Clinical Study and Biostatistics

Hy everyone,
I have measured a biological parameter in a cohort of 20 patients before they start a therapy (baseline) and then every 2 months for a maximum of 2 years (follow-up).
So, I have 20 measurement at the beginning of the study but the last months of the study I have only 4/5 measurement, because of the patients start therapy at different moment.
My question is simple, it is possible to analyze the trend of my parameter during the time with a test for repeated measures (such as Mixed Model under SAS), despite of missing data?
Thank you very much for your help.
I think you could bootstrap the data using a time series and then analyze that data for repeated measures, but I admit I'm not 100% sure. Bayesian statistics seem like the answer here.
Thank you very much Andy for your answer, but my knowledge about statistic is limited. I don't know how bootstrap the data and I'm not familiar with Bayesian statistics. Could you suggest me a link to better understand this issue or a (free) software to performe these analyses?
Thank you