The two-hit hypothesis presented by Knudson in 1971 explains the development of tumours deficient in anti-oncogenes. Hamartomas in patients with tuberous sclerosis usually fit into this model, the first hit is a congenital lesion of either of the tuberous sclerosis genes (TSC1 or TSC2), and the second hit is loss of heterozygosity of this gene. Although this mechanism is true for most tumours associated with tuberous sclerosis, only 30-60% of brain and cardiac tumours show loss of heterozygosity--the remaining tumours develop despite the presence of an intact allele. Tumours in which loss of heterozygosity is rare, such as subependymal giant-cell astrocytoma, might all share a common feature that mimics loss of heterozygosity either by inactivation of the TSC complex or by direct activation of mammalian target of rapamycin (mTOR) or its downstream targets. Because phosphorylation of the TSC complex can inactivate it, expression and activation patterns of protein kinase B (AKT) and extracellular signal-regulated kinase (ERK), two potent protein kinases that are activators of the mTOR pathway, have been implicated. AKT activation is detected only in few samples, whereas ERK is hyperactive in all subependymal giant-cell astrocytomas. We postulate that ERK activation consistently detected in different tuberous-sclerosis-associated tumours is a molecular trigger for the development of these neoplasms.