sample things other than copy number (e.g.,
HER2 or EGFR amplifications). That type of
technology I think will be critical.
KP: I would also argue that we need to exploit
more testing at autopsy. For example, we
have analyzed all the metastases in the same
patient using a rapid autopsy. And when we
did the mutational analysis, the metastases
had different variants. When we compared
metastases in the lung and the liver and some
other organs using FISH [fluorescence in situ
hybridization], they were not the same. This
raises the question, do we really understand a
patient’s cancer if we just analyze the primary
tumor? Because at the moment this is what
we are doing to diagnose the patient and then
design the treatment. Instead, one would
like to sample metastatic lesions or develop
models that predict what kind of metastatic
lesions may develop from a primary tumor.
RMH: I think one of the most useful additions
to our present diagnostic tools would
be to have a marker—especially a fluorescent
marker—that would work in the clinic that
tells us whether a cancer cell is dividing or
not. That would help guide what kind of drug