Epigenetic progenitor theory: cancer without a first mutation

The claim. Feinberg, Ohlsson and Henikoff proposed in 2006 that the first step in cancer is an epigenetic disruption of progenitor cells within a tissue, polyclonal and driven by environment, injury or age, which expands a pool of cells primed for transformation. A gatekeeper mutation then initiates a tumour in that pool, and epigenetic plasticity supplies the heterogeneity that lets the tumour evolve and resist treatment. The broader epigenetic view (Baylin, Jones, Bernstein) holds that DNA methylation, histone marks and chromatin state are causes of cancer behaviour and not only consequences.

Who and when. Feinberg and Vogelstein reported global hypomethylation in human cancers in 1983, the first cancer epigenetic lesion. Baylin and Jones described promoter hypermethylation silencing tumour suppressor genes in the 1990s. Feinberg's progenitor model appeared in 2006 and was updated in 2016; Flavahan, Gaskell and Bernstein connected epigenetic plasticity to every hallmark in 2017.

Evidence for. Every human cancer studied shows methylation changes, and a methylation profile classifies brain tumours more accurately than histology. Some childhood cancers have almost no recurrent mutations: Mack and colleagues showed in 2014 that lethal infant ependymomas are defined by a methylation programme rather than mutations, and diffuse midline gliomas are driven by a single histone change (H3 K27M). Chromatin regulators are among the most frequently mutated genes (ARID1A and other SWI/SNF subunits, DNMT3A, TET2, EZH2, KMT2A), which shows that the genome selects for epigenetic disruption. Drug-tolerant persister cells survive by reversible epigenetic states, and lineage switching under treatment happens without new mutations. Oncometabolites from IDH mutations cause cancer by blocking demethylation.

Evidence against and limits. Most epigenetic changes could be consequences of transformation rather than causes, and causality is hard to prove without a mutation to trace. Epigenetic drugs work in myeloid cancers and lymphomas and have largely failed in solid tumours. The polyclonal epigenetic field predicted by the model is difficult to observe directly in human tissue.

Predictions that held or failed. Held: Hanahan added non-mutational epigenetic reprogramming as a hallmark in 2022; azacitidine, decitabine, HDAC inhibitors, tazemetostat (EZH2), IDH inhibitors and menin inhibitors are approved; methylation-based cell-free DNA tests detect and classify cancers; persister states are reversible. Failed or unfulfilled: broad reversal of the cancer epigenome in solid tumours; epigenetic priming to make cold tumours respond to immunotherapy has not yet succeeded in a phase 3 trial.

Therapies that came from it. Hypomethylating agents in myelodysplastic syndromes and acute myeloid leukaemia, EZH2 and menin inhibitors, IDH inhibitors, methylation profiling for diagnosis and liquid biopsy, and the emerging field of epigenetic editing. It feeds the cancer stem cell and plasticity view and the 2022 hallmarks, and it is one of the theories framed against the somatic mutation theory.

Status: partly confirmed. Epigenetic disruption is established as a driver of cancer behaviour and, in a minority of cancers, as the initiating event; whether it precedes mutation in the common adult cancers is still being worked out.