Does it make sense to have less than 30 documents and more than 10000 words in Latent Dirichlet Allocation?

Question

We are trying find out topics more covered by political parties in italy using LDA and we have about 30 parties and lots of words (8000). Does it makes sense?

Since the words are taken from daily documents which are mixtures of parties speeches we can replace the parties with the documents and just study them (which are around 1000)