Advertisement

THE DIAGNOSTIC VALUE OF AMYLOID PET IN AN UNSELECTED COHORT OF MEMORY CLINIC PATIENTS

      Background

      Earlier studies evaluating the diagnostic value of amyloid PET used highly selected study samples. We offered amyloid PET to all patients visiting our memory clinic and studied how amyloid PET impacted clinical diagnosis. In addition, we investigated whether appropriate use criteria (APUC) adequately identified patients that benefitted most from amyloid PET.

      Methods

      From March to December 2015, we offered [18F]florbetaben PET to all patients (n=443) visiting our memory clinic. Of all patients, 170/443 (38%) participated (64±7yrs; 61%M). PET scans were visually assessed as amyloid positive or negative. Before and after disclosure of PET results, one dedicated neurologist determined syndrome diagnosis and suspected etiology for each patient. Retrospectively and blinded to amyloid status, we applied APUC, which were positive when patients [A] had persistent or unexplained mild cognitive impairment (MCI), [B] satisfied core clinical criteria for possible AD with atypical clinical course or etiologically mixed presentation, or [C] that were young (≤65) with dementia.

      Results

      Patients who participated (n=170) did not differ in age and sex compared to non-participants (n=273), but had slightly higher MMSE (25±5 vs. 23±6; p0.02). Of participating patients, 95(56%) were demented, 25(15%) had MCI and 50(29%) were cognitively normal. After disclosure of PET results, syndrome diagnosis changed in 8(5%) patients. In demented patients, suspected etiology changed in 16/68(24%) patients with pre-PET suspected AD etiology, and in 4/27(15%) with suspected non-AD etiology. In non-demented patients, suspected etiology changed in 6/18(33%) with pre-PET suspected AD etiology, and in 15/57(26%) with suspected non-AD etiology. When we applied APUC (Table 1), 59/95 dementia and 18/25 MCI patients met the criteria. By definition, cognitively normals were APUC-. In dementia, suspected etiology changed in 19/59(32%) for APUC+, compared to 1/26(4%) in APUC-. In non-demented patients, suspected etiology changed irrespective of APUC status (APUC+28%; APUC-28%).

      Conclusions

      In an unselected sample of memory clinic patients, amyloid PET has diagnostic value as it contributes to the ideas regarding suspected underlying etiology in a substantial number of patients. APUC are helpful to identify patients with dementia who may benefit from amyloid imaging, but do not adequately identify non-demented patients benefitting from amyloid PET.
      Table 1Appropriate Use Criteria in Demented and Non-Demented
      Dementia (n=95)MCI (n=25)CN (n=50)
      AD (n=68)Non-AD (n=27)
      APUC+ (n=45)APUC- (n=23)APUC+ (n=14)APUC- (n=13)APUC+ (n=18)APUC- (n=7)APUC- (n=50)
      Age (SD)65 (±7)66 (±7)65 (±10)67 (±7)65 (±7)66 (±7)61 (7%)
      Males (%)25 (56%)9 (39%)13 (93%)9 (69%)12 (67%)6 (86%)29 (58%)
      MMSE (SD)22 (±4)21 (±5)24 (±3)23 (±6)27 (±2)27 (±2)28 (±3)
      [18F]FBB+ (%)30 (67%)21 (91%)7 (50%)1 (8%)10 (56%)5 (71%)13 (26%)
      Change in syndrome diagnosis (%)2 (4%)1 (4%)0 (0%)0 (0%)1 (6%)1 (14%)3 (6%)
      Change in suspected etiology (%)15 (33%)1 (4%)4 (29%)0 (0%)5 (28%)2 (29%)14 (28%)