your cv genetics, without genome mysticism

tl;dr

• the missing problem was not “we need more genes”.
• the missing problem was that the old pipeline stopped halfway.
• now the two biggest holes are much cleaner:
• ABCG2 no longer looks like a vague rosuvastatin warning
• LDLR/APOB/PCSK9 no longer look like scary row counts without meaning

what changed

1. statin pharmacogenomics

we already knew:

• SLCO1B1 rs4149056 = CT
• SLCO1B1 rs2306283 = AG

that still means: statin entry is not perfectly neutral.

but the old read left ABCG2 rs2231142 = GG hanging as “maybe risk”.

that ambiguity is now resolved.

the issue was strand language:

• CPIC describes the variant at the transcript level (c.421C>A)
• ABCG2 sits on the minus strand
• 23andMe reports genomic letters

once you harmonize those layers, your GG does not carry the decreased-function allele at this locus.

practical meaning:

• ezetimibe-first still looks elegant
• but rosuvastatin low-dose now looks cleaner than before
• the remaining genetic friction is mostly SLCO1B1, not ABCG2

2. hidden FH panic went down

before, the WGS layer only said:

• there are many non-reference rows in LDLR, APOB, PCSK9

that was true, but not useful.

now those regions were run through consequence-aware annotation.

LDLR

• no missense / stop / frameshift hit emerged
• only a splice-region variant and several synonymous/common-like variants remained

APOB

• no classic R3527Q hit (rs5742904 = CC)
• coding variation exists, but this stage does not support a monogenic-FH claim

PCSK9

• no common protective R46L (rs11591147 = GG)
• common coding/splice texture exists
• still not a clear monogenic-FH verdict

practical meaning:

your genome does not currently read like:

“hidden severe monogenic LDL disease explains everything”

the phenotype still leads:

• elevated Lp(a)
• decent apoB
• decent LDL-C
• prevention intensity should be set mostly from that, plus imaging and blood pressure

what the genome does say clearly

• LPA support is real:
• rs10455872 = AG
• rs3798220 = TT
• rs6415084 = CT
• APOE looks like ε2/ε3, which is a softer LDL background than ε4
• 9p21 common CAD-risk texture is present
• F5 Leiden is not present

what this changes in treatment logic

before

• ezetimibe-first looked clean
• rosuvastatin had unresolved transporter ambiguity
• FH-region layer added raw anxiety

now

• ezetimibe-first still looks like the cleanest first pharma move
• rosuvastatin low-dose now looks more defensible than before
• the genome does not force a dramatic “hidden FH, therefore brute-force statin now” story

current verdict

if the goal is:

elegance + low friction

start with ezetimibe-first

deeper LDL/apoB lowering now

rosuvastatin low-dose is more acceptable than it looked in the older read

very aggressive prevention

that still depends more on:

• fresh apoB
• clean home BP
• maybe CAC / PWV

than on adding more common SNPs

bottom line

the best thing genetics did here was not “find the answer”.

it removed two kinds of fake complexity:

1. fake rosuvastatin fear from unresolved ABCG2
2. fake hidden-FH fear from raw region row counts

that leaves a cleaner real choice:

• ezetimibe-first
• or low-dose rosuvastatin if you want stronger lowering and accept a less frictionless entry