I'm just speculating on your application's design here, but why would you need to use the UrlFetch API to retrieve something you could easily get instantiating and using your business logic classes directly?
Also note this from the Urlfetch documentation:
To prevent an app from causing an endless recursion of requests, a
request handler is not allowed to fetch its own URL.
If you are really forced to do this, then set the proper Cache-Control
header on your request:
headers = {'Cache-Control':'no-cache,max-age=0', 'Pragma':'no-cache'}
result = urlfetch.fetch("http://foo/route", None, urlfetch.GET, headers)
Have a look to this Issue.