RESUMO
PURPOSE: We performed a retrospective analysis of a sarcoidosis cohort who had sACE obtained at their initial clinic visit, but the treating physician was blinded to the results. We examined the relationship between sACE and the treating physician's decision to escalate sarcoidosis treatment. METHODS: Treatment was considered escalated if the prednisone dose was increased or if the prednisone dose was not changed but an additional anti-sarcoidosis drug was added or the dose was increased. RESULTS: 561 sarcoidosis patients were analyzed. The most common target organ was the lung (84%). Using a cut-off of > 82 units/L for an elevated sACE, 31/82 (38%) with an elevated sACE had treatment escalation whereas 91/497 (18%) had treatment escalation with a normal sACE (p < 0.0001). For the need of treatment escalation, a sACE (cut-off of > 82) had sensitivity 0.25, specificity 0.89, positive predictive value 0.38, negative predictive value 0.81. These results were not appreciably different using other sACE cut-off values such as 70, 80, 90, or 100. A multivariable logistic regression model that included demographics, the target organ, spirometry results estimated that sACE level and lower FVC were significantly associated with the likelihood of treatment escalation. These findings held when sACE > 82 replaced sACE level in the multivariable logistic regression model. CONCLUSIONS: Although there was a strong correlation between sACE at the initial sarcoidosis clinic visit and subsequent treatment escalation of sarcoidosis, the predictive power was such that sACE is not adequately reliable to be used in isolation to make this determination.