tumors might respond to treatment. In particular,
it challenges the simplistic notion of neatly
binning cancer mutations as tumorigenesispromoting
‘drivers’ versus incidental ‘passengers’.
“It may help to get away from the ‘once
a passenger always a passenger’ mentality and
start thinking about some passenger events as
‘drivers in waiting’,” says Swanton. “It could be
following treatment or metastatic outgrowth or
during growth in hypoxic conditions, but whatever
the insult, it may be good to consider some
passenger mutations as simply drivers waiting
for the right context.”
But there is a silver lining. Amid all the
heterogeneity, Swanton and colleagues
observed clear signs of convergent evolution,
such that even ‘branches’ with different
mutations might ultimately acquire the same
defects—for example, different inactivating
mutations in PTEN. “This tells us there’s a
strong selective pressure for these events to
happen,” says Swanton. He also notes that
this mutational complexity obviously does
not negate the existence of common driver
mutations in genes that tend to be early triggers
of tumorigenesis and are likely to consistently
be found in a tumor’s evolutionary
‘trunk’. Nevertheless, it’s clear that much
remains to be learned about how these mutations
function in individual patients and how
this tumor diversity originates. “How much
heterogeneity exists in different tumor types
for the real drivers of the oncogenic process?”
asks Phil Stephens, vice president of
Cancer Genomics at Foundation Medicine.
“I think that’s largely unknown, and it’s likely
to be different for different tumor types, and
to depend on the treatment of that cancer
specimen