I am going to examine the miRNA expression in FFPE samples but my archive is 5 years old. This age made me worry about the reliable expression mode, what is your opinion?
In my opinion 5 years of storage may be a problem with miRNA quality but you may try to assess small RNA quality and concentration using the small RNA chips for agilent bioanalyzer on a few samples before to start the full study.
It has already been described that miRNA expression from FFPE samples even older than 5 years (up to 10 years) is quite comparable to younger samples (Yaguang Xi, Go Nakajima, et al. Systematic analysis of microRNA expression of RNA extracted from fresh frozen and formalin-fixed paraffin-embedded samples. RNA. 2007 Oct;13(10):1668-74). In our group we even used up to 17 years old FFPE samples and we got comparable miRNA expression results to those isolated from cryopreserved tissue (Wach S, Nolte E, et al. MicroRNA profiles of prostate carcinoma detected by multiplatform microRNA screening. Int J Cancer. 2012 Feb;130(3):611-21).
I'd say go with Elke's suggestions. For FFPE tissues it depends more how you normalize and what you exactly want to do. The RNA will be degraded with the very strange exception of microRNAs. You can look at RNA profiles from very good samples (RIN close to 10) and from very bad samples (RIN 4 or lower) and the peak for small RNAs is more or less completely unchanged. As said, if your normalizer is outside this peak you may be doomed. Best to use a microRNA or a few microRNAs as normalizer(s) in case you want to go for qRT-PCR. For some weird reason, microRNAs are the last thing to get degraded...