A working classification of scleroderma spectrum disorders: a proposal and the results of testing on a sample of...

To propose a working classification for scleroderma spectrum disorders (SDS), based on criteria shown in the past to be specific for the condition: sclerodermatous skin involvement, specific autoantibodies and specific microvascular abnormalities, to obtain mutually exclusive subgroups of SDS and to test this classification on a sample of patients. This is an attempt to improve the current classifications used in research studies, which are all based on the extent of skin involvement only.

Patients (n= 165) referred to the Microvascular Laboratory of the Division of Rheumatology between 1976 and 1992, who had participated in various other studies in the Division and for whom therefore systematically collected clinical and laboratory data were available. Evaluation of skin involvement, clinical, laboratory, and microvascular studies were performed independently and a subgroup assignment was made "blindly" from coded data.

Results demonstrated that significant differences were present between our subgroups of SDS with regard to meaningful variables. Comparisons of our results with previously used classifications such as the ACR preliminary criteria study and later subsets used by more recent investigators, demonstrated that our classification permitted us to include less severe cases than those meeting ACR criteria.

Our results show that by dividing SDS into groups using the cutaneous, microvascular and autoantibody characteristics, while using well-defined criteria and inclusion-exclusion rules to obtain mutually exclusive subgroups, can reveal differences that may otherwise be masked. This study has suggested new avenues of research. A replication study should be performed with a systematic follow-up. This is a working classification, i.e. it can be modified as new findings using updated technology become available. The focus should be on clearly defining criteria for classification and collected data. Then between-center comparisons will be possible.