Blinded Sample Size Re-estimation for Longitudinal Overdispersed Count Data in Randomized Clinical Trials with an Application in Multiple Sclerosis