Thread by @Lpa_Doc

1/20 Saturday Morning Class #12 Lp(a) and clinical trials

Welcome back to class, and a healthy and fulfilling 2021. Despite the ongoing chaos of life and society and all it entails, if one looks at events historically over centuries or millennia, humans continue to evolve...

2/20 ..to make life fairer and just. I am optimistic that progress will continue to be made for all people, even though in the moment it may not seem like it to many.

3/20 Now back to science, medicine and the search for Lp(a) truth.

The highest evidence for a medical ‘truth’ is a randomized, double blind (patient and investigators don’t know the treatment one is assigned to), placebo controlled trial, with an adequate number of patients

4/20 Some trials don't need to be done- i.e. a trial of the safety of parachutes by jumping out of planes with or without a parachute to see who survives. There are other kinds of trials, but the ability to these to generate “medical truth” is weaker and will not be discussed.

5/20 The statistically metric used is the p-value (or confidence intervals, like COVID-19 vaccine trials). A p<0.001 showing a benefit of a drug over placebo can be interpreted as: if the study was repeated 1000 times, only one time would the study findings be due to chance.

6/20 Usually, a p<0.05 (1 out of 20 being due to chance) is used as the minimal cutoff for a significant results, and a metric investigators, journals, the FDA uses for approvals. getting 100% certainty is impossible -rare people survive falling out of airplanes with no parachute

7/20 Fun fact: In a prior long voyage, 1400/1900 sailors died of scurvy (vit C deficiency). James Lind gets credit for first trial: he randomized 12 sailors with scurvy to 6 groups; cider, sulfuric acid, vinegar, seawater, barley water and 2 oranges/1 lemon, but no placebo group.

8/20: They ran out of fruit by day 5 but the 2 sailors in the fruit group had already improved and one returned to work. The rest or evolution of trials is too complex to go into here, but in the 20th century is when trials were done properly.

9/20 Back to Lp(a): despite being discovered in 1963, randomized, double blind clinical trials in Lp(a) did not occur until the 2010s.

To do a proper trial in Lp(a), one needs to recruit subjects with elevated Lp(a), randomize to a therapy vs placebo, and see the results.

10/20 Despite much data on Lp(a) and apheresis, statins, niacin, PCSK9i and other therapies, I can only find 5 published, randomized double blind studies in subjects enrolled with elevated Lp(a) (normal level <75 nmol/L) as part of inclusion criteria

11/20 4 with an antisense oligonucleotide (ASO), 1 with a PCSK9i:
1-Tsimikas et al Lancet 2015, Phase 1; ncbi.nlm.nih.gov/pubmed/26210642. Main findings: baseline Lp(a) ~80-150 nmol/L, 40-78% reduction in Lp(a) with ASO vs placebo

12/20 2-Viney et al Lancet 2016, Phase 2- ncbi.nlm.nih.gov/pubmed/27665230 Main findings: Baseline Lp(a) 125–437 nmol/L or ≥438 nmol/L, 67-72% reduction in Lp(a) with ASO vs placebo

13/20 3-Viney et al Lancet 2016, Phase 1- ncbi.nlm.nih.gov/pubmed/27665230 Main findings: Baseline Lp(a) ~110-219 nmol/L, 66-92% reduction in Lp(a) with GalNac-ASO vs placebo

14/20 4-Stiekema et al- EHJ 2010, phase 4- ncbi.nlm.nih.gov/pubmed/30561610 Main findings: baseline Lp(a) ~200 nmol/L, 14% reduction in Lp(a) with PCSK9i (evolocumab), no change in FDT-PET uptake in carotids/aorta

15/20 5-Tsimikas et al NEJM 2020, Phase 2B- ncbi.nlm.nih.gov/pubmed/31893580 main findings; baseline Lp(a) ~200-250 nmol/L- 80% reduction in highest dose GalNac-ASO. 98% of patients got to goal of <125 nmol/L (<50 mg/dL)

16/20 There are several other studies ongoing where abstracts have been presented. Lp(a) HORIZON (@Novartis) is now performing a 7680 trial in patients with prior CV events and Lp(a) >70 mg/dL (~175 nmol/L) to assess reduction in new cardiac events- clinicaltrials.gov/ct2/show/NCT04…

17/20 Summary:

1-Despite discovery in 1963, it took over 50 years to come up with a therapy to lower Lp(a) effectively
2-New drugs can lower Lp(a) to normal levels by inhibiting the production of apo(a) in the liver

18/20 3-The evidence base for Lp(a) lowering and improving outcomes is lacking but we should know in 2024 when Lp(a) HORIZON is complete
4-Hope for having a therapy for patients with high Lp(a) to reduce CVD risk is greater than ever

19/x Quiz: 5 points: Choose correct answer.

The best evidence for determining whether a therapy works is:

1-Experts’ opinions and consensus
2-Years of observation of patient outcomes
3-A study where neither the patient or doctor know who got what
4-An experiment in animals

20/20 Bonus question: 1 point

Why do ASOs work so well to lower Lp(a):
1-They enhance apo(a) clearance
2-They block DNA from coding apo(a)
3-They inhibit the specific mRNA that codes for the apo(a) protein
4-They are magic