This review, available in free full text from Mayo Clinic Proceedings, is the best I've seen on this topic. Points I found interesting:
The traditional classification of SS into diffuse and limited (aka scleroderma vs CREST) is simplistic. Finer gradations based on the extent of skin involvement or rate of progression of skin disease may be more meaningful. Also, specific disease profiles based on auto antibody associations are emerging. Antibody-phenotype associations are shown in one of the tables in the article. Although the associations are not absolute (and there is a certain amount of overlap) the specific antibodies are fairly predictive such that antibody profiling early in the course of the disease is beneficial.
There are four pathophysiologic dimensions to SS: vasospasm, obliterative vasculopathy, tissue fibrosis and autoimmunity.
There is no “treatment for SS” per se. Treatment is directed at disease manifestations. Immunosuppression directed against autoimmune processes that drive certain clinical manifestations is recommended in selected circumstances but none of it is supported by high level data.
Corticosteroids, in disfavor because they increase the risk of renal crisis, should be used sparingly if at all.