You have a way to check your gene? because you could make a ESTs like library, cloning your cDNA to vectors, check which colony has your gene of interest, and then use primers in the vector to sequence your gene.
The problem with this is the ammount of clones you are going to check if your gene is low represented in the whole cDNA.
I think RACE PCR still a best choice, but this sounds like it could work.
I suppose you could use a gene-specific second stand synthesis primer, end-polish/dA-tail your cDNA fragments, and TOPO-TA clone them into plasmid? You'd probably need quite a large amount of input material for this, and you'll probably need to use a colony PCR screen using the gene-specific primer + backbone derived primer to select correct clones. Should work though.