I am a Stata user, so I can only give some broad advice for SAS. I am assuming that you have your treatments specified as indicator (dummy) variables in your data (with one holdout for the base category). If that is the case, then you just need to add an additional dummy variable for the additional treatment. When a subject has this treatment level, all the other dummies will equal 0, and this will equal 1.

If you have all your treatments listed in one variable (ie., 0, 1, 2, 3, ...T), then I assume you can tell SAS that this variable is categorical. In Stata, categorical variables are noted with a prefix "i."

Thank's Ariel, but I could perform the analysis by using the contrast statement of GLM procedure.

I was responding to your statement that you don't know how to put in the additional treatment in the model. That is not a clear statement. I assume that if you are able to run contrasts between treatments, that you have all your treatments in the model.

Perhaps have a look at the following paper.  It includes SAS code examples for the dummy variables referred to by Ariel to achieve what you are looking to do. 

Marini, R.P. 2003. Approaches to Analyzing Experiments of Treatments Plus Other

Treatments. Hort. Sci. 38:117-120

Hope this helps.

Thank's Robert. That was the exactly paper that I used for my analysis. Thank's anyway.

It's a good paper, Robert.

Jurandy, have we helped you with your issue? If not, I suggest that you clarify it better for us

Ariel