Serum S-100 protein in acute stroke.

To the Editor: We read with interest the work of Missler et al1 and Büttner and colleagues2 and agree with both sets of researchers on the need for reliable, noninvasive markers of neuronal damage following acute stroke. Such markers may allow prognostication of future clinical outcome and may be useful in acute therapy trials as surrogate markers of efficacy. However, neither study cited our publication on serum S-100 protein levels in acute stroke,3,4 a study that examined 81 patients (68 with ischemic stroke and 13 with hemorrhagic stroke) within 48 hours of stroke ictus and compared them with 51 age-, race-, and gendermatched control subjects.3 As with both recent papers in Stroke, we found significantly higher serum S-100 protein levels in the stroke population than in the control group. Furthermore, the highest S-100 protein levels were seen in the hemorrhagic stroke group, and differentiation between the two stroke populations almost reached statistical significance. We also performed a temporal study (24, 48, 72, and 96 hours after ictus) in 13 patients and, unlike the recent Stroke studies, did not find differences between these time points. Convalescent samples were also analyzed in 57 of the 74 patients still alive at 3 months after stroke; S-100 protein levels had significantly fallen (2 P,.0001) at this time but were still above the level seen in the control population, suggesting partial restoration of glial cell integrity. We have previously shown5 that neuron-specific enolase level and carnosinase activity are individually associated with clinical outcome, determined with use of the Rankin Scale score and Barthel Index, although their ratio has the strongest association. We also found that S-100 levels were associated with clinical outcome.3 The results of our S-100, neuron-specific enolase, and carnosinase activity work and the recent S-100 protein studies show that serum biochemical markers can be readily measured using commercially available assays and appear to provide useful prognostic information. We believe the combination of the results of two (or more) markers, as used in our neuron-specific enolase/human serum carnosinase article,5 may improve their usefulness and reliability. We are currently assessing whether early drug therapy can modulate these serum markers in patients with acute stroke, thereby supporting their use as surrogate markers of efficacy.