I'm exploring a theoretical cancer treatment strategy that approaches tumors as complex, evolving ecosystems rather than isolated molecular targets. The idea stems from a fundamental limitation I see in current therapies:

Most mainstream approaches — immune checkpoint inhibitors, CAR-T, targeted drugs — all attempt to reprogram or stimulate the body's own biological systems (immune cells, signaling pathways, etc.) to fight cancer. But these systems are highly adaptive, noisy, and embedded in feedback loops we don't fully control. We're essentially trying to rewrite parts of the system using its own unstable components.

What if we flipped the logic?

Instead of fighting cancer from outside, what if we intentionally designed a new population of genetically engineered cancer cells that could invade, dominate, and replace the existing tumor cells — and then be selectively destroyed?

The proposed strategy would involve:

• Engineering cancer-like cells with enhanced fitness in the tumor microenvironment, so they can outcompete native tumor cells for nutrients, signaling access, and space.
• Equipping these engineered cells with custom synthetic targets or antigens not found in normal tissue.
• Once these cells dominate the tumor mass, we would trigger a targeted therapy (drug, antibody, or immune response) specifically designed to recognize and eliminate the synthetic targets.

This would be a kind of Trojan horse approach: turn the tumor into something predictable and targetable — a “homogenized”, engineered tumor — and then destroy it.

This strategy has echoes of:

• Immunotoxins or suicide switches in CAR-T therapy
• Synthetic biology-based payload cells or viral carriers
• Tumor evolution modeling via engineered clones

But to my knowledge, no one has yet tried to build a full “competitive engineered cancer cell → tumor replacement → synthetic antigen targeting” pipeline.

I realize this comes with massive risks:

• Stability and mutational drift of engineered cells
• Ensuring selective expansion in the tumor but nowhere else
• Avoiding premature immune clearance of the engineered cells
• Ethical concerns around injecting proliferative, engineered cells into patients

That said, it might be a way to achieve orthogonality — creating a controllable subsystem that doesn't rely on fragile native interfaces. I believe solving cancer will require this kind of orthogonal control logic, not just patching failing pathways with internal tweaks.

Has anyone explored something like this experimentally — even in organoids or mouse models? Would love to hear thoughts, objections, or references.

Also open to collaborators if someone wants to formalize this into a position paper or conceptual model.