I'm not sure about the 4 years thing. I don't think its much of a bump.
 Maybe you could ask your doctor for a literature reference. I seem to
recall having seen one somewhere.  The answer should be there in the
SEER data, but I don't know of any available extract graph that
specifically shows this information.

http://seer.cancer.gov

The main reason why the risk curve starts low and initially rises is the
choice of staring point.  At the time of surgery you are known to have
no symptomatic or detectable metastases, those patients that do aren't
part of these statistics.  That means the likelihood of you developing
symptoms the next day is very small - you aren't being examined and
staged every day, and it takes a few months for any metastases to grow
from undetectable to symptomatic.

Beyond this 'statistical artefact' the incidence of recurrence generally
falls steadily over time as the population of residual cancer cells (or
cancer stem cells according to some theories) falls as they are mopped
up by the immune system.  But it never falls as low as the general
population's base risk because some breast cells have shown a propensity
for going cancerous, and you've got billions more similar cells still in
your body, that might someday choose to take the same path.

In looking at survival curves you have to take into account general life
expectancy.  If you consider women diagnosed with cancer at say age 80
it is hardly surprising that most don't survive beyond 20 years simply
because most people don't live to be 100, nothing to do with their
cancer. So for your purposes you are better to look at recurrence data
than survival data, unless you restrict the search to your own age group.

Tim

Yes, Tammy, just make the most of what you have and don't spend time trying
to understand statistics. You're not a statistic, you're a woman with a life
to lead :-)

Mary