5Friedewald Misclassifies Risk Category When Triglycerides > 177 mg/dL Estimating LDL-C is UnreliableWhen triglycerides (TG) > 400 everyone knows the LDL may be 25% falsely low and you have to order a directly measured LDL. You may not realize that even at relatively modest elevations of TGs when TG > 200 the estimated, or Friedewald LDL is already 10% falsely low or more, at TG > 300 it may be 15%-20% falsely low. What’s more when TG < 100 Friedewald breaks down again and you get a falsely low LDL.In this study by Marnieri of 225 patients whose LDL was measured with the preparative ultracentrifuge, it was found that the Friedewald equation would have put persons in the wrong LDL risk category as much as half the time – this occurred for patients with TG > 177 mg/dl (which is > 2.0 mmol/l).Of course you’re supposed to fast for Friedewald else TG’s go up and the estimated LDL goes down. In my mind we systematically underestimate LDL with Friedewald mainly because so many patients do not really fast.LDL Risk CategoryMarniemi J et al. Clin Biochem 1995 June; 28:
6When LDL-C Target is < 100 Friedewald Should Not Be Utilized Directly Measured LDL CholesterolWhen LDL-C Target is < 100 Friedewald Should Not Be UtilizedThe third and perhaps most important reason to move immediately to direct LDL measurement in all patients is that Friedewald never intended for us to use his equation on persons’ with low LDL values, I.e. LDL-C < 100 mg/dL. The Friedewald equation is 15% - 19% falsely low in persons whose directly measured LDL is under 100 mg/dL. We are in an era where all CAD patients should have LDL below 100 and it must be measured directly to be accurate.The fact that VAP provides a direct LDL-C, reliable at values below 100, reliable in persons with high or low TGs, and reliable in non-fasting persons - is sufficient reason alone to replace the old lipid panel with the VAP. It’s just a Better Cholesterol Test. When you switch you will notice that your patient’s LDLs suddenly look higher because the Friedewald underestimates LDL.*Scharnagl H et al. Clin Chem Lab Med May;39(5):
7Evaluation of Four Homogeneous Direct LDL-C Methods Study: LDL-C methods from Genzyme, Reference Diagnostics (RD), Roche, and Sigma were evaluated for precision, accuracy, and specificity for LDL in the presence of abnormal lipoproteins.Results: Precision was < 2% CV for all methods; and correlation to the CDC reference method was r = ; total error was Genzyme = 12.6%, RD = 16.5%, Sigma = 38.3%, Roche = 41.6%.Conclusion: The methods show nonspecificity toward abnormal lipoproteins, thus compromising accuracy. These direct methods are no better than the Friedewald LDL-C.From: Miller, WG et.al. Clin Chem 48: (2002)
8Review of LDL-C Methods 1. Ultracentrifugation:Reference method, accuracy basedTime-consuming, tedious, technically difficult, costly2. Electrophoresis:Simultaneous separation of major lipoprotein fractions, can quantitate, visualize some unusual bandsTechnically difficult to do; unless automated, it can be tedious and time consuming, can be costly+-+-
9Review of LDL-C Methods (Cont.) 3. Third Generation Direct Methods:+No pretreatment, full automation, improved analytical precision, fasting specimen not required, save on labor costCan have lack of LDL specificity; inaccuracy because of analytical interferences from TG, bilirubin, IDL-C, Lp-X, VLDL-C, Lp(a), apo E-rich HDL, and/or hemoglobin; can be costlyFrom: Nauck M et. al. Clin Chem 48: (2002)-
11Homogenous HDL-c Methods There are 4 types of homogeneous HDL-C methods: immunologic, PEG, synthetic polymer, enzymatic.All can be automated & directly measure HDL.Most have excellent precision (< 3% CV).Many have CRMLN certification, suggesting that accuracy is possible with proper instrument, reagents, and calibrator.Specificity and interferences: most are robust and can tolerate TG < 900 mg/dL, but biases can occur with atypical lipoprotein patterns.From: Warnick et. al. Clin Chem 47: (2001)
12Review of HDL-c Methods 1. Ultracentrifugation:Comparison method for accuracyTime-consuming, tedious, technically difficult, costly, can have interference from Lp(a) and others2. Electrophoresis:Simultaneous separation of major lipoprotein fractions, can quantitate, visualize some atypical bandsTechnically difficult to do; unless automated, it can be tedious and time consuming, can be costly+-+-
13Review of HDL-C Methods (Cont.) 3. Third Generation Direct Methods:+No pretreatment, full automation, improved analytical precision (<3% CV), fasting specimen not required, save on labor cost (20% savings)Can have lack of HDL specificity; inaccuracy because of analytical interferences from bilirubin, hemoglobin, high TG (chylomicrons and VLDL-c), Lp(a), apo E-rich HDL, and/or HDL variants (Apo A-IMilano)From: Nauck M et. al. Clin Chem 48: (2002)-
14New Approaches to Quantitative Lipoprotein Electrophoresis Completely automated; better precisionBetter gel/buffer system (high-resolution)Precision control of temperature during electrophoresisDensitometer designed for higher sensitivitySimultaneously measures VLDL-c, LDL-c, HDL-c & Lp(a)-cLDL-c and HDL-c compares well with CDC reference methodsLp(a) compares well with Lp(a) candidate reference method
15Within-Run with Helena Electrophoretic Method Pool VLDL-C LDL-C HDL-Cmg/dL mg/dL mg/dLB 8.1 ± ± ± 1.5(5.3% CV) (1.3% CV) (3.6% CV)From: Naito HK, et al. Handbook of Lipoprotein Methods. AACC Press, Washington D.C., Pgs (1997)
16Comparison of Methods: Helena Electrophoresis vs. CDC Reference Method Fractions Regression rVLDL-C y= xLDL-C y= xHDL-C y= xFrom: Naito HK, et al. Handbook of Lipoprotein Methods. AACC Press, Washington D.C., Pgs (1997)
17Comparison of Methods: Helena vs. Ultracentrifugation Fractions Regression rVLDL-C y= 0.929xLDL-C y= 1.006xHDL-C y= 1.158xLp(a) y= 0.23xFrom: Nauck M, et al. Clin Chem 1995; 41:
18LDL-c and HDL-c Summary LDL-C calculation method is not dependable when compared to some of the current direct LDL-C methods.Ultracentrifugation methods for LDL and HDL still remains the gold standard for accuracy.The 3rd-generation electrophoresis methods for lipoprotein-C quantitation have several attractive features.The 3rd- generation direct or homogenous LDL-C and HDL-C methods have better precision but may suffer from lack of LDL or HDL specificity; thus, accuracy may be compromised.
19LDL-c and HDL-c Summary (cont) The selection of your instrument-reagent-calibrator system for LDL-C or HDL-C determinations should be based on the approved list of the CDC CRMLN certificate of traceability.Periodic verification of accuracy of LDL-C and HDL-C should be done, particularly if the reagent and/or calibrator lot number changes.More robust LDL-C and HDL-C methods should be employed that are not affected by interfering substances (i.e., abnormal lipoproteins, hypertrigly- ceridemia)
20Lipoprotein(a) Chemistry Structurally resembles LDLHas a second large polypeptide, Apo(a)Is polymorphic in size; Molecular weight kDaHas 10 types of kringle 4, which is the basis of the different isoform size variability
21Physiology of Lp(a)Is an acute phase protein. Apo(a) is made by the liver and is assembled with apo B-100 on the hepatocyte surface. Lp(a) catabolism is unclear.Compete with plasma plasminogen for binding sites, resulting in decreased synthesis of plasmin and inhibition of fibrinolysisIncreases cholesterol deposition in the arterial wallEnhances foam cell formationMakes O2-free radicals in monocytesPromotes SMC proliferationInduces monocyte-chemotactic activity in subendothelial space
22Physiology of Lp(a) (Cont.) Mechanism of CAD: Atherogenesis and ThrombogenesisEmerging Risk Factor for Vascular DiseaseMost prospective and retrospective studies suggest an independent association between Lp(a) and presence and extent of CAD, premature CAD, MI, restenosis after balloon angioplasty, and CVD.There is evidence for a benefit of lowering Lp(a)Response to Intervention TherapyDiet and exercise have no effect: (maybe with monounsaturated fats or caloric restriction with weight loss)Effect of statins are controversalNiacin and aspirin will lower
23Lp(a) and CHD: Meta-analysis of 27 Prospective Studies n The study (n = 5436) with a mean follow-up of 10 years showed that if an individual in the general population is in the upper third at baseline, you are at 70% increased risk for CHD compared to persons at the lower one-third.n The Lp(a) association to CHD risk is significant and is independent of the standard vascular risk factors.From: Danesh J . Circulation 2000; 102:
24Lp(a) Increases CHD Risk In Men With Other Risk Factors n PROCAM Prospective Study: 788 males were followed for 10 yrs; [Lp(a) measured on fresh blood]n The overall risk of a coronary event was 2.7 times higher if Lp(a) was > 20 mg/dL. The risk increased further if there were other risk factors, I.e., LDL-C > 160 mg/dL, HDL-C < 35 mg/dL, BP > 140/90 mmHgn CONCLUSION: Lp(a) is an important independent CHD risk factor that aggravates the coronary risk exerted by elevated LDL-C, low HDL-C, hypertension or the combined effects of multiple risk factors (TG, smoking, diabetes, angina pectoris, and family history of MI)From: Am Coll Cardiol 2001:37:
25Who Should Have an Lp(a) Test Done? n Patients with a normal lipid profile, but have documentation of definite CHD (MI, angina, CABAG, angioplasty, stent implants)n Patients with parents or 1st-degree relatives who died of premature CHDn Patients with known elevation of Lp(a) or parents with elevation of Lp(a)n High-risk African American malesn Postmenopausal womenn Men with traditional and/or global CHD risk factors; diabetics and patients with renal disease
26It All Began with Some Observations by a Clinical Pathologist and a Cardiologist… Let’s prove it!
27Clinical StudiesRetrospective study of 1124 subjects with lipoprotein electrophoresis (including Lp (a)) performed in a 26 month period. Lipoprotein electrophoresis performed only when abnormal lipids where found apriori and/or in cardiac patients with high hsCRP/histamine. These patients are followed to date.Prospective study of CHD compared with healthy subjects matched for age and sex: Distribution according to race and Lp(a).100 patients, (72 w, 28 aa), and 50 healthy subjects.Prospective study of 51 patients with poorly controlled T2DM before and after treatment with infusion with a external insulin pump.
28Major Diagnosis2 major groups: Caucasians + (Hispanics) and African Americans were considered for statistics.T2DMCHDCVDOtherCaucasians82902010African Amer.88842414CHD: coronary heart disease (MI, angina, coronary insufficiency, angioplasty, bypass surgery, restenosis)CVD: cerebrovascular disease (ischemic stroke, TIA, brain infarction)T2DM: Type 2 diabetes mellitus ( nephrotic syndrome)Other: PAD, chronic renal failure
29Distribution of Lp(a) Cholesterol According to Race Caucasians34485184African Amer.4089210Hispanics491220Asians4264Pakistani+Indians3423Total (%)509 (45.28)194 (17.26)421 (37.45)
30Distribution of Lipids in the Cohort with Lp(a) > 4.1 Chol > 200TG > 150HDL < 40# (%)Caucasians130 (37.8)110 (32.0)86 (25)African Amer.29 (72.5)14 (35)16 (40)Hispanics15 (30.6)10 (20.4)14 (8.6)Asians12 (28.6)8 (19.1)6 (14.3)Pakistani+Indians30 (88.2)14 (41.2)2 (5.9)
31Distribution of Increased Lipoproteins in the Cohort with Lp (a) > 4.1 VLDL > 35LDL > 120HDL < 40BROAD BCaucasians #(%)48 (23.5)62 (30.4)40 (19.6)34 (16.6)African Amer. #(%)50 (23.8)71 (33.8)38 (18.1)10 (4.8)
32Distribution According to Race & Lp(a) (36w, 21aa) CaucasiansAfrican American# (%)0 – 425(69.4)12(57.1)4.1 – 104(11.1)4(19.1)> 10.17(19.5)5(23.8)
33Normal Lp(a) Values Caucasian African American Lp(a) Non-Diseased Population (Caucasian & African American)CaucasianAfrican AmericanAllMalesFemalesMean22.214.171.124.43.23.8SD126.96.36.199Median188.8.131.52.22.095th Percentile184.108.40.206.68.3% ≥10 mg/dL98.398.897.797.998.297.0Min0.20.49.2Max11.410.911.811.311.5N2541461081025646
34Lp(a) Stratified by Risk Score Diseased PopulationFramingham Risk AssessmentAll≤45 and 67≥8Mean220.127.116.110.914.9SD18.104.22.168.88.0Median22.214.171.124.010.65th Percentile126.96.36.199.25.495th Percentile22.49.814.818.026.3% ≥10 mg/dL51.622.559.465.2Min0.20.61.63.3Max24.414.623.228.8N530103105195127
35ROC of Lp(a) Cut-offs vs Clinical Sensitivity & Specificity at Framingham Risk Assessment Cut-off of < 40.22.03.95.98.19.912.51518.2202528406080100Area under the curve = ± 0.007Sensitivity (%)1-Specificity (%)
36Cardiovascular Disease is the Leading Cause of Death in the US CV Cancer Accident Pneumonia AIDS disease InfluenzaMillionsCP
37Ischemic Events in the U.S. 1,500,000 heart attacks: 500,000 deaths500,000 strokes: 150,000 deathsOne third of individuals who experience anischemic event will die as a result of that event.Many who have an event have no priorsymptoms.
38Prevention of Ischemic Cardiovascular Events is Key!! How do we determine who will have an ischemic event?Risk FactorsAge, smoking status, hypertension, diabetes, Cholesterol, HDL.The ATP-III: NCEPProvide risk factor screening guidelines and treatment guidelines based on risk factor analysis.
39Cholesterol Screening: NCEP Everyone over 20 years of ageEvery 5 yearsTotal CholesterolHDL cholesterolTriglycerideLDL cholesterolTreatment guidelines based on LDL cholesterol concentration
42Prevention of Ischemic Cardiovascular Events is Key!! One-third to one-half of ischemic events occur in individuals with LDL < 130 mg/dL and in current guidelines for primary prevention the target is < 130 mg/dL.
43Cardiovascular Risk Factors There is a need for additional risk factors.Additional risk factors would improve accuracy of decisions regarding preventative therapiesRader DJ. NEJM. 2000;343:
45Emerging Risk Factors Proposed Panel HomocysteineLipoprotein(a) [Lp(a)]High-sensitivity CRPFibrinogenSmall dense LDLCurrently no guidelines for measurement; may guide intensity of risk reduction therapy in selected patients.
46Emerging Risk FactorsThe NCEP ATP-III panel identified these novel markers and indicated that clinicians may utilize them in selected persons to guide intensity of risk reduction therapy and modulate clinical judgment when making therapeutic decisions.They do not however, identify in which group of patients these markers are best used or how to respond to elevated values.JAMA 2001:285;
47Use of Novel Risk Markers Mayo Test Volumes: 2003Homocysteine 58,000hsCRP 26,000Lp(a) 22,000LDL subfractions 3,600Fibrinogen ?????
48Mayo Recommendations for use of Extended Risk Marker Panel Acute coronary syndromeIncreasing risk>20%10 year riskCAD andCAD risk equivalentsExtended MarkerPanel10% - 20% 10 year risk6%-9% 10 year riskLow risk population
49Mayo Recommendations for Use of Extended Risk Marker Panel When to MeasureUse to enhance clinical decision making in persons at intermediate risk for developing an ischemic event as assessed by the Framingham 10 year risk score: 10-20% risk.Provide with RequestAge, gender, smoking status, blood pressure (treated or untreated)
51Mayo Recommendations Clinical Response to Elevated Markers Interpretations are provided with every report of the novel risk marker panel.Interpretations are made by doctoral level staff in Laboratory Medicine or in the Mayo Cardiovascular Health Clinic.Interpretations include a description of abnormal values, as well as suggestions for appropriate treatment, given the noted abnormalities.A separate document describing up to date information on the background, interpretation and recommended therapy for abnormalities of each of the risk markers is included with the report.
52Novel Risk Markers Relation to Angiographic CAD and Events 504 consecutive patients undergoingcoronary angiographyVessel Disease3VNone2VMild1VMean Age 60 ± 11 years, 62% Male46% patients have none or mild coronary occlusion54% patients have significant occlusion (>50% stenosis)Median Follow-up: 4 years
53Patient Demographics Age (yr) 60±11 Male (%) 62 Hypertension (%) 46 BMI (kg/m2)29±6Prior infarction (%)15Heart failure (%)12Significant CAD (%)54ACS (%)34
54Study Population Indication for coronary angiography Acute coronary syndrome (34%)Positive stress test (25%)Exertional dyspnea (27%)Other (14%)
55Aims To analyze for associations between novel risk factors including Lp-PLA2 and:Other CAD risk Factors(both traditional and emerging)Acute coronary syndromesAngiographic CADClinical outcomes
56Lipoprotein Associated Phospholipase A2 (Lp-PLA2) AKA: Platelet-activating factor acetylhydrolase50kDa, Ca-insensitive lipaseProduced predominantly by macrophages and contributes to foam cell formation> 80% bound to LDLNot responsive to IL-1, IL-6, TNF-alpha.Hydrolyzes oxidized phospholipids
57Multivariate Logistic Regression Analysis After adjusting for age, gender, smoking history, hypertension, cholesterol, HDL cholesterol, and triglyceride, Lp-PLA2 was no longer was no longer independently predictive of angiographic CAD.CRP was not predictive of CAD in either the univariate or multivariate model.Data Presented at American College of Cardiology Annual Meeting, March 2003.
58HR For Angiographic CAD: Multivariate Variable HR 95% CI P-ValueAge <0.0001Male Gender – 6.60 <0.0001HypertensionSmokingTotal cholesterolHDL cholesterolLog triglycerideLog CRPFibrinogenLp-PLALp(a) protein (per 20 mg/dL)Lp(a) cholesterol (per 5 mg/dL)
59Lp(a) Cholesterol Electrophoresis Helena Laboratories LDLVLDLHDLLp(a)UCNon-UCSample 1Sample 2Sample 3Sample 4Sample 5UC Lp(a) = 6Non-UC Lp(a) = 8Lp(a) mass = 125UC Lp(a) = 8Non-UC Lp(a) = 19Lp(a) mass = 131UC Lp(a) = < 2Non-UC Lp(a) = < 2Lp(a) mass = <7UC Lp(a) = 19Non-UC Lp(a) = 21Lp(a) mass = 72UC Lp(a) = <2Non-UC Lp(a) = <2Lp(a) mass = 11Lp(a) cholesterolcan be measuredin whole serumVerified in samplesfrom 470 subjectsBaudhuin, et.al., Clinical Biochemistry: 2004
66Overview and Conclusions Prevention of Cardiovascular events is Key!!!Know Your Cholesterol, Know Your RiskLipid profile in everyone > 20 years of ageUse ATP III guidelinesDetermine Framingham Risk ScoreEducate OthersNovel/Emerging Risk MarkersPrimarily in patients at intermediate riskGuide intensity and type of therapyMuch more work needs to be done
67ConclusionsLp-PLA2 is emerging as an independent risk marker for cardiovascular events: orally active specific inhibitors make it a potential therapeutic target.Lp(a) cholesterol is a strong marker for angiographic coronary disease as well as cardiovascular events.Differences observed between Lp(a) cholesterol and Immunologic Lp(a) mass assays need to be further investigated, but may be due to dependance of the mass assays on apo(a) isoform sizeEfforts to standardize Lp(a) methods need to continue.
69Panel Comparison 266 consecutive subjects 45% had desirable TC, TG by NCEPOf those, 44% had Lp(a) above 55th percentile28% had Lp(a) above 75th percentile28% had desirable TC, TG, LDL-c, HDL-cOf those, 49% had Lp(a) above 55th percentile
70Lp(a) CorrelatesNone of the following are correlates or predictors of Lp(a) values:Total cholesterolLDL-cHDL-cTGApo A1Apo B100
71Lp(a)/Lp(a)-c and Platelet Function Study of 100 subjects with existing Lp(a) and Lp(a)-c measurementsLp(a) ranged from <5-231 mg/dLLp(a)-c ranged from mg/dLMeasured platelet aggregation to collagen/ADP and collagen/EPI by PFAMedication history: 34% on ASA or NSAIDS
72ResultsClosure times with either EPI/collagen or ADP/collagen were not decreased as Lp(a) concentrations increasedNo difference between Lp(a) and Lp(a)-c