(Hi)story of Lp(a) into the CV arena Florian Kronenberg Medical - PDF document

(Hi)story of Lp(a) into the CV arena Florian Kronenberg Medical University of Innsbruck, Division of Genetic Epidemiology Disclosures ■ Speaker honoraria: Kaneka, Fresenius, Miltenyi Biotec ■ Advisory activities: Kaneka, Amgen ■ No support for research activities from companies 1

Overview or how to set the stage 1. The long bumpy ride into the CV arena 2. The black hole of LPA 3. Therapy: the future has already started I apologize that not every milestone can be mentioned First indications that Lp(a) might make it into the CV arena 2

First time mentioning of a 30 and 50 mg/dL threshold ? The discovery of the apo(a) size polymorphism F B B S1 S1 S1/S2 S1 S2 S3 S4 Utermann et al.: J. Clin. Invest. 80: 458-65, 1987 3

Apolipoprotein(a) 10 to >50 KIV copies (Mr ≈ 300‐800 kDa) 5‘ 3‘ KIV 1 KIV 2 KIV 3 KIV 4 KIV 5 KIV 6 KIV 7 KIV 8 KIV 9 KIV 10 KV P n = 1 to >40 KIV 2 copies 20 KIV 2 repeats (=29 KIV repeats) Fibre‐FISH Association of apo(a) isoforms and Lp(a) concentrations Molecular weight . low (LMW) = small high (HMW) = large 60 53 48 50 47 Median Lp(a) mg/dL 40 30 20 13 10 9 10 7 7 4 0.6 0 11-16 17-19 20-22 23-25 26-28 29-31 32-34 35-37 >37 Null Number of K-IV repeats J. Int. Med. 273: 6-30, 2013 (updated) 4

The hard start in prospective studies Editorial Lp(a) and risk for myocardial infarction Results from the Copenhagen City Heart Study Kamstrup et al.: JAMA 301:2331-9, 2009 5

Oxidized phospolipids arrive in the arena Results from the Bruneck Study Tsimikas et al.: J.Am.Coll.Cardiol. 47:2219-28, 2006 Oxidized phospolipids arrive in the arena Results from the Bruneck Study HR = 2.4 HR = 2.8 (1.3-4.3) (1.6-5.0) Kiechl et al.: Arterioscler.Thromb.Vasc.Biol. 27:1788-95, 2007 6

Mendelian Randomization studies: the strong support for causality Lp(a) and CHD: Mendelian randomization Small apo(a) determined at the time of conception isoforms strong association association? (explains  causality? about 50%) association High Lp(a) CHD "reverse causation" Do carriers of small apo(a) isoforms more often have CHD? 7

Apo(a) isoforms and risk for CHD % of controls / patients with small apo(a) isoforms 60 Controls CHD 50 Combined OR = 1.78 40 p<0.001 % 30 20 10 0 Tyrolean Welsh German Israeli Chinese Indian Populations Sandholzer et al.: Arterioscler Thromb 12: 1214-26, 1992 Lp(a) concentrations, apo(a) isoforms and CVD Small apo(a) isoforms 25-35% of the population strong association strong association (explains about 50%) association causal association High Lp(a) CVD Kronenberg & Utermann: J. Int. Med. 273: 6-30, 2013 8

SNPs as risk factor for CHD: revitalization for the field % Clarke et al.: NEJM 361: 2518-28, 2009 rs10455872 as risk factor for CHD ■ This SNP is associated with higher Lp(a) levels ■ Claimed that this SNP tags small apo(a) isoforms Compared to 2.08 for small apo(a) isoforms Clarke et al.: NEJM 361: 2518-28, 2009 9

Carrier-status of the 2 SNPs as marker for small isoforms? Results from the KORA Study in 5999 subjects rs10455872 rs10455872 or rs3798220 or rs3798220 Non‐Carrier Non‐Carrier Apo(a) phenotypes 97% 51% Do they get the wrong message? Large Small 76% 24% 3% 49% Carrier Carrier J. Int. Med. 276: 243-7, 2014 (updated) Effect of 49 independent Lp(a)-associated SNPs on CAD risk CAD data retrieved from CARDIoGRAM consortium For each copy of the minor (GWAS on CAD risk; n =60,000 cases + 120,000 controls) allele (MAF~1%): • Effect on Lp(a): ~65 mg/dL • CAD risk: OR= 1.73  Risk increase proportional to Lp(a)-effect  7 SNPs even genome-wide significant in LD with rs10455872 in LD with rs3798220 Mack et al.: J. Lipid Res. 58:1834-44, 2017 10

Overview or how to set the stage 1. The long bumpy ride into the CV arena 2. The black hole of LPA 3. Therapy: the future has already started Apo(a) K-IV repeat polymorphism Special characteristics: • 95% are heterozygous at DNA level • 50% express only 1 isoform in plasma • <1% express no isoform in plasma 11

Large variance in Lp(a) levels in various apo(a) isoform classes KORA F3+F4, n=6221 Small isoforms Large isoforms 11-18 19-22 23-28 29-34 >34 K-IV 33% with low 8% with high % concentrations concentrations <15 15-30 30-45 45-60 >60 <15 15-30 30-45 45-60 >60 <15 15-30 30-45 45-60 >60 <15 15-30 30-45 45-60 >60 <15 15-30 30-45 45-60 >60 Lipoprotein(a) mg/dL Other genetic variants that modulate Lp(a) levels? Structure of apolipoprotein(a) Molecular weight of apo(a) 300-800 kD ■ Hard to resolve with Next Generation Sequencing ■ Mutations in this repetitive region? 12

Variation in the LPA KIV-2 region is not easily accessible Coverage in the ExAC project browser (>60.000 exomes) KIV-2 repeat region No (reliable) variation information http://exac.broadinstitute.org/gene/ENSG00000198670 ■ Up to 70% of the protein are coded in the KIV-2 and are thus not accessible to common mutation screening ■ A robust KIV-2 screening method is required for mutation screening in this white spot Ultra‐deep “batch NGS sequencing” ■ All KIV-2 are amplified as a mixture using primer binding to all repeats (“batch sequencing”) ■ Developed a low level variant caller for the KIV-2 region: all reads are mapped to one KIV-2 repeat ■ Variants down to 0.5-1% mutational level can be called Coassin et al.: Eur. Heart J. 38:1823-31, 2017 Coassin et al.: J. Lipid Res. 60:186-99, 2019 13

KIV2 region might be more variable than previously thought Screening in 123 samples from GCKD and SAPHIR G4925A located at an exonic splice donor site 135 120 Number of probands [n] 105 90 75 60 45 30 15 0 KIV‐2 exon 1 KIV‐2 exon 2 Coassin et al.: Eur. Heart J. 38:1823-31, 2017 Coassin et al.: J. Lipid Res. 60:186-99, 2019 G4925A frequency and effect on Lp(a) levels and CVD ■ Splice site variant ■ Competitive allele-specific TaqMan PCR („castPCR“) ■ Genotyping in population-based study (n=2892) ■ 22.1% Carrier frequency ■ Clusters in isoforms 19-25 ■ Reduces Lp(a) by 31 mg/dL ■ Explains 20.6% of Lp(a) variance in LMW carriers ■ Associated with lower CVD risk Coassin et al.: Eur. Heart J. 38:1823-31, 2017 14

A bit more complicated due to a mutation in a white spot Coassin et al.: Eur. Heart J. 38:1823-31, 2017 Overview or how to set the stage 1. The long bumpy ride into the CV arena 2. The black hole of LPA 3. Therapy: the future has already started 15

Lp(a)-lowering by IONIS antisense to apo(a) Sam Tsimikas Viney et al.: Lancet 388:2239-53, 2016 How much should Lp(a) be decreased? Brian Ference A change in Lp(a) levels of 101.5 mg/dL is required to have the same effect on CHD risk as a 38.67-mg/dL change in LDL-cholesterol Burgess et al.: JAMA Cardiol 3:619-27, 2018 16

SADE.ALI.19.03.0722 Summary of the estimates for required Lp(a)-lowering ■ HPS2-THRIVE 2018: 50-60 mg/dL can be deducted ■ Burgess et al. JAMA Cardiol 2018: 101.5 mg/dL (MR-approach) ■ Lamina & Kronenberg , JAMA Cardiol. 2019 : 65.7 mg/dL (MR-approach) ■ FOURIER trial 2019 : patients with higher Lp(a) levels showed a tendency to a stronger risk reduction ■ ODYSSEY trial 2019 : lowering of Lp(a) and LDLC contribute independently to MACE reduction - lowering of Lp(a) by 41 mg/dL lowers MACE by 22% 17

Take-home messages ■ Lp(a) is one of the most important genetically determined risk factors for CVD ■ There were many ups and downs during the ride into the CV arena ■ Genetic Mendelian randomization studies strongly support causality ■ These studies also provide estimates by how much Lp(a) should be lowered Genetic and first clinical trials provide reasonable support that an isolated lowering of Lp(a) might result in a clinical benefit 18