From genome to therapy: cancer as a genomic disease, CRISPR editing and mRNA vaccines

From a molecular standpoint, cancer is best understood as a disease of the genome that accumulates over the course of a lifetime. Tumor cells carry acquired genetic alterations, and one conceptual distinction organizes the entire field: driver mutations — few and selectively advantageous, which effectively propel growth — versus passenger mutations, the vast majority, biologically inert from the standpoint of selection. A classic synthesis estimated that typical tumors harbor 2 to 8 driver mutations surrounded by many passengers, providing the basis for separating what is a target from what is genomic noise (Vogelstein, 2013).

Against this backdrop acts genomic instability. In the framework of the hallmarks of cancer, genomic instability and mutation are described as enabling characteristics — they are not the engine of the tumor itself, but the mechanism that generates the genetic diversity upon which selection operates. The 2022 update to that framework added dimensions such as phenotypic plasticity, non-mutational epigenetic reprogramming and the role of polymorphic microbiomes, broadening the reading beyond strict genetics (Hanahan, 2022).

It is from this biology that the logic of precision medicine is born. If it is a specific molecular alteration — and not merely the organ of origin — that underlies the tumor's behavior, then characterizing that alteration can guide treatment. The promise is appealing, but it demands interpretive discipline: identifying a driver mutation does not guarantee that a targeted therapy exists against it, nor that the response will be durable. The rest of this article traces where this logic has already translated into approved practice and where it remains a hypothesis under test.

CRISPR-Cas9 gene editing is no longer exclusively laboratory-based, but it is essential to map precisely what has reached the clinic. The approved milestone is ex vivo editing: the patient's own hematopoietic stem cells are removed, edited outside the body — in this case, targeting the BCL11A gene to reactivate fetal hemoglobin — and reinfused. The trial that grounded this strategy eliminated vaso-occlusive crises and transfusion dependence in the first patients (Frangoul, 2021). This foundation supported exagamglogene autotemcel (Casgevy), approved by the FDA on December 8, 2023 for sickle cell anemia and on January 16, 2024 for transfusion-dependent beta-thalassemia; in 2026, the indication was expanded to children from 2 years of age. Let the essential be recorded: these are hereditary hematologic diseases, not cancer.

In vivo editing — in which the editor is administered directly into the bloodstream to act inside the body — had its first human proof of concept with NTLA-2001 in transthyretin amyloidosis, reducing circulating TTR protein after a single infusion (Gillmore, 2021). This program (nexiguran ziclumeran) advanced to phase 3 studies by 2026, but remains investigational, without approval. Its trajectory also illustrates the real limits of the technology: the phase 3 trials even underwent a temporary regulatory pause after a serious hepatic adverse event, a reminder that long-term safety is a hypothesis to be verified, not a premise.

In cancer specifically, CRISPR enters via the route of cell therapy. The first human trial of CRISPR-edited T cells — with deletion of TRAC, TRBC and PDCD1 and transduction of an anti-NY-ESO-1 receptor — proved feasible and safe in refractory cancer, with cell persistence for up to 9 months (Stadtmauer, 2020). It is crucial to read this result for what it is: a phase 1 study designed to assess feasibility and safety, not to prove efficacy. The distinction between 'approved' and 'investigational' organizes the following table.

No honest discussion of gene editing can dispense with governance. The central distinction is between somatic and germline editing. Somatic editing alters cells of the patient's own body, without transmission to offspring — it is the one in approved clinical use (such as exa-cel) and in trials. Germline editing, heritable by subsequent generations, remains widely proscribed by international ethical-scientific consensus, for reasons that combine safety, irreversibility and social implications. Under no circumstances should it be suggested that germline editing is an accepted practice: it is not.

On the technical plane, three uncertainties weigh. First, off-target effects: the editor may, in principle, cut unintended sites of the genome, which demands vigilance. Second, long-term durability and safety are unknown — the approved therapies are recent, and the follow-up horizon is still short; the regulatory pause experienced by an in vivo editing program after a serious hepatic adverse event is a concrete reminder of this. Third, even efficacy demonstrated in a trial does not eliminate individual variability of response.

There is also the dimension of cost, access and equity, frequently omitted amid technological enthusiasm. Gene-editing therapies and personalized vaccines involve high cost and complex logistics — individual sequencing, bespoke manufacturing, specialized centers. Without confronting this point, a distorted expectation is created: a technology that exists but is not accessible is not, yet, a public health solution. Separating evidence from expectation, here, is also a matter of informational justice.