You can supplement the CPCI with the SOPA if you want more information on pain attitudes.
FWIW, when I've seen an MMPI used in an assessment report (predominantly neuropsych), I can't think of any instances where the version used has been anything other than the MMPI-2-RF -- That might be idiosyncratic, but I mention it to underscore that the -RF is likely a solid investment if you're doing a lot of assessment work, even if it's assessment work with non-pain/pre-surg patients.
In terms of using the first 370 questions of the MMPI, I'd be wary of deviating too far from standard administration practices (e.g., having to defend uncommon methodology in court) -- I don't know the literature at all on this practice, so I don't how empirically/psychometrically sound of a decision it is. IMO, the peace of mind that I would enjoy from using something well validated like the RF would likely be worth the initial cost of purchasing it.