References 1. Sjaastad O, Saunte C, Salvesen R, Fredriksen TA, Seim A, Roe OD. Short-lasting unilateral headache attacks with conjunctival injection, tearing, sweating and rhinorrhea. Cephalalgia 1989;9:147–156. 2. Goadsby PJ, Lipton RB. A review of paroxysmal hemicranias, SUNCT syndrome and other short-lasting headaches with autonomic feature, including new cases. Brain 1997;120:193–209. 3. Leone M, Rigamonti A, Usai S, D’Amico D, Grazzi L, Bussone G. Two new SUNCT cases responsive to lamotrigine. Cephalalgia 2000;20:845– 847. 4. Graff-Radford NR. SUNCT syndrome responsive to gabapentin (Neuron- tin). Cephalalgia 2000;20:515–517. 5. D’Andrea GD, Granella F. SUNCT-Syndrome: the first case in childhood. Cephalalgia 2001;21:701–702. 6. ter Berg JWM, Goadsby PJ. Significance of atypical presentation of symptomatic SUNCT: a case report. J Neurol Neurosurg Psychiatry 2001;70:244 –246. 7. Penart A, Firth M, Bowen JRC. Short-lasting unilateral neuralgiform headache with conjunctival injection and tearing (SUNCT) following pre- sumed dorsolateral brainstem infarction. Cephalalgia 2001;21:236 –239. Elevation of serum copper levels discriminates Alzheimer’s disease from vascular dementia R. Squitti, PhD; P. Pasqualetti, PhD; E. Cassetta, MD; G. Dal Forno, MD, PhD; S. Cesaretti, PhD; F. Pedace, MD; A. Finazzi-Agrò, MD; and P.M. Rossini, MD Serum copper elevation and total radical trapping antioxidant capacity (TRAP) decrement were reported to parallel cognitive deficits in patients with AD, suggesting that copper measure- ments may help to noninvasively discriminate AD from normalcy. 1 Iron, copper, and zinc homeostasis is perturbed in AD and these metals concentrate in senile plaques, neurofibrillary tangles, and CSF. 2 To assess whether serum copper elevation is specific for AD, we studied copper levels, along with iron, total peroxides, TRAP, transferrin, and APOE genotype in patients with vascular demen- tia (VAD) and patients with AD. Subjects and methods. Forty-eight patients with probable AD (mean age = 75.7, SD = 6.8), National Institute of Neurological and Communicative Disorders and Stroke/Alzheimer’s Disease and Related Disorders Association (NINCDS-ADRDA criteria), 3 without additional pathologic conditions, thoroughly studied, as described elsewhere, 1 were compared with 20 patients with VAD (mean age = 78.8, SD = 7) (NINCDS-Association Internationale pour la Recherche et l’Enseignement en Neurosciences criteria) with a Mini-Mental State Examination (MMSE) score of 25. 4,5 Patients with VAD were selected on the basis of clinical history consistent with VAD as well as brain MRI criteria suggestive of microangiopathic (small vessel) pathology (Jellinger class 2). 6 Ex- clusion criteria were conditions known to affect copper and periph- eral oxidative metabolism. 1 The study was approved by the internal review board. All participants or legal guardians had signed an informed consent. Materials included copper (Randox Laboratories, Crumlin, UK); d-ROMs test (Diacron, Italy) for hydro- and lipo- peroxides assessment; TAS (Randox Laboratories, Crumlin, UK) for TRAP evaluation; FerroZine (Roche Diagnostics, Mannheim, Germany) for iron; and Transferrin (Roche Diagnostics, Mannheim, Ger- many) for transferrin. All biochemical and APOE genotyping as- says were performed according to published methods and described in detail elsewhere. 1 Demographic, genetic, and cognitive characteristics of patients with AD and VAD were compared with a Student t-test (with corrected degrees of freedom, when heteroschedasticity was present), Mann–Whitney U, and  2 square, according to data type and distribution. To select independent variables discriminating AD from VAD, we used binary multiple logistic regression with the forward likelihood-ratio method and provided odds ratios (ORs) and 95% CIs to measure effect sizes. Receiver operator characteristics curves were applied to identify biologic variables, their diagnostic validity, and a useful cut-off value for sensitivity and specificity (SPSS 10.0 for Windows; SPSS, Inc., Chicago, IL). Results. The AD (15 men, 33 women; mean MMSE score = 17.5, SD = 4.5) and VAD (13 men, 7 women; mean MMSE score = 22.6, SD = 2.9) groups differed for sex [ 2 (1) = 6.639; p = 0.010)] and MMSE [t(51.7) = 5.23; p  0.001)], but not for disease dura- tion (median AD duration = 33 months, range = 12 to 120; me- dian VAD duration = 24, range = 8 to 72; Mann–Whitney U = 191.5, p = 0.398). The 4 frequency distribution was not different in patients with AD (17%) and those with VAD [(19%;  2 (1) = 0.553, p = 0.457)]. The table reports copper, peroxides, TRAP, iron, and trans- ferrin levels and related statistics. After adjusting for MMSE scores and sex, copper discriminated AD from VAD (OR = 2.06; 95% CI: 1.28 to 3.31; p = 0.003). An unitary increment in serum copper concentration doubled the probability of belonging to the AD group. Transferrin and iron did not fit model criteria (p  0.05). Inspection of the receiver operator characteristic curves and a specificity constraint of 85% identified a cut-off level of 16 mol/L (1.02 mg/L; data not shown) providing a 63% sensitivity. Perox- ides and TRAP did not distinguish AD from VAD. Because MMSE scores were lower in the AD group, two MMSE-matched sub- groups of 14 patients with AD and 16 with VAD were compared. The difference in copper levels was confirmed [mean = 16.7, SD = 4.9 in AD, mean = 13.7, SD = 2.2 in VAD; t(17.4) = 2.31, p = 0.033]. Discussion. This study shows that, by setting a cut-off of 16 mol/L, serum copper level elevations discriminate AD from VAD, confirming previous findings in which the same cut-off discrimi- nated patients with AD from healthy controls (94.7% specificity, 60% sensitivity). 1 Current results suggest that copper may be a marker for AD. Total peroxides and TRAP disruption were noted in either AD and VAD, suggesting a significant oxidative stress in both dementias in comparison with normalcy. 1 Despite contrasting reports about specific pro- or antioxidant agents, general agree- ment exists on a perturbation of oxidative balance in dementia. We can therefore assume that oxidative abnormalities accompany brain tissue degeneration and that susceptibility to dementia is increased in individuals who have a defective antioxidant machin- ery. Some authors have described a characteristic antioxidant blood profile in patients with AD and VAD that allows a discrimi- nation between these two diseases. 7 Our results confirm this hy- pothesis, and in particular AD and VAD can be discriminated on the basis of copper levels. Copper homeostasis, therefore, seems specifically disrupted in AD, 1,2,7 supporting the hypothesis of a selective copper-mediated toxicity as part of the pathogenic pro- cess of this disease. 2 Measurements of serum copper, oxidative stress, and antioxi- dant variables in other dementias and neurodegenerative disor- ders are needed to validate our preliminary observations. Table Serum copper, peroxides, TRAP, iron, and transferrin concentrations in patients with AD and vascular dementia Patients with AD (n = 48) Patients with VAD (n = 20) t-test Copper* (mol/L) 18.3 (6) 13.2 (2.2) t(65.9) = 5.368 p  0.001 Peroxides (U.CARR † ) 364 (79) 340 (64) t(66) = 1.207 p = 0.232 TRAP (mmol/L) 1.30 (0.16) 1.29 (0.18) t(66) = 0.364 p = 0.717 Iron (g/dL) 72.1 (31.7) 80.4 (36.8) t(66) =-0.831 p = 0.408 Transferrin (g/L) 2.56 (0.60) 2.57 (0.68) t(66) =-0.045 p = 0.964 Values are mean (SD). * Values are referred to copper measured with spectrophotome- try. Atomic absorption measurements gave superimposable re- sults (r = 0.89, p  0.001). See Squitti et al. 1 for details. † U. CARR corresponds to 0.08 mg/100 mL of hydrogen peroxide. June (2 of 2) 2003 NEUROLOGY 60 2013