THE PROGNOSTIC VALUE OF LESION CHARACTERISTICS FOR POST-STROKE UPPER EXTREMITY REHABILITATION IN THE ACUTE SETTING: A POST-HOC ANALYSIS.

Abstract

PURPOSE/HYPOTHESIS: An important objective in today's cost-contained health care environment is the development of valid and reliable clinical measures that can be used to select the most appropriate rehabilitation candidates and predict functional outcomes to rehabilitation protocols. The purposes of this post-hoc analysis were: 1) to determine the relation between lesion characteristics (size and location) and the Orpington Prognostic Scale (OPS); 2) to determine the prognostic value of lesion characteristics on functional outcomes with focused post-stroke upper extremity treatment. A data subset was obtained from a randomized controlled comparison of UE rehabilitation strategies in acute stroke (Phase II STAR, Stroke Arm Recovery trial; Winstein et al., 2004). For STAR, we showed that OPS was useful for stratifying eligible candidates across two intervention and a usual care group. Those patients with more severe OPS showed no differential benefit from 20 additional hours of focused UE rehabilitation delivered over a 4-6 week period compared to usual care. By contrast, those with less severe OPS showed a significantly better functional outcome from the focused UE rehabilitation. NUMBER OF SUBJECTS: The acute magnetic resonance image (MRI) from a subset of STAR patients (11 less severe, 5 more severe) were analyzed. MATERIALS/METHODS: We used a computer-based analysis program (MRIcro) to estimate the size and location for each lesion from the T1 images. Baseline OPS, immediate and post-intervention outcome data from the Functional Test of the Hemiparetic Upper Extremity (FTHUE) were used for our analyses. Functional outcomes were classified as poor, moderate, or good based on items accomplished with the FTHUE. RESULTS: For 16 participants, the correlation between lesion size alone and OPS was moderate, r = 0.64 (p = .004), however, this correlation improved substantially (r = 0.84, p =.0001) for n -1 subgroup. One subject, with a relatively small lesion (7.1 cc) and severe OPS (6.0), showed 73% of the lesion located in the hand area of the primary motor cortex (MI); in this case, lesion location more importantly determined OPS than did lesion size. Surprisingly, there was a significant correlation between lesion size alone and functional outcome (r =.-54, p = .017) using the FTHUE post-test raw score. When categorical outcome (FTHUE) was examined in relation to lesion size and location [pontine (n=4), subcortical (n=7), cortical + partial M1 (n=4), MI only (n=1)], the only participants with good outcomes were those with small (< 20cc) lesions in either the subcortical or pontine location. Further, within the subcortical lesion group, better outcomes were achieved for those who received focused UE treatment. CONCLUSIONS: Lesion size and location are important prognostic factors for determining UE recovery potential after stroke. CLINICAL RELEVANCE: While limited by a small sample, our multi-dimensional approach provides complementary information to the recent clinical practice guidelines for UE rehabilitation (Barreca et al., 2003).