Limitations of invasive modalities in the diagnosis of primary angiitis of the central nervous system.

OBJECTIVE

Rheumatologists are often consulted to evaluate patients suspected of having primary angiitis of the central nervous system (PACNS). The diagnostic process relies heavily on interpreting the results of cerebral angiography and brain biopsy. We have assessed the operating characteristics of those invasive modalities in the diagnosis of PACNS:

METHODS

The records of 30 consecutive patients referred for the evaluation of possible PACNS were retrospectively analyzed. Patients were evaluated on clinical grounds, and the diagnostic process was extended accordingly until a reasonable probability of a definable disease was present. All patients had cerebral angiography and/or brain biopsy. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were assessed for each of the following diagnostic tests: cerebral angiography, brain biopsy, cerebrospinal fluid (CSF) examination, and magnetic resonance imaging (MRI).

RESULTS

The final diagnostic outcomes were: PACNS in 7, lymphoproliferative disease in 4, infection in 4, demyelinating disease in 2, reversible vasospastic disorder in 6, and a variety of other nonvasculitic conditions in 7. Cerebral angiography had less than 30% specificity and PPV for PACNS: Brain biopsy had limited sensitivity and NPV (53% and 70%, respectively). CSF examination and MRI, although sensitive, lacked specificity.

CONCLUSION

In patients suspected of having PACNS, the results of invasive diagnostic modalities should be interpreted with caution. Accurate diagnosis should rarely rely on any single study and should only follow careful clinical, radiographic and pathologic correlation.