Unfortunately Emanuel, several studies have been published that show RNA degradation occurs over long storage periods even at -80°C, at -20°C this would only be exacerbated.  Your best bet would probably be to mix your plasma samples with an appropriate amount of RNAlater prior to freezing.

Thank you for your reply. It would be interesting evaluate the difference of  RNA integrity between -80°C and -20°C. If the loss at -20°C after 1 month from the collection is small, I think it is accettable for my purpose. 

If you are interested in more information, consider reading The procurement, storage, and quality assurance of frozen blood and tissue biospecimens in pathology, biorepository, and biobank settings by Shabihkhani, et al. (2014).  Although I may assume that the degradation incured over only a month would be negligible, your best bet would be to establish a RIN for your RNA sample prior to use in the future.

I have found this protocol on http://www.sigmaaldrich.com/technical-documents/protocols/biology/rnalater.html

White Blood Cells White blood cells can be effectively preserved in RNAlater if they are separated from the red blood cells and serum, and treated as tissue culture cells. RNAlater is not recommended for preserving RNA in whole blood, plasma, or serum. Because of their high protein content, these fluids will form an insoluble precipitate if they are mixed with RNAlater. 

Nevertheless in this paper http://www.ncbi.nlm.nih.gov/pubmed/12076286 the authors found that:

"Quantitative PCR data showed that a 60:40 or greater ratio of RNAlater:plasma volume successfully stabilized HCV RNA and HIV RNA in plasma for up to 28 days at 37 degrees".

I think if I use a lower ratio of RNAlater:plasma (i.e 20:80 or 40:60) and I store the samples at -20°C maybe I can stabilized RNA and at the same time reduce the formation of insoluble precipitate.

Does it make sense?