Yes, we live in one. What was regarded as mathematics 2000 years ago is not what we regard as mathematics today. Gauss published the first acceptable proof of the Fundamental Theorem of Algebra; but Gauss's proof would not be acceptable from an undergrad today. Standards of rigor, as well as our understanding of the topology of the real line, have changed considerably since then.
Mathematics is a historically-contingent activity of humans. Not only could mathematics be different on a different planet or in another universe; which are of course unprovable one way or the other; but mathematics could and actually has been different at different eras on this planet.
Just consider the rise of computers, experimental mathematics, machine proof systems, and computatibility theory. It's likely that math in 100 years will be very different than math is now. Zermelo-Fraenkel set theory is less than 100 years old. What if on some other planet they never discovered it, but rather skipped to some other framework?
Now, you may be referring not to the mathematics as a historically and culturally contingent human activity; but rather as some sort of Platonic thing that is "out there" that we can discover. To which I'd ask: Where is your evidence that such a thing exists? And if it does, then which human mathematics is the one, true mathematics? The math of 1000 years ago? The math of today? Or the math of 1000 years from now?
I do realize that you're asking if it's possible that in some other universe, 2 + 2 is 3. I have no idea. I don't think the question is meaningful. I think I'm wearing my formalist hat today.