The Aliquot

Cancer is not one disease. It is hundreds of diseases, each arising in a different tissue, driven by different mutations, following a different clinical course. For decades, this diversity made the field feel ungovernable — a catalog of exceptions rather than a science of rules. Then, in 2000, Hanahan and Weinberg proposed something audacious: that beneath all this variation, every human tumor acquires the same small set of biological capabilities. Six hallmarks. A shared logic for a disease that had seemed to defy logic.

Eleven years later, the same authors returned to that framework. The original six hallmarks held up. But the decade between had been extraordinarily productive, and the map needed redrawing. Two new capabilities had earned their place on the list. Two underlying processes had been recognized as the engines driving hallmark acquisition. And a whole dimension of tumor biology — the non-cancerous cells that surround and sustain a growing tumor — had moved from footnote to center stage.

The Original Six, Revisited

The six hallmarks that anchored the original framework remain the conceptual core of the updated paper. Cancer cells must sustain chronic proliferation, evade the growth suppressors that would normally restrain them, resist programmed cell death, achieve replicative immortality, induce the formation of new blood vessels, and ultimately invade surrounding tissue and spread to distant organs. Each of these capabilities, the authors argued in 2000, is a prerequisite for malignancy. That argument has only grown stronger.

What the intervening decade added was mechanistic depth. Take proliferative signaling. The broad outlines were known: cancer cells hijack growth factor pathways to drive their own division. What genome sequencing revealed was the specific circuitry involved. Roughly 40% of human melanomas carry activating mutations in BRAF, locking the B-Raf kinase into a constitutively active state and driving continuous signaling through the MAP kinase pathway without any upstream stimulus. Mutations in the catalytic subunit of PI3-kinase, similarly, hyperactivate the Akt/PKB pathway across a wide range of tumor types.

The more conceptually surprising finding was about feedback. Normal cells are wired with negative-feedback loops that dampen signaling after it has served its purpose. Ras, the archetypal oncoprotein, is not simply a hyperactive accelerator in cancer cells — its oncogenic mutations specifically disable its own GTPase activity, the intrinsic brake that would normally terminate signaling. PTEN, which counteracts PI3-kinase by degrading its lipid product, is frequently silenced by promoter methylation in human tumors. The disruption of these self-attenuating circuits is now recognized as a widespread mechanism by which cancer cells achieve proliferative independence, and it carries a direct clinical implication: blocking mTOR pharmacologically can paradoxically increase PI3K activity by releasing a negative-feedback loop, blunting the drug's intended effect.

The cell death hallmark also gained new complexity. Apoptosis — the orderly, immunologically silent form of programmed cell death — had been well characterized by 2000. The decade since revealed two additional layers. Autophagy, a cellular recycling program that breaks down organelles to generate fuel under nutrient stress, turns out to be a double agent. Mice lacking key autophagy components like Beclin-1 show increased cancer susceptibility, suggesting autophagy normally acts as a tumor-suppressive barrier. Yet in established tumors, the same program can protect cancer cells from starvation and from cytotoxic therapy, enabling a state of reversible dormancy. The conditions that determine which role autophagy plays remain an open question.

Necrosis, long viewed as passive cellular collapse, also proved more consequential than assumed. Unlike apoptosis, necrotic cell death releases proinflammatory signals into the surrounding tissue, recruiting immune cells that — in the context of a growing tumor — can actively promote angiogenesis, proliferation, and invasion. Tolerating a degree of necrosis may therefore benefit tumors by drawing in growth-stimulating inflammatory cells, even as it eliminates some cancer cells.

Replicative immortality, meanwhile, acquired a new wrinkle around telomerase. The enzyme, expressed in roughly 90% of spontaneously immortalized cells including cancer cells, was known to maintain telomere length and prevent the chromosomal crisis that would otherwise kill dividing cells. What emerged was a more dynamic picture: early in tumor development, before telomerase is activated, progressive telomere shortening can actually generate tumor-promoting mutations through cycles of chromosomal breakage and fusion. The delayed acquisition of telomerase then stabilizes this mutant genome and confers the unlimited replicative capacity that macroscopic tumors require. Telomerase, it turns out, also has functions entirely independent of telomere maintenance — including amplifying Wnt signaling as a cofactor of the β-catenin/LEF transcription factor complex.

The Engines Underneath: Genome Instability and Inflammation

The six hallmarks describe what cancer cells do. Two enabling characteristics describe how they get there.

Genome instability is the more fundamental of the two. Acquiring the full roster of hallmark capabilities requires a succession of specific genetic changes, and the normal rate of spontaneous mutation is far too low to generate them within a human lifetime. Cancer cells solve this problem by dismantling the machinery that normally detects and repairs DNA damage. Defects in so-called caretaker genes — those encoding components of mismatch repair, nucleotide excision repair, and double-strand break surveillance — accelerate the accumulation of mutations across the genome. TP53, the "guardian of the genome," sits at the center of this surveillance network; its loss removes a critical checkpoint that would otherwise force genetically damaged cells into senescence or apoptosis. The result is a mutator phenotype: a cell that generates genetic variation far faster than normal, providing the raw material on which selection can act to produce hallmark capabilities. Epigenetic changes — DNA methylation, histone modifications — can produce the same heritable phenotypic shifts without altering the DNA sequence itself, adding another layer to this evolutionary process.

Tumor-promoting inflammation is the second enabling characteristic, and its recognition represents one of the more striking conceptual shifts of the past decade. Immune cells were long assumed to be adversaries of tumors, and in some contexts they are. But the inflammatory cells that infiltrate premalignant and malignant lesions — macrophages, neutrophils, mast cells, myeloid progenitors — can supply tumors with growth factors, survival signals, proangiogenic factors, and matrix-degrading enzymes that actively support hallmark acquisition. The same immune system designed to eliminate pathogens and heal wounds can be co-opted to build a more hospitable environment for a growing tumor.

Two New Territories: Metabolism and Immune Evasion

The two emerging hallmarks proposed in this update reflect areas where the evidence had grown compelling enough to warrant inclusion, even if not yet fully generalized across all cancer types.

The first is the reprogramming of energy metabolism. Normal cells generate ATP primarily through oxidative phosphorylation in mitochondria, a highly efficient process that requires oxygen. Cancer cells, even in the presence of adequate oxygen, preferentially use aerobic glycolysis — converting glucose to lactate rather than fully oxidizing it. This is the Warburg effect, first described in the 1920s and long regarded as a metabolic curiosity. Its functional logic has become clearer: glycolysis, though less efficient at generating ATP, produces the biosynthetic precursors — nucleotides, amino acids, lipids — that rapidly dividing cells need to build new cellular material. The shift is not simply a passive consequence of mitochondrial dysfunction; it is an active reprogramming, regulated by oncogenes including MYC and by the hypoxia-inducible factor HIF-1α, that supports the anabolic demands of continuous proliferation.

The second emerging hallmark is immune evasion. The immune system does eliminate many nascent tumors — a process called immunosurveillance. The tumors that persist are those that have found ways to escape it. Some cancer cells reduce the surface display of antigens that would flag them for destruction by cytotoxic T lymphocytes. Others secrete immunosuppressive signals, or recruit regulatory T cells and other immunosuppressive cell populations, that paralyze the immune response within the tumor. The dichotomy is important: the immune system both antagonizes and, through its inflammatory arm, promotes tumor development. Recognizing immune evasion as a hallmark opened the conceptual door to a class of therapies — immune checkpoint inhibitors, among them — that have since transformed oncology.

The Hidden Ecosystem: A Tumor Is Not Just Cancer Cells

Perhaps the most consequential conceptual shift in the updated framework is the elevation of the tumor microenvironment from supporting character to co-protagonist. A solid tumor is not a mass of cancer cells. It is a complex tissue, and the non-cancerous cells within it — cancer-associated fibroblasts, endothelial cells, pericytes, and a diverse array of immune cells — are not passive bystanders. They are active participants in tumor progression, recruited and reprogrammed by cancer cells, and in turn supplying the signals that enable hallmark acquisition.

The signaling between these cell types is reciprocal and elaborate. Cancer-associated fibroblasts secrete hepatocyte growth factor (HGF) in response to TGF-β signals from cancer cells; the HGF then feeds back to promote cancer cell proliferation and invasion. Tumor-associated macrophages supply matrix-degrading metalloproteinases and cysteine cathepsin proteases that clear a path for invading cancer cells, and in experimental models of metastatic breast cancer, macrophages supply epidermal growth factor to cancer cells while the cancer cells reciprocally stimulate the macrophages with CSF-1, a paracrine loop that facilitates intravasation into the bloodstream. Mesenchymal stem cells in the tumor stroma secrete CCL5/RANTES in response to cancer cell signals, which then acts back on the cancer cells to stimulate invasive behavior.

The epithelial-mesenchymal transition (EMT) — a developmental program normally used during embryogenesis — is a key mechanism by which cancer cells acquire invasive capacity. Transcription factors including Snail, Slug, Twist, and Zeb1/2 orchestrate a shift from an epithelial to a mesenchymal cell state, suppressing CDH1 (E-cadherin) expression, increasing motility, and heightening resistance to apoptosis. Contextual signals from the surrounding stroma can trigger this transition, meaning that the invasive phenotype is not purely cell-autonomous — it depends on the tumor's ecosystem. The reverse transition, MET, may occur when disseminated cells reach a distant tissue, allowing metastatic colonies to re-establish an epithelial architecture.

Metastatic colonization remains the least understood step in this cascade. Many disseminated cancer cells successfully reach distant tissues but never form macroscopic tumors, persisting instead as dormant micrometastases for years or decades. The barriers to colonization likely include the inability to activate angiogenesis in a foreign tissue, anti-growth signals embedded in normal extracellular matrix, and active suppression by the immune system. When those barriers are eventually overcome — through further genetic evolution, changes in the tissue microenvironment, or both — the clinical consequences can be devastating.

A Roadmap for Treatment

The hallmark framework was always intended to be more than a classification scheme. Its practical value lies in providing a logical basis for therapy: if every tumor must acquire the same core capabilities, then drugs targeting those capabilities should work across tumor types, and combining drugs that target different hallmarks should be harder for tumors to escape.

That logic is now being tested in the clinic. EGFR inhibitors target proliferative signaling. VEGF pathway inhibitors target angiogenesis. CDK4/6 inhibitors target the cell cycle machinery downstream of the RB pathway. Immune checkpoint antibodies targeting CTLA-4 address the immune evasion hallmark. Aerobic glycolysis inhibitors are in development to target the metabolic reprogramming hallmark. The framework predicts that combining agents across hallmarks, rather than hitting the same pathway harder, will produce more durable responses — because a tumor that evolves resistance to one targeted therapy must simultaneously maintain all its other hallmark capabilities, and each represents a potential vulnerability.

What the Framework Cannot Yet Explain

The hallmarks paper is a synthesis, not an experiment, and its limitations are those of the field it summarizes. Three gaps stand out.

The role of epigenetics remains underspecified. DNA methylation and histone modifications can produce heritable changes in gene expression that mimic the effects of mutations — silencing tumor suppressors, activating oncogenes — without altering the DNA sequence. The authors acknowledge this, but the mechanistic integration of epigenetic reprogramming into the hallmark framework is still rudimentary. As single-cell epigenomic tools mature, this will need to be addressed more directly.

MicroRNAs present a similar gap. Hundreds of small non-coding RNAs are dysregulated in cancer, and some have clear functional roles in regulating hallmark-relevant genes. Their systematic incorporation into the circuit diagrams that the paper uses to represent cancer cell signaling has not yet happened, and the authors flag this as a priority.

The tumor microenvironment sections, while conceptually rich, rest on a relatively thin experimental base for some cell types. The signaling diagrams are, by the authors' own admission, rudimentary. The field has since developed far more sophisticated tools — single-cell RNA sequencing, spatial transcriptomics, multiplexed tissue imaging — that are beginning to reveal the true complexity of stromal-cancer cell interactions at a resolution the 2011 paper could not access. The framework provides the right questions; the answers are still being assembled.

One productive direction the paper gestures toward but does not fully develop: the possibility that some hallmark capabilities can be acquired not through cancer cell-intrinsic mutations but through signals from the microenvironment. If invasive behavior can be induced by stromal signals without requiring additional mutations in the cancer cell, then targeting the stroma may be as important as targeting the cancer cell itself — and the two strategies may need to be combined to achieve durable control.