Search

BACKGROUND Perinatal depression (PND) is a severe complication of pregnancy, but there are no established risk factors predicting the disease. Evening chronotype has been associated with unhealthy lifestyle habits and adverse outcomes during pregnancy. In this study, we aimed to clarify whether chronotype can predict symptoms and/or occurrence of PND. METHODS Two hundred ninety-nine women were followed-up from the first trimester of pregnancy until 6 months postpartum. Chronotype was assessed at baseline using the MEQ, while mood was repeatedly assessed by depression rating scales (EPDS, HDRS, MADRS). The influence of time and chronotype on EPDS, HDRS and MADRS, was estimated by constructing multilevel linear mixed regression models. A Cox proportional-hazard regression model was built to evaluate the association between chronotype and incidence of depression. RESULTS Chronotype modulated PND symptom severity depending on time of assessment, with evening chronotypes having a higher risk for developing PND symptoms, as assessed by EPDS, at postpartum visits V4 (5-12 days) and V5 (19-26 days). These also had less healthy lifestyle habits and were more likely to suffer from gestational diabetes mellitus and undergo cesarean delivery as compared to other chronotypes. LIMITATIONS Only a minority of women were classified as evening chronotypes. The long follow-up phase of the study led to missing data. CONCLUSIONS Pregnant evening chronotypes show unhealthy lifestyle habits and sociodemographic characteristics commonly associated with a higher risk for PND. They also have a higher risk of developing PND symptoms in the first month after delivery. Chronotype should therefore be routinely assessed during pregnancy to identify women potentially at risk for developing PND.

Purpose: To evaluate the role of F-18-Fluorothymidine (FLT) PET/CT in lymphoma patients with suspected recurrent or residual disease. Methods: Adult lymphoma patients presenting with positive or equivocal F-18-FDG PET/CT at end-treatment or follow-up were prospectively addressed to an additional F-18-FLT-PET/CT. SUV max and tumour-to-background ratios (TBRs) were recorded for the most avid lesion. Biopsy or, when not available, clinical or imaging assessment were employed as standard of reference. Results: Overall 52 patients were recruited. Histology was available in 20/52 patients (38%), proliferation-index (Ki-67) in 14/20. Disease was excluded in 13/52 patients (25%) (one reactive follicular hyperplasia, five reactive-inflammatory tissues, four reactive nodes, two nodal sarcoid-like and one non-specific peri-caecal finding). FDG and FLT scans were concordant in disease restaging in 34/52 patients (65%), whereas in 18/52 cases (35%) relevant discrepancies were recorded. SUV max and TBR were significantly higher in the disease versus the disease-free group, with both tracers (p = 0.0231 and 0.0219 for FDG; p = 0.0008 and 0.0016 for FLT). FLT-SUVmax demonstrated slightly better performance in discriminating benign from malignant lesions (ROC-AUC: 0.8116 and 0.7949 for FLT-SUV max and TBR; 0.7120 and 0.7140 for FDG). Optimal FLT-SUV max cut-offs were searched: three would lead to 95% sensitivity, 81% accuracy, and 39% specificity, whereas seven led to 100%, 41%, and 56% respectively. No statistically significant correlation was observed between the two FLT indices and Ki-67. Conclusions: According to our results in a clinical setting of recurrent or residual lymphoma, FLT is not significantly superior to FDG and it is unlikely that it will be employed independently. FLT may be restricted to a few specific cases, as complementary to standard FDG imaging, to confirm a diagnosis or to define a better target to biopsy. However, due to FLT suboptimal performance, many findings would remain inconclusive, requiring further diagnostic procedures and reducing the effectiveness of performing an additional FLT scan.

The primary aim of the study was to evaluate the role of [18F]Fluciclovine PET/CT in the characterization of intra-prostatic lesions in high-risk primary PCa patients eligible for radical prostatectomy, in comparison with conventional [11C]Choline PET/CT and validated by prostatectomy pathologic examination. Secondary aims were to determine the performance of PET semi-quantitative parameters (SUVmax, target-to-background ratios [TBRs], using abdominal aorta, bone marrow and liver as backgrounds) for malignant lesion detection (and best cut-off values) and to search predictive factors of malignancy. A six sextants prostate template was created and used by PET readers and pathologists for data comparison and validation. PET visual and semi-quantitative analyses were performed: for instance, patient-based, blinded to histopathology, subsequently lesion-based, un-blinded, according to the pathology reference template. Among 19 patients included (mean age 63 years, 89% high and 11% very-high-risk, mean PSA 9.15 ng/mL), 45 malignant and 31 benign lesions were found and 19 healthy areas were selected (n = 95). For both tracers, the location of the “blinded” prostate SUVmax matched with the lobe of the lesion with the highest pGS in 17/19 cases (89%). There was direct correlation between [18F]Fluciclovine uptake values and pISUP. Overall, lesion-based (n = 95), the performance of PET semiquantitative parameters, with either [18F]Fluciclovine or [11C]Choline, in detecting either malignant/ISUP2-5/ISUP4-5 PCa lesions, was moderate and similar (AUCs ≥ 0.70) but still inadequate (AUCs ≤ 0.81) as a standalone staging procedure. A [18F]Fluciclovine TBR-L3 ≥ 1.5 would depict a clinical significant lesion with a sensitivity and specificity of 85% and 68% respectively, whereas a SUVmax cut-off value of 4 would be able to identify a ISUP 4-5 lesion in all cases (sensitivity 100%), although with low specificity (52%). TBRs (especially with threshold significantly higher than aorta and slightly higher than bone marrow), may be complementary to implement malignancy targeting.

INTRODUCTION AND OBJECTIVES:The primary objective was the evaluation of Gallium 68(68Ga)-prostate-specific membrane antigen (PSMA)-11 positron emission tomography/computed tomography (PET/CT) detec...

Objectives: Visceral leishmaniasis (VL) is the most severe manifestation of the infection caused by the protozoan Leishmania, recently on increase in Italy and Spain. The aim of the study was to describe FDG uptake patterns in VL patients (pts) who underwent 18F-FDG PET/CT. Methods: A retrospective monocentric study of pts who underwent FDG PET/CT between 2008 and 2017 and later diagnosed with VL was performed. Semi-quantitative parameters were calculated in FDG-positive lesions: SUVmax, SUVmax spleen/SUVmax liver ratio (SLR), SUVmax focal/diffuse spleen ratio (FDR). Results: Overall, 23 pts were included. PET/CT was negative in 2 immunocompromised pts, positive in 21/23 (91%) [6 spleen only, 2 spleen + nodes, 7 spleen + bone marrow (BM), 4 spleen + BM + nodes, 1 spleen + BM + lung, 1 BM only + nodes, 2 nodes only]. Splenic involvement was demonstrated in 20/23 (87%) pts. Two different splenic patterns were observed: diffuse (13/20 pts, mean spleen SUVmax = 7.3 ± 4.2 [4.0–14.1], mean SLR = 2.2 ± 1.6 [1.3–6.7]) and focal over diffuse (7/20 pts, mean SUVmax = 12.6 ± 4.5 [9.5–20.5], mean SLR = 2.8 ± 0.8 [2.1–4.4], mean FDR = 2.1 ± 0.8 [1.2–3.6]). Extra-splenic FDG-avid findings were detected in 15/21 pts (65%): bone marrow in 13/15 (mean SUVmax = 4.0 ± 1.3 [2.8–6.0]), nodes in 67/15 and lung in 1/15. Conclusions: PET/CT demonstrated splenic FDG uptake in all immunocompetent VL pts; two splenic patterns (diffuse/focal over diffuse) were observed and indistinguishable from splenic involvement by other disorders. The most frequent extra-splenic FDG-positive sites were BM and lymph nodes. Considering the potential disease aggressiveness and recent outbreaks in north-eastern Italy, VL should be considered in the differential diagnosis of FDG-positive splenic findings in pts from endemic areas or reporting travels to endemic countries.

OBJECTIVES: The primary objective was the evaluation of Gallium 68 (68Ga)-prostate-specific membrane antigen (PSMA)-11 positron emission tomography/computed tomography (PET/CT) detection rate, for identifying the site of prostate cancer (PCa) relapse (local vs systemic), stratifying the population according to different clinical stages of biochemical recurrence (BCR). Secondary aims were: 1) to evaluate the association of clinical/pathologic features and 68Ga-PSMA-11 PET/CT detection rate, 2) to compare 68Ga-PSMA-11 PET/CT with other imaging procedures, and 3) to evaluate the positive predictive value (PPV) in a per-patient analysis. MATERIAL AND METHODS: This population was enrolled through a prospective, open label, single-center trial performed at the Nuclear Medicine of the University Hospital of Bologna (Eudract: 2015-004589-27 OsSC). The inclusion criteria were: (1) proven PCa, (2) surgery or radiotherapy as definitive therapy, (3) proven BCR, (4) prostate-specific antigen (PSA) 0.2-2 ng/ml, (5) age ≥ 35 years, and 6() willing to sign an informed consent. Three-hundred and thirty-two (332) patients were enrolled between March 2016 and June 2017; mean/median PSA was 0.84/0.61 ng/ml, 97.9% (325/332) of patients received radical prostatectomy and 2.1% (7/332) radiotherapy. Different patterns of BCR were identified by referent physicians as follows: (a) persisting detectable PSA after radical prostatectomy in 13.5% (45/332) of patients (subgroup 1), (b) first-time PSA failure after radical therapy in 44.9% (149/332) (subgroup 2), and (c) PSA increase after salvage or hormonal therapy in 41.6% (138/332) (subgroup 3). RESULTS: Primary objective: 68Ga-PSMA-11 PET/CT detection rate was 53.6% (CI 95% 48.1%-59.1%). In a patient-based analysis, disease confined to pelvis (prostate bed and/or lymph-nodes) was detected in 24.7% of cases (82/332). The presence of at least one distant lesion was observed in 28.9% of cases (96/332). The detection rate in different subgroups was: subgroup 1 = 64.5%, subgroup 2 = 45.6%, and subgroup-3 = 58.7%. Secondary objectives: 1) PSA (p = 0.041) and PSAdt (p = 0.001) showed association with 68Ga-PSMA-11 PET/CT detection rate, and 2) correlative imaging was available in 73.2% of patients (243/332). When 68Ga-PSMA-11 PET/CT was positive, correlative imaging resulted negative in 83% of cases (108/130). 3) The calculated PPV was 96.2%. CONCLUSION: Our data confirmed the efficacy of 68Ga-PSMA-11 PET/CT for detecting local vs systemic disease in PCa patients presenting PSA failure after radical therapy. Furthermore, 68Ga-PSMA-11 PET/CT detection rate is different depending on the clinical stage of BCR, and this information should be taken into consideration by referring physicians.

PURPOSE: To compare the accuracy of 18F-FACBC and 11C-choline PET/CT in patients radically treated for prostate cancer presenting with biochemical relapse. METHODS: This prospective study enrolled 100 consecutive patients radically treated for prostate cancer and presenting with rising PSA. Of these 100 patients, 89 were included in the analysis. All had biochemical relapse after radical prostatectomy (at least 3 months previously), had 11C-choline and 18F-FACBC PET/CT performed within 1 week and were off hormonal therapy at the time of the scans. The two tracers were compared directly in terms of overall positivity/negativity on both a per-patient basis and a per-site basis. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy were calculated for both the tracers; follow-up at 1 year (including correlative imaging, PSA trend and pathology when available) was considered as the standard of reference. RESULTS: In 51 patients the results were negative and in 25 patients positive with both the tracers, in eight patients the results were positive with 18F-FACBC but negative with 11C-choline, and in five patients the results were positive with 11C-choline but negative with 18F-FACBC. Overall in 49 patients the results were false-negative (FN), in two true-negative, in 24 true-positive (TP) and in none false-positive (FP) with both tracers. In terms of discordances between the tracers: (1) in one patient, the result was FN with 11C-choline but FP with 18F-FACBC (lymph node), (2) in seven, FN with 11C-choline but TP with 18F-FACBC (lymph node in five, bone in one, local relapse in one), (3) in one, FP with 11C-choline (lymph node) but TP with 18F-FACBC (local relapse), (4) in two, FP with 11C-choline (lymph nodes in one, local relapse in one) but FN with 18F-FACBC, and (5) in three, TP with 11C-choline (lymph nodes in two, bone in one) but FN with 18F-FACBC. With 11C-choline and 18F-FACBC, sensitivities were 32 % and 37 %, specificities 40 % and 67 %, accuracies 32 % and 38 %, PPVs 90 % and 97 %, and NPVs 3 % and 4 %, respectively. Categorizing patients by PSA level (

PURPOSE: We present findings concerning (18)F-fluorodeoxyglucose (FDG) positron emission tomography (PET) at end-treatment evaluation in follicular lymphoma (FL) in order to establish possible predictive factors for progression-free survival (PFS) and patient outcome. METHODS: We retrospectively analysed data from 91 consecutive FL patients (M:F = 51:40, mean age 61) referred to our PET Unit at therapy completion: 38 with an indolent form (grade 1-2) and 53 with an aggressive FL (grade 3a and b) according to the World Health Organization (WHO) classification. A total of 148 FDG PET/CT scans were analysed and findings reported as positive or negative for disease. The overall response to treatment was assessed according to the revised International Workshop Criteria (IWC). The final outcome was defined as remission or disease by taking clinical, instrumental and histological data as standards of reference, with a mean follow-up period of 3 years (range 1-8). A statistical analysis was performed with respect to PFS and patient outcome for FDG PET result, tumour grading, Follicular Lymphoma International Prognostic Index (FLIPI), disease stage and number of relapses, on uni- and multivariate analyses, with p < 0.05 considered as significant. RESULTS: Overall patients presented a mean PFS of 35 months (range 3-86), with a relapse rate of 42%. At final outcome, remission was achieved in 67 of 91 patients (74%). Of the different predictive factors, only FDG PET result significantly correlated with patient outcome (p = 0.0002). PET/CT performance at the end of treatment was as follows: 100% sensitivity, 99% specificity, 89% positive predictive value and 100% negative predictive value. The Kaplan-Meier analysis demonstrated a statistically significant correlation with PFS for FDG PET (p < 0.0001), FLIPI score (0-1 versus ≥ 2) (p = 0.0451) and number of relapses (none versus ≥ 1) (p = 0.0058). These findings were confirmed at the univariate analysis, whereas at the multivariate analysis only FDG PET (p = 0.0006892) and number of relapses (p = 0.01947) were independent predictive factors for PFS. CONCLUSION: End-treatment PET/CT in FL has high accuracy and appears to be a good predictor of PFS and patient outcome, irrespective of grading. As expected, patients facing more than one relapse seem to have significantly shorter PFS in the presence of a positive FDG PET.

Background: The aim of this study is the 2-[fluorine-18]fluoro-2-deoxy-D-glucose (FDG)–positron emission tomography (PET) evaluation following radioimmunotherapy (RIT) with ibritumomab tiuxetan Y 90 in patients with non-Hodgkin’s follicular lymphoma (FL). Materials and methods: We retrospectively analyzed data from 59 relapsed or refractory FL patients treated with ibritumomab tiuxetan Y 90 in four different PET centers who had a PET scan carried out before and after RIT. Possible predictive factors of progression-free survival (PFS) were studied through univariate and multivariate analysis. Results: The post-RIT PET documented 45.8% complete responders (CR), 25.4% partial responders (PR) and 28.8% nonresponders [stable disease + progressive disease], with an overall survival of 71.2% (range 59.5%–90.9%). With a median follow-up period of 23 months, the univariate analysis documented a statistically significant relation between disease extent before RIT and response to treatment with respect to PFS (P = 0.015), while all the other prognostic factors showed no significant correlation. When carrying out the multivariate analysis, post-RIT PET resulted as the lonely independent predictor of PFS (P < 0.00001). Conclusions: RIT is an effective therapy in FL patients, as confirmed in our study too. Disease extension before treatment and response to RIT, as assessed by FDG–PET, result as main predictors of PFS, with the post-RIT PET result being the only independent predictive factor.

A group of 20 patients with probable Alzheimer disease (AD) and a control group were tested in a verb generation task, in a verb synonym task and several cognitive tests. Three types of verbs and novel verbs were presented in simple sentence frames, in two different conditions. In one condition participants were presented with the verb in the infinitive, providing information about the conjugation of the verb and the most likely type of past participle mapping required. The second was an ambiguous condition in which the suffix of the input verb did not provide any clue to the conjugation. The aim was to investigate if different types of verbs and input mapping affected patients' performance, and to what extent the deficit increased when the illness became more severe. Dependent measures were accuracy rates, rates of different morpho-phonological transformations and error type rates. Patients showed a more marked deficit in verb generation, when input was ambiguous. Correlation between the verb synonym test and accuracy in verb generation indicated that a deficit in lexical-semantic memory was partially responsible for impaired performance. Data suggest that patients maintained information about frequency distribution of different verb types and verb classes in each conjugation, but were impaired in operating complex phonological transformations.

Introduction. A monocentric prospective study was designed to evaluate the role of F-18-fluoro-3-deoxy-3-L-fluorothymidine (FLT) PET/CT in lymphoma patients presenting with positive or equivocal F-18-deoxyglucose (FDG) PET/CT at end-treatment or follow-up evaluation. Methods. From May 2010 to March 2015, 40 patients were enrolled in the study, all undergoing FLT PET/CT within 3 weeks from a previous positive FDG scan and, when possible, biopsy confirmation as standard of reference to define PET results (true positive-TP; true negative-TN; false positive-FP; false negative-FN). The highest lesion standardized uptake value (SUVmax) was measured with both tracers. Median age was 55 (range 20-76); 24 patients were male and 16 female. Thirty patients had non-Hodgkin lymphoma and 10 Hodgkin lymphoma (HL). At diagnosis, 8 patients had stage II, 8 stage III and 24 stage IV. Six patients were evaluated at the end of their treatment and 34 during follow-up. Results. FDG-PET was judged positive in 36 out of 40 cases (SUVmax range 3-31; mean 10.1); 28/36 resulted also FLT positive (SUVmax range 3-16.8; mean 7.8) and 16/28 patients had the most active lesion biopsied, demonstrating 15 TP but 1 FP (cervical node, patient with a history of HL, SUVmax 6.9 and 6.8 for FDG and FLT respectively: reactive follicular hyperplasia). The remaining 12 patients could not be biopsied: however, clinical judgment and subsequent follow-up data confirmed 11 TP cases and 1 FP (reactive node). Eight FLT scans resulted inconclusive because of the low SUVmax values (range: 1.7-11.7; mean 4.8); 5 were judged as TP at biopsy (2 patients) or clinical evaluation (3 patients), whereas 3 represented reactive-inflammatory tissue. Among the 4 FDG-PET considered inconclusive (SUVmax range 4.9-8.8; mean 7.2), 3 out of 4 also resulted FLT inconclusive (SUVmax range 2.6-4.9; mean 4), but the final clinical evaluation excluded lymphoma (1 reactive cervical node, 1 nodal sarcoid-like disease, 1 aspecific pericecal finding). The remaining 1/4 resulted FLT positive (SUVmax 6.9 and 6.5 for FDG and FLT, respectively) and was clinically confirmed to be a TP case. Conclusions. In the particular setting of residual/recurrent lymphoma, our preliminary data suggest that FLT can be complementary to FDG, although the elimination of inconclusive PET findings remains a controversial step. Further lesion-based, semi-quantitative analyses (i.e. target to background/mediastinal blood pool/liver ratio) and correlation with Ki67 proliferation index are ongoing. Disclosures Cavo: Janssen: Honoraria; Celgene: Honoraria, Speakers Bureau; Amgen: Honoraria; Bristol Myers Squibb: Honoraria; Novartis: Honoraria. Zinzani:Pfizer: Membership on an entity's Board of Directors or advisory committees; Gilead: Membership on an entity's Board of Directors or advisory committees; Takeda: Membership on an entity's Board of Directors or advisory committees; J&J: Membership on an entity's Board of Directors or advisory committees; Celgene: Membership on an entity's Board of Directors or advisory committees.

Objectives: Antegrade selective cerebral perfusion (ASCP) represents the best method of cerebral protection during surgery of the thoracic aorta. However, brain integrityand metabolism after antegrade cerebralperfusion have not yetbeen investigated. We assessed cerebralpositron emission tomography (PET), diffusion-weighted imaging, proton magnetic resonance spectroscopy and cognitive functions in patients undergoing either ASCP or coronary artery bypass grafting (CABG) to elucidate whether cerebral perfusion was associated with postoperative neuronal alterations, metabolic deficit or cognitive decline. Methods: Seventeen patients undergoing aortic arch surgery using ASCP with moderate hypothermia (26 8C) (ASCP group) and 15 patients undergoing elective on-pump CABG (CABG group) were prospectively enrolled in the study. Brain PET, diffusion-weighted imaging, proton magnetic resonance spectroscopy and neuropsychometric testing were performed preoperatively, and at 1 week and 6 months postoperatively (T1, T2 and T3, respectively). Patient data were compared for statistic analysis with a normal database made up of healthy subjects. Results: One patient in each group was excluded because they refused postoperative evaluation. There were neither strokes nor hospital deaths. Two patients suffered from temporary neurological dysfunction (one in each group). Proton magnetic resonance spectroscopy did not reveal significant alterations in cortical N-acetyl-aspartate (NAA) content within and between the groups at T2 and T3. In the ASCP group, brain diffusion-weighted magnetic resonance showed a significant increase of the apparent diffusion coefficient values, reflecting vasogenic cerebral oedema, at T2, that disappeared at T3. Magnetic resonance detected new focal brain lesions in two CABG group patients. In seven ASCP group patients, PETscan showed glucose hypometabolism in the occipital lobes at T2, which disappeared in five patients at successive examination (T3). Significant cognitive decline was not observed in any patient. Test score changes between and within groups were not significant. Conclusions: There was no evidence of ischaemic brain injury after ASCP even if some degree of reversible brain oedema secondary to cardiopulmonary bypass (CPB) was present. The cognitive outcomes in patients undergoing ASCP were comparable to patients undergoing coronary artery bypass. The lack of left subclavian artery perfusion during cerebral perfusion leads to temporary glucose hypometabolism in the occipital lobes without neuronal injury. # 2009 European Association for Cardio-Thoracic Surgery. Published by Elsevier B.V. All rights reserved.

We report the clinical findings and neuropsychological profiles of a sample of patients exhibiting a focal retrograde amnesia (FRA) seen consecutively during the period 1992–2007. The cohort comprised 13 patients, five males, with a mean age of 30 years (range 16–49). They were given a neurologic examination, psychiatric interview and formal neuropsychological examination (all but one) during the amnesic phase, underwent neuroimaging, and were followed up for six months to ten years. All presented with an acute amnesia characterized by an impaired recollection of memories predating the acute event, with spared or minimally and transiently affected anterograde memory, thus consistent with FRA. The events triggering FRA varied widely: mild to severe head injury, road accident without head injury, seizure, dissociative fugue, BDZ overdose, posttraumatic headache, syncope, migraine attack, acute distress. The neuropsychological hallmark of FRA was a selective or prominent impairment of autobiographical memory. The defect was often so severe as to cover most or all of the patients' lives and, in some cases, to erase the knowledge of their own identity. Conventional neuroimaging (brain CT and MRI) was unimpressive. Cerebral SPECT/PET disclosed unilateral frontal hypoperfusion in three (two left). All but one patient fully recovered, time of recovery ranging from three days to six months. FRA is a condition reflecting a block of memory function triggered by heterogeneous events, including both physical and psychic insults. Analogies shared with the more frequently encountered and better known condition of transient global amnesia suggests common pathogenetic mechanisms. A tentative nosographic classification of FRA is finally offered.

We describe three patients exhibiting an acute reversible amnesia characterised by an impaired recollection of past events with preserved anterograde memory, thus consistent with a focal retrograde amnesia (FRA). This occurred after variable events: state of fugue, road accident, post-traumatic headache. Retrograde amnesia affected autobiographical memory so severely as to cover all of the patients' lives and to erase knowledge of their own identity. The retrieval of public events was variably affected, ranging from normality to severe impairment. No lesions were found on neuroimaging, and neurophysiological findings were unimpressive. FRA subsided in a few days, leaving a gap for the onset. The hypothesis of a psychogenic amnesia is considered, but overcoming the organic/psychogenic dichotomy the episodes appear as examples of "functional" memory inhibition, potentially triggered by different conditions, including events classifiable as psychic trauma. The clinical and neuropsychological traits of functional FRA are discussed. According to a current theory of autobiographical memory, the memory profile may be explained by a lost access to abstract autobiographical knowledge. Given some analogies with the more common transient global amnesia, a mechanism of spreading depression may also be hypothesised for functional FRA.

The influence of lexical-semantic impairment and of executive dysfunction on word naming performance was investigated in a group of patients with probable Alzheimer dementia (AD). The patients, who varied in the severity of the illness, were tested in a word naming task where they had to read aloud Italian three-syllable words with a dominant or subordinate stress pattern. These types of words have been shown to interact with frequency in normal adults [J. Exp. Psychol.: Hum. Percept. Perform. 18 (4) (1992) 987], so that the effect of the subordinate stress pattern (slower reading times) is only apparent for low frequency words. The frequency and stress effects on accuracy increased across dementia severity levels. Regression analyses showed that the impairment in reading low frequency words with subordinate stress depended largely on the level of lexical-semantic impairment, measured by a test of semantic memory and comprehension. Implications for the current reading models are discussed.