Mild Cognitive Impairment in newly diagnosed Parkinson's disease: A longitudinal prospective study

Highlights

• The prevalence of Parkinson's Disease-Mild Cognitive Impairment (PD-MCI) was 32.9% at baseline.

• Executive dysfunctions predicted development of PD-MCI after few years from onset.

• Reversal from PD-MCI to PD-Cognitive Normal underlined the heterogeneity in cognitive deficits.

Abstract

Introduction

In PD, Mild Cognitive Impairment (PD-MCI) occurs since early stages of disease. The aims were to assess presence of PD-MCI in untreated, drug-naive PD patients, and to follow-up the sample over 4 years to ascertain evolution of neurocognitive profile.

Methods

Seventy-six patients underwent neuropsychological testing at baseline (T0), and after 2 (T1:n = 62) and 4 years (T2:n = 55). Diagnosis of PD-MCI and PD-associated dementia (PDD) was made according to current consensus criteria.

Results

PD-MCI occurred in 25/76 patients (32.9%) at baseline, and 4 of them reverted from PD-MCI to Normal Cognition (Reverters), 7 remained stable (Non-Reverters) and 2 developed PDD at T2; 12 patients were lost to the follow-up. Among the 51 patients with normal cognition (PD-CN) at T0, 27 had normal cognition at T2 (5 of them were Reverters with respect to diagnosis at T1), 5 had MCI at T1 and T2 (Non-Reverters), 9 had MCI at T2 only, whereas 1 developed PDD; 9 patients were lost to the follow-up.

At baseline, Reverters (n = 9) had younger age at onset and better performance on constructional visuospatial task than Non-Reverters (n = 12). Compared to patients without PD-MCI at all evaluations (n = 19), Reverters had poorer performance on verbal immediate recall and attention tasks and higher level of apathy at T0. Reduced performance on the Stroop Test at baseline predicted PD-MCI at T2.

Conclusion

Executive dysfunctions predicted development of PD-MCI after few years from onset. Reversal from PD-MCI to PD-CN was related to young age at onset and high level of apathy at baseline evaluation.

Introduction

Mild Cognitive Impairment (MCI) is often thought to be an intermediate condition between normality and dementia [1]. In Parkinson's Disease, MCI (PD-MCI) occurs since early stages of disease and is mainly characterized by impairments of executive functioning and attention [2], [3]. PD-MCI becomes more frequent with ageing and duration of PD [4], and is a risk factor for development of PD-associated dementia (PDD) [5]. The prevalence rate of PD-MCI ranges 18.9–38.2% across studies [4], depending on classification criteria, PD sample (treated or untreated patients), and neuropsychological tests used to assess cognitive functioning.

Recent operative criteria for diagnosis of PD-MCI allow to diagnose PD-MCI on the basis of “comprehensive” (level II) or “abbreviated” (level I) cognitive assessment, with the latter being more suitable for clinical practice and large-scale studies [6]. The “comprehensive” (level II) cognitive evaluation should assess all five relevant cognitive domains and should include two tests within each of the five cognitive domains [7]. The “comprehensive” (level II) cognitive evaluation has been employed in some cross-sectional studies aimed at identifying PD-MCI in untreated and treated patients with PD [8], [9], [10].

The MDS PD-MCI criteria were proposed to promote a uniform definition of PD-MCI across different research and clinical settings, although some issues still warrant discussion or evaluation: the feasibility of level II versus level I criteria (i.e., time-consuming level II criteria might not be easily applied in a clinical setting); the optimal cut-off scores for defining PD-MCI, since MDS proposed a range of cut-off scores (1–2 Standard Deviations below normative data), but recent studies observed good sensitivity and specificity at a cut-off score of 2 SD [10] (see Geurtsen et al. [11], for a large multicentric ongoing study employing such criteria). Similarly, the MDS Task Force proposed diagnostic criteria for PDD, based on a short checklist (Level I) or on a protocol of neuropsychological tests (Level II). These criteria have been validated in clinical practice, but still present issues similar to those delineated above for PD-MCI criteria.

A few longitudinal studies are available on PD-MCI. Only one longitudinal study adopted level II cognitive assessment in a sample of treated PD patients, and reported a 35% prevalence rate of PD-MCI at the time of diagnosis, with an increase to approximately 50% after 3 and 5 years [12]. Three longitudinal studies adopted level I cognitive assessment [5], [13], [14], but only two of them followed the current criteria for PD-MCI. Pedersen et al. [13] reported that MCI within the first year of PD diagnosis signals a highly increased risk for early incident dementia, and underlined both prognostic value and validity of the MCI concept in patients with early PD. More recently, Hobson and Meara [14] reanalysed data on a PD cohort retrospectively and prospectively (4 years, 6 years and 16 years after baseline assessment), and found that incidence of progression from PD-MCI to PDD was 98.0 per 1000 person-year, with an annual conversion rate to PDD of 11%; impaired semantic fluency, praxis and visuospatial functions were neuropsychological predictors for conversion from PD-MCI to PDD.

Although PD-MCI patients are at increased risk of dementia compared with cognitively intact patients, recent longitudinal studies showed that some PD-MCI patients reverted to Normal Cognition (PD-CN). Measurement errors and practice-related learning effects, depression, mild psychiatric conditions or stress, general ill-health or poor motivation are among the possible factors associated with reversal from MCI to NC [13], [15], [16].

Taking into account the abovementioned studies, some issues seem to be explored yet: 1) progression of PD-MCI identified on the basis of “comprehensive” (level II) cognitive assessment after controlling for depression; 2) identification of baseline cognitive measures, not modulated by antiparkinsonian treatment [17], useful to identify PD patients at high risk for PD-MCI. We therefore designed a longitudinal prospective study on drug-naïve, non-depressed de novo PD patients in which we evaluated: occurrence of PD-MCI at baseline and over the following 4 years, and neuropsychological predictors for development of PD-MCI. For these purposes, we administered a “comprehensive” Level II cognitive assessment at baseline (T0), and two (T1) and four years (T2) after baseline assessment to all enrolled patients, and adopted recent clinical criteria for diagnosis of PD-MCI [6] and PDD [18].

Since the PD-MCI likely represents a transitional period between normal ageing and the diagnosis of clinically probable PDD, the accurate and early identification of patients with PD-MCI can shed light on course and treatment of this condition and even on possible strategies to delay or prevent development of PDD [19], [20]. Within this perspective, longitudinal studies are most needed to identify clinical or cognitive predictors for development of PD-MCI. The present study aims to contributing to such issues, by assessing the natural course of this clinical condition, and setting the reference frame useful to ascertain possible efficacy of disease-modifying therapeutic options.