In late 2026, the first useful observation did not come from a cardiovascular trial.
It came from a materials-science group studying why certain synthetic lipoprotein particles accumulated so efficiently inside diseased arterial tissue.
The particles were supposed to be imaging agents. They had been engineered to resemble apoB-containing lipoproteins just enough to cross dysfunctional endothelium and interact with the matrix of an atherosclerotic plaque, but they contained almost no cholesterol. Instead, their cores carried an inert fluorescent tracer.
In mice with advanced atherosclerosis, something odd happened.
The brightest particles did not distribute evenly through the plaque. They accumulated disproportionately in lipid-loaded macrophages, especially cells expressing a particular combination of scavenger receptors and stress markers associated with advanced foam-cell biology.
The investigators jokingly called them "Trojan LDL."
A year later the joke became a project.
The idea was simple enough to sound reckless.
If diseased plaque cells had spent years swallowing apoB particles, why not exploit that habit?
The first therapeutic construct was called AtheroVector-1.
It was not actually LDL. The team deliberately avoided creating a cholesterol-rich particle. Instead, they manufactured a biodegradable nanoparticle that reproduced selected physical and surface characteristics of an apoB particle while carrying no meaningful atherogenic lipid payload.
Its targeting was deliberately redundant.
The particle could bind weakly to matrix features enriched in atherosclerotic lesions.
It could also be taken up by scavenger-receptor-heavy foam cells.
And its outer shell was designed to become permeable only in an intracellular environment showing the oxidative and lysosomal characteristics common in severely lipid-loaded macrophages.
The researchers wanted three locks on the door.
Reaching the plaque was lock one.
Entering the wrong kind of cell was lock two.
Encountering the wrong intracellular environment was lock three.
Only after all three conditions were satisfied could the payload become active.
The first payload was modest.
It did not kill the cell.
It did not dissolve calcium.
It did not attempt to remove the plaque.
It carried two short-lived RNA components.
The first reduced expression of a protein involved in accelerated degradation of ABCA1.
The second blocked a microRNA pathway that normally suppresses ABCA1 and ABCG1 synthesis.
Individually, neither component did very much.
Together, in the intended cell state, they temporarily increased the amount of functional cholesterol-export machinery on the cell membrane.
The result, at least in culture, was dramatic.
Macrophages that looked like swollen droplets of cholesterol gradually changed shape.
Cholesteryl esters were mobilized.
ABCA1 transferred cholesterol toward ApoA-I.
ABCG1 transferred additional cholesterol toward HDL.
The cells did not disappear.
They became less foamy.
That distinction turned out to matter enormously.
Earlier plaque-directed therapies had sometimes made lesions worse by injuring macrophages. Dead macrophages that were not rapidly cleared became part of the necrotic core.
AtheroVector-1 tried to do the opposite.
It attempted to rescue the cell before killing it became necessary.
By 2028, the project had acquired a second component.
The team realized that unloading cholesterol solved only part of the problem.
Advanced plaques contained dying cells, oxidized lipids, extracellular debris, cholesterol crystals and poorly functioning macrophages. Cholesterol efflux alone did not guarantee healthy tissue repair.
So AtheroVector-2 was created.
It used the same targeting architecture but carried a different transient program.
Instead of increasing cholesterol export, it enhanced efferocytosis: the ability of macrophages to recognize and clear apoptotic cells before those cells ruptured.
The two treatments were not given simultaneously.
That was deliberate.
The researchers began describing the treatment as "plaque remodeling in phases."
Phase one reduced the lipid burden of surviving foam cells.
Phase two improved cleanup.
Phase three came later.
By then the investigators had become almost obsessive about the fibrous cap.
They knew that atherosclerosis was not merely a problem of excess material occupying space.
The most dangerous lesion could be relatively modest in size but possess a thin collagen-poor cap over a large lipid-rich necrotic core.
Rapidly removing cellular lipid while leaving the structural shell weak could, theoretically, create new mechanical problems.
So the third vector did almost nothing to macrophages.
Instead, it targeted cells at the fibrous-cap interface and delivered a short-duration signal intended to favor collagen production and stable smooth-muscle-cell behavior without provoking uncontrolled fibrosis.
This became the strangest feature of the therapy.
The treatment was not:
"destroy the plaque."
It was:
"change the plaque's ecology."
The sequence became known internally as U-C-R:
Unload.
Clear.
Reinforce.
The first convincing large-animal study appeared in 2029.
The animals had established coronary atherosclerosis.
They received intensive LDL lowering first.
That requirement never went away.
The investigators understood that directly unloading plaque while continuing to flood the arterial wall with apoB particles would be biologically absurd.
So circulating apoB had to be driven very low before treatment began.
Then came the three-vector sequence.
Imaging over several months showed something the team had hoped for but had refused to put in their grant proposal.
Low-attenuation plaque volume declined.
Macrophage lipid burden declined.
Necrotic cores became smaller.
Fibrous caps became thicker.
Total plaque volume fell modestly rather than spectacularly.
Calcified plaque changed much less.
Some calcium remained exactly where it was.
Nobody cared.
The treatment had never been designed to make CT scans look pristine.
Its purpose was to alter the biological part of the plaque most associated with future rupture.
The most important safety finding was what did not happen.
There was no surge in myocardial injury markers.
No wave of plaque hemorrhage.
No increase in thrombosis.
No evidence that macrophage populations throughout the body had been broadly reprogrammed.
The targeting was imperfect, but sufficiently selective to justify the next step.
In 2030, the first human trial began.
It was tiny.
Forty-eight patients.
All had established coronary disease, very low LDL levels despite maximal therapy, and persistent high-risk noncalcified plaque features on coronary CT or intravascular imaging.
Nobody in the trial was told that the treatment would reverse coronary disease.
The consent form used much less exciting language.
"Investigational plaque-directed RNA nanoparticle therapy."
The first six patients received approximately one-tenth of the predicted biologically active dose.
Nothing happened.
That was considered excellent news.
The doses increased slowly.
At intermediate doses, imaging tracers demonstrated that the particles reached coronary plaque.
Then researchers began finding the RNA payload inside plaque-associated cells sampled from patients undergoing clinically indicated procedures.
The platform had crossed its first human hurdle.
It could find its target.
The next question was much harder.
Did changing those cells change the plaque?
By 2032, the first meaningful imaging data appeared.
The effect was not miraculous.
Coronary calcium scores did not fall.
Many plaques remained visibly present.
But some of the features cardiologists feared most began changing.
Low-attenuation lipid-rich regions contracted.
Fibrous caps thickened.
Inflammatory imaging signals decreased.
Noncalcified plaque volume fell more than expected from LDL lowering alone.
Several lesions that had looked biologically active became smaller, denser and structurally quieter.
The press called it "arterial plaque reversal."
The researchers hated that phrase.
They preferred "directed plaque remodeling."
The second-generation therapy arrived almost immediately afterward.
The investigators had discovered that not all foam cells were macrophages.
Some originated from vascular smooth-muscle cells that had changed phenotype inside the plaque.
So AtheroVector-2B stopped asking:
"Is this a macrophage?"
Instead, it asked:
"Is this a lipid-loaded, oxidatively stressed plaque cell with impaired cholesterol export?"
The targeting logic became phenotype-based rather than lineage-based.
That improved specificity.
It also produced the most elegant version of the therapy.
The third-generation particle contained no permanently active drug.
It carried fragments of an RNA program that could only assemble inside cells displaying two disease-associated molecular signals.
One RNA component arrived with the first infusion.
A second arrived days later.
Neither could activate alone.
Only a cell that had accumulated both components and expressed the appropriate intracellular RNA signature could assemble the functional message.
Researchers called it coincidence gating.
Patients called it the two-key treatment.
The completed RNA program lasted only several days.
Then cellular enzymes destroyed it.
ABCA1 and ABCG1 activity rose temporarily.
Cholesterol export accelerated.
Lysosomal processing improved.
Efferocytosis increased.
Once the cell's lipid burden declined and its disease-associated signals faded, subsequent particles became much less likely to activate there.
The treatment effectively lost interest in cells it had successfully rehabilitated.
That feature turned out to be one of the platform's greatest advantages.
The healthier the cell became, the less strongly it attracted or activated the therapy.
By 2034, researchers were beginning to talk about something that would have sounded extravagant eight years earlier.
Not replacing statins.
Not replacing PCSK9 inhibitors.
Not replacing inclisiran.
But dividing coronary treatment into two complementary tasks.
The first task was prevention:
keep apoB concentrations extremely low so fewer atherogenic particles entered the arterial wall.
The second was remediation:
identify the residual pathological cell states left behind by decades of earlier exposure and selectively rehabilitate them.
For the first time, the treatment model resembled environmental cleanup.
Stop adding pollution.
Then clean the contaminated site.
Some plaques regressed substantially.
Others barely changed.
Heavily fibrotic and calcified lesions were especially resistant.
The therapy could evacuate lipid and improve cellular behavior, but it could not turn mature calcium and dense collagen back into a normal youthful artery.
That became an important boundary.
The drug was not a coronary fountain of youth.
It was much better at changing biologically active plaque than ancient scar-like plaque.
Another surprise came from stenosis.
Investigators originally hoped that dramatic plaque unloading would reopen arteries.
Sometimes it did improve lumen dimensions modestly.
Often it did not.
Arteries remodeled in complicated ways.
But clinical events appeared to decline even when stenosis measurements barely changed.
That reinforced an old lesson:
a lesion's danger was not determined solely by how narrow it made the artery.
Its biological composition mattered enormously.
Eventually the treatment received a name:
REMORA.
Regulated Endosomal Modulation of Atheroma by RNA Architecture.
The acronym was engineered shamelessly.
The metaphor was not.
Like the fish, the particle attached itself to something larger moving through the body.
But unlike a remora, it carried instructions.
The mature regimen looked almost anticlimactic.
Patients first achieved aggressive apoB lowering.
Advanced imaging identified lesions with substantial modifiable noncalcified plaque.
A targeting infusion mapped uptake.
Then, over several weeks, patients received sequential waves of plaque-seeking particles.
The first encouraged cholesterol export.
The second enhanced cleanup.
The third reinforced structural stability.
Afterward, treatment stopped.
No permanent gene editing.
No continuously active immune therapy.
No attempt to erase every calcified lesion.
The arterial wall was allowed to settle.
A year later, imaging determined whether another cycle was warranted.
The largest unresolved question remained the same one that had haunted the project from the beginning.
Could deliberately changing plaque biology ever create an unforeseen form of instability?
The developers never stopped worrying about it.
That caution shaped the entire platform.
Slow unloading instead of explosive dissolution.
Cell rescue instead of cell killing.
Transient RNA instead of permanent alteration.
Multiple targeting requirements instead of one receptor.
Sequential treatment instead of maximal simultaneous activation.
And continued LDL suppression throughout.
The treatment succeeded, in this fictional future, not because someone discovered a molecule capable of dissolving atherosclerosis.
It succeeded because researchers finally treated plaque as a living tissue ecosystem.
They stopped asking only:
"How do we lower cholesterol in the bloodstream?"
And began asking:
"What instructions would we give the diseased cells that have already spent twenty years living inside the plaque?"
The answer turned out not to be one instruction.
It was a conversation.
First:
"Stop storing cholesterol."
Then:
"Export what you can."
Then:
"Clean up the damage."
Then:
"Repair the structure."
And finally:
"Stand down."
That was the part the original investigators found most satisfying.
The treatment did not defeat the plaque by attacking it harder.
It taught the plaque to become less dangerous.
