caused a sharp drop in Lp(a) serum levels (between 58 and 88%) confirming previous findings [22]. The rebound rate could be best predicted by the quadratic equation: y (%) ¼ 22.49 þ 18.64 * x 1.14 * x2, demonstrating the gradual increase up to 97% of the initial value before the next apheresis session within 7 days. The coefficient of determination R2 ¼ 0.874 demonstrated a good model fit for a prediction of percentage deviation from the pretreatment Lp(a) concentration dependent from the day after the treatment (x). The time course of LDL-C levels was similar to that of Lp(a). This may be of importance for the design of Lp(a) targeting apheresis end point studies, because most systems effectively remove both atherogenic lipoproteins. The first analysis of the serum levels under extracorporeal methods was conducted in 1977 in 3 hyperbetalipoproteinemic patients, who received daily plasmapheresis for 5e9 days [23]. Interestingly, despite an entirely different study design the cholesterol concentrations in a non-homozygous patient returned to the pretreatment values within 7 days after cessation of treatment course which is in accordance with our data. On the other hand, only cholesterol, but not Lp(a) levels were analyzed and in contrast to lipoprotein apheresis, plasma exchange resulted in contraction of plasma volume that had to be considered in the analysis. Our findings also confirmed the results of a Japanese group on Lp(a) kinetics performed in 13 patients who were
treated with dextran sulfate columns [24]. They demonstrated similar rebound curves with an increase nearly to the baseline within 7 days after apheresis. They showed a significant slowdown of LDL-C recovery and at the same time an increase of Lp(a) levels in patients undergoing pravastatin treatment. The curves for both lipoproteins in our study did not differ, while the measured daily LDL-C values were elevated despite the fact that all patients received the maximum tolerated lipid lowering therapy. It can be partly attributed to lower pretreatments LDL-C values in our patients (2.3 mmol/l vs. 3.7 mmol/l in patients of the aforementioned study) resulting in only slight fluctuations within the normal range. In contrast to the Japanese trial which was conducted in 1993, we would not pause statins in persons with high cardiovascular risk for testing effects of a lipid-altering medication on rebound rates. Patients from the most recent study on this topic also took statins and underwent a biweekly LA [25]. Acute reduction rates of Lp(a) and LDL-C were similar to our data and also the rebound curves of both lipoproteins were not significantly different, but the rebound to baseline took longer than that in our patients. Factors influencing the rebound of lipoproteins after LA are of importance. Basing on the results of the Kroon’s study [25] and our findings it is prudent to note the following evidence: the longer the time between two sessions is, the longer the recovery time for the serum concentrations of removed lipoproteins will be. It is important to mention that some research groups demonstrated no increase in the cholesterol synthesis after LA. In rats it was shown that the elevated cholesterol production was associated with albumin removal, caused by plasmapheresis [26]. Since a previous analysis from our group demonstrated a maximum 15% protein and albumin reduction, this might lead to a limited induction of
S. Tselmin et al. / Atherosclerosis Supplements 30 (2017) 209e216
cholesterol synthesis [27]. On the other hand Pfohl reported on transient stimulating of cholesterol biosynthesis after LDL-apheresis despite concomitant statin-therapy and the degree of stimulation was inversely related to the level to which the LDL-C was reduced [28]. The frequency of LA therapy might play a relevant role. Our analysis included two patients receiving apheresis twice weekly. In both patients we noted a rapid rebound of Lp(a) over baseline during the 4 days following apheresis, i.e. to the next treatment. Most likely this finding relates to the much higher Lp(a) levels these subjects had without LA, but it is also related to the high risk both subjects derived from their very high Lp(a) levels. Without doubt, it will be of interest to analyze more individuals with a necessity for a higher treatment frequency. Of note, our model to predict the Lp(a) rebound considered only the day following LA, not including the values before and at the end of a single session or the reduction rate that makes the calculation easier than in the papers mentioned above. In the same time the “R” parameter representing the impact of the factor time on lipids concentration was high (R2 ¼ 0.874, therefore R ¼ 0.935) for our formula, comparable with Apstein’s equation (0.900), though it was modeled for plasmapheresis and not for Lp(a) [23] and exceeded that from Kroon’s formula (0.74) derived from an analysis having a similar design to our study [25]. The linear correlation between the acute Lp(a) reduction and interval mean Lp(a) level revealed by the regression analysis enabled us to develop a very simply and reliable equation to predict this value without integration of the area under the rebound curve. This empirical formula is much more practicable for daily clinical routine, nevertheless enabling optimized and individualized protocols for this costly extracorporeal therapy. As a major drawback, this study had a monocentric design and therefore a limited number of participants. On the other hand, this enabled a very homogeneous preanalytic processing of samples. 5. Conclusions 1. A higher Lp(a) level before LA apheresis leads to a greater absolute decrease at the end of the treatment; 2. the rebound of Lp(a) after LA apheresis is best described by a square function: y (%) ¼ 27.16 þ 21.17 * x e 1.58 * x2, where the independent variable (x) is the day after apheresis; 3. the interval mean Lp(a) value between two weekly apheresis sessions can be predicted from an easy and reliable empiric formula: y (nmol/l) ¼ x 12, where x is the absolute acute Lp(a) reduction after a single LA session; 4. the rebound courses of Lp(a) and LDL-C following LA are similar; 5. patients with high cardiovascular risk and higher LA intensity demonstrate a rapid rebound of serum Lp(a)
215
concentrations; the causality of this association should be investigated in further studies and could contribute to carrying out the individual strategies of the costly LA therapy. Conflict of interest S. Tselmin: no financial relationships or conflicts of interest regarding the content herein. G. Mu¨ller declared no conflict of interest. U. Schatz travel expenses by Sanofi-Aventis, Amgen, honoraria from Amgen, Sanofi-Aventis, Abbot, Fresenius. U. Julius travel expenses by Aegerion, Diamed, Fresenius Medical Care, Kaneka; honoraria from Aegerion, Amgen, Chiesi, Sanofi, Kaneka, Diamed, Fresenius Medical Care. SR. Bornstein declared no conflict of interest. B. Hohenstein received honoraria, travel support and project funding from Miltenyi Biotec, B Braun Avitum, Fresenius Medical Care and Kaneka. Author contributions S. Tselmin: data gathering and analyzing, writing the draft of publication. G. Mu¨ller: statistical data processing. U. Schatz: data gathering, data interpretation, draft editing. U. Julius: data interpretation and editing. SR. Bornstein, B. Hohenstein: idea of the study, data interpretation and editing of the manuscript. Acknowledgements The authors are grateful for the excellent work of the staff of the Extracorporeal Treatment and Apheresis Center at the University Hospital Carl Gustav Carus Dresden. The skilled help of Kerstin Haaser is gratefully acknowledged. This study was supported by a research grant from B. Braun Avitum to BH. References [1] Julius U, Tselmin S, Fischer S, Passauer J, Bornstein SR. The Dresden Apheresis Center e experience with LDL apheresis and immunoadsorption. Atheroscler Suppl 2009;10(5):12e6. [2] Julius U, Frind A, Tselmin S, Kopprasch S, Poberschin I, Siegert G. Comparison of different LDL apheresis methods. Expert Rev Cardiovasc Ther 2008;6(5):629e39. [3] Waldmann E, Parhofer KG. Lipoprotein apheresis to treat elevated lipoprotein(a). J Lipid Res 2016 Oct;57(10):1751e7. [4] Nordestgaard BG, Chapman MJ, Ray K. Lipoprotein(a) as a cardiovascular risk factor: current status. Eur Heart J 2010;31(23): 2844e53. [5] Tselmin S, Julius U, Mueller G, Fischer S, Bornstein SR. Cardiovascular events in patients with increased lipoprotein (a) e retrospective data analysis in an outpatient department of lipid disorders. Atheroscler Suppl 2009;10(5):79e84.
216
S. Tselmin et al. / Atherosclerosis Supplements 30 (2017) 209e216
[6] Kamstrup PR, Benn M, Tybjærg-Hansen A, Nordestgaard BG. Extreme lipoprotein(a) levels and risk of myocardial infarction in the general population: the Copenhagen City Heart Study. Circulation 2008;117(2):176e84. [7] Danesh J, Collins R, Peto R. Lipoprotein(a) and coronary heart disease. Meta-analysis of prospective studies. Circulation 2000; 102(10):1082e5. [8] Stra¨ter R, Becker S, von Eckardstein A, Heinecke A, Gutsche S, Junker R, et al. Prospective assessment of risk factors for recurrent stroke during childhood - a 5-year follow-up study. Lancet 2002; 360(9345):1540e5. [9] Dieplinger B, Lingenhel A, Baumgartner N, Poelz W, Dieplinger H, Haltmayer M, et al. Increased serum lipoprotein(a) concentrations and low molecular weight phenotypes of apolipoprotein(a) are associated with symptomatic peripheral arterial disease. Clin Chem 2007;53(7):1298e305. [10] Kamstrup PR, Tybjærg-Hansen A, Nordestgaard BG. Elevated lipoprotein(a) and risk of aortic valve stenosis in the general population. J Am Coll Cardiol 2014;63(5):470e7. [11] Bundesministerium fu¨r Gesundheit. Bekanntmachung eines Bes¨ nderung chlusses des Gemeinsamen Bundesausschusses u¨ber eine A der Richtlinie Methoden vertragsa¨rztliche Versorgung: Apherese bei isolierter Lp(a)-Erho¨hung. BAnz, 2008;138:3321. [12] Safarova MS, Ezhov MV, Afanasieva OI, Matchin YG, Atanesyan RV, Adamova IY, et al. Effect of specific lipoprotein(a) apheresis on coronary atherosclerosis regression assessed by quantitative coronary angiography. Atheroscler Suppl. 2013 Jan;14(1): 93e9. [13] Jaeger BR, Richter Y, Nagel D, Heigl F, Vogt A, Roeseler E, et al. Longitudinal cohort study on the effectiveness of lipid apheresis treatment to reduce high lipoprotein(a) levels and prevent major adverse coronary events. Nat Clin Pract Cardiovasc Med 2009;6(3): 229e39. [14] Roeseler E, Julius U, Heigl F, Spitthoever R, Heutling D, Breitenberger P, et al. Pro(a)LiFe-Study Group. Lipoprotein apheresis for lipoprotein(a)-associated cardiovascular disease: prospective 5 years of follow-up and apo(a) characterization. Arterioscler Thromb Vasc Biol. 2016 Sep;36(9):2019e27. [15] von Dryander M, Fischer S, Passauer J, Mu¨ller G, Bornstein SR, Julius U. Differences in the atherogenic risk of patients treated by lipoprotein apheresis according to their lipid pattern. Atheroscler Suppl 2013;14:39e44. [16] Kassner U, Vogt A, Rosada A, Barz F, Giannakidou-Jodan E, Berthold HK, et al. Designing a study to evaluate the effect of apheresis in patients with elevated lipoprotein(a). Atheroscler Suppl 2009;10(5):85e8.
[17] Armstrong VW, Schleef J, Thiery J, Muche R, Schuff-Werner P, Eisenhauer T, et al. Effect of HELP-LDL-apheresis on serum concentrations of human lipoprotein(a): kinetic analysis of the posttreatment return to baseline levels. Eur J Clin Invest 1989;19(3): 235e40. [18] Groß E, Hohenstein B, Julius U. Effects of Lipoprotein apheresis on the Lipoprotein(a) levels in the long run. Atheroscler Suppl 2015 May;18:226e32. [19] Kronenberg F. Causes and consequences of lipoprotein(a) abnormalities in kidney disease. Clin Exp Nephrol 2014;18(2):234e7. [20] Julius U, Fritsch H, Fritsch W, Rehak E, Fu¨cker K, Leonhardt W, et al. Impact of hormone replacement therapy on postprandial lipoproteins and lipoprotein(a) in normolipidemic postmenopausal women. Clin Investig 1994;72(7):502e7. [21] Perk J, De Backer G, Gohlke H, Graham I, Reiner Z, Verschuren WM, et al. European Guidelines on cardiovascular disease prevention in clinical practice (version 2012): The Fifth Joint Task Force of the European Society of Cardiology and Other Societies on Cardiovascular Disease Prevention in Clinical Practice (constituted by representatives of nine societies and by invited experts). Atherosclerosis 2012 Jul;223(1):1e68. [22] Leebmann J, Roeseler E, Julius U, Heigl F, Spitthoever R, Heutling Dfor the Pro(a)Life-Study Group. Lipoprotein apheresis in patients with maximally tolerated lipid lowering therapy, Lp(a)hyperlipoproteinemia and progressive cardiovascular disease. Circulation 2013;128(24):2567e76. [23] Apstein CS, Zilversmit DB, Lees RS, George PK. Effect of intensive plasmapheresis on the plasma cholesterol concentration with familial hypercholesterolemia. Atherosclerosis 1978 Oct;31(2):105e15. [24] Koizumi J, Koizumi I, Uno Y, Inazu A, Kajinami K, Haraki T, et al. Reduction of lipoprotein(a) by LDL-apheresis using a dextran sulfate cellulose column in patients with familial hypercholesterolemia. Atherosclerosis 1993 Apr;100(1):65e74. [25] Kroon AA, Van’t Hof MA, Demacker PNM, Stalenhoef AFH. The rebound of lipoproteins after LDL-apheresis. Kinetics and estimation of mean lipoprotein levels. Atherosclerosis 2000;152:519e26. [26] Takeuchi N, Yamamura Y. The effect of plasmapheresis on cholesterol synthesis in the rat: relationship to protein synthesis. J Lab Clin Med 1972 May;79(5):801e13. [27] Julius U, Siegert G, Kostka H, Schatz U, Hohenstein B. Effect of different lipoprotein apheresis methods on serum protein levels. Atheroscler Suppl 2015 May;18:95e102. [28] Pfohl M, Naoumova RP, Klass C, Knisel W, Jakober B, Risler T, et al. Acute and chronic effects on cholesterol biosynthesis of LDLapheresis with or without concomitant HMG-CoA reductase inhibitor therapy. J Lipid Res 1994 Nov;35(11):1946e55.