Frontal Network Syndrome Testing

Abstract

Background:

Dementia diagnosis and the various subtypes are challenging in the absence of biomarkers.

Aim:

To examine available tests and neuroimaging procedures that may help distinguish these disorders.

Methods:

Alzheimer’s disease (AD), cognitive vascular disorder (CVD), and Frontotemporal lobe disorders (FTLD) were tested with a hierarchical neuropsychological battery that included the Frontal Systems Behavior Scale, Mini-Mental State Examination, Montreal Cognitive Assessment Test, and subtests. All patients had multimodality magnetic resonance imaging and ¹⁸F fluorodeoxyglucose-positron emission tomography (FDG-PET) brain scans.

Results:

Of the 161 patients evaluated for dysmemory and cognitive impairment, 31 satisfied the full protocol. The mean T scores for the 3 principal frontal system syndromes for the AD group were all abnormal save disinhibition. For the CVD and FTLD groups, all the 4 subcategory scores were abnormal. Disinhibition differed significantly between the AD and FTD group (analysis of variance [ANOVA], P = .02) and there was a strong association between the memory for 5 words test and a significant difference in the word list generation test score among the 3 groups (ANOVA, P = .0233). There was a strong association between the FDG-PET and the disease subtype (P < .0001).

Conclusion:

Evaluation for disinhibition, word list generation, 5-word memory testing and PET brain imaging may help distinguish the 3 most common dementia subtypes.

Background

Dementia diagnosis and the various subtypes are challenging in the absence of biomarkers. Four different frontotemporal lobe degeneration (FTLD) subtypes are recognized and at least 4 different clinical Alzheimer’s disease (AD) subtypes, with both entities having a frontal variant with predominant behavioral presentation. ¹ In addition, the neuropathology is becoming more complex with a steadily increasing stream of new discoveries. However, treatment options and clinical trials depend on accurate diagnosis. ² For example, anticholinergic therapy is of proven benefit in AD, serotonergic therapy has moderate scientific support in the treatment of FTLD, ³ and cognitive vascular disorder(CVD) may benefit from dopaminergic, cholinergic as well as serotonergic therapies. ⁴ Importantly, all therapies are reliant on accurate diagnosis and incorrect treatment may lead to worsening. ⁵ It is a commonplace experience in clinical practice that cognitive evaluation is challenging in the various stages of dementia. In some, the degree of cooperation or attention is limited to no more than a few minutes at best. Research based and emanating from cognitive stroke registries, for example, revealed that higher function abnormalities, including frontal network syndromes (FNSs) were common in acute and sub-acute stroke. ^{6 , 7} In addition, FNS may manifest no matter where the brain lesion, whether frontally located, subcortically, posteriorly, or even subtentorially. An analysis of subtentorial stroke, found FNS in at least half of patients with this location of stroke. Many of these patients were unable to have comprehensive testing in these settings. ⁸

Hypothesis

Frontal systems testing and neuroimaging do not distinguish the most important dementia disorders.

Aim 1

To determine whether examination with frontal systems tests and neuroimaging the most common dementia disorders may be distinguished.

Aim 2

Evaluate the utility of a context-appropriate FNS test battery incorporating behavioral neurological, neuropsychiatric, and neuropsychological components and compared to magnetic resonance imaging (MRI) brain (structural) and metabolic PET brain scanning (functional) to facilitate the diagnosis of the 3 most common dementia syndromes: AD, CVD, and FTLD.

Methods

Setting

Consecutive cognitive and memory impairment patients, aged 18 to 90 years, were accrued through a prospectively coded, dedicated cognitive and memory disorders registry in a tertiary referral center. The Stroke registry was approved by the University Institutional Review Board and in compliance with Health Insurance Portability and Accountability Act (HIPAA) regulations. All patients signed informed consent for the evaluation and the collection of the their neurological, medical, and neurocognitive data. Analysis of the dementia subtypes was performed retrospectively.

Diagnosis of Dementias

The Diagnostic and Statistical Manual of Mental Disorders fourth edition criteria were used for AD and CVD diagnosis. ⁹ For the Frontotemporal lobe disorders (FTLDs), the core diagnostic criteria by Neary et al were used. ¹⁰ In brief, these included insidious onset and gradual progression, early decline in social interpersonal conduct, early impairment in regulation of personal conduct, early emotional blunting, and early loss of insight.

Neuropsychological Testing

A hierarchical and time-orientated clinical approach was adopted. A brief, intermediate, and comprehensive system of frontal tests for clinical application was devised. Assessment was tiered according to the clinical need and indications into 3 different time options, up to 5 minutes, 15 to 30 minutes, and several hour assessment protocols. The tests conform to the 4 core components of FNS, namely working memory, disinhibition, initiation, and monitoring in addition to neuropsychiatric syndrome diagnosis.

Hierarchical Clinical Assessment

1. A 5-minute FNS battery geared toward emergent assessment in the emergency room or primary care outpatient clinics using the Montreal Cognitive Assessment (MOCA).

2. A 15- to 30-minute battery that incorporates behavioral neurological and neuropsychiatric syndromes with abbreviated neuropsychological tests geared toward inpatients, neurology, and psychiatry outpatient clinics including the Frontal Systems Behavior Scale (FRSBE), ¹¹ Mini-Mental State Examination (MMSE), ¹² MOCA, ¹³ orientation for 5 items, serial 7s × 5, memory for 5 words at 5 minutes, word list generation test (WLT) using the letter “F” ¹⁴ and Luria Motor Sequence test. ¹⁵

3. A longer version, typical duration of several hours that incorporates contemporary frontal behavioral neurological, neuropsychological, speech and language, and neuropsychiatric tests need for precise determination of nature and extent of cognitive deficit typically needed for research protocols, forensic situations, or covert brain lesions. These tests include the computerized Wisconsin Card Sorting Test, ¹⁶ the Tower of London Test, ¹⁷ Behavioral Rating Inventory for Executive Function (BRIEF), ¹⁸ FRSBE, ¹¹ Emotional Intelligence Quotient (Bar-On), ¹⁹ computerized Iowa Gambling Test, ²⁰ Stroop Test, ²¹ Comprehensive Trail Making Test, ²² and letter/category fluency tests. ²³

Neuropsychiatric and Behavioral Neurological

The FRSBE ¹¹ is a self-administered and caregiver administered test, yielding scores from both before the onset of illness and at the time of illness. The Likert-type scale questions are converted to age, gender, and education normative data in T scores where scores more than 60 are abnormal.

Cognitive and Neuropsychological

The MMSE and components of the MOCA, orientation for 5 items, serial 7s × 5, memory for 5 words at 5 minutes, WLT using the letter “F,” and Luria Motor Sequence test.

Neuroradiology

All patients had multimodality MRI, MRI T1 and T2, fluid-attenuation inversion recovery, diffusion-weighted imaging, and magnetic resonance angiography to exclude secondary dementia causes such as brain tumor, stroke, multiple sclerosis as well as assessing for leukoaraoisis and degree of generalized and focal atrophy. Positron emission tomography (PET) ¹⁸ F Fluorodeoxyglucose (FDG-PET) brain scans were performed if any uncertainty existed with respect to dementia subtype in accordance with the Food and Drug Administration regulations. In addition, they provided an indication of cognitive reserve pertaining to a particular individual.

Following intravenous injection of FDG, with a dose of 15 milli Curies, the patient was kept in a quiet, darkened room for 60 minutes during the uptake phase. Standard acquisition time was 15 minutes. A single bed PET and co-acquired, low dose computed tomography scan of the same areas were performed. Attenuation-corrected PET images of the brain were created in sagittal, coronal, and transverse projections and reviewed on a computer workstation. Using GE cortex ID software (General Electric Company Corporate Office & Headquarters, Wildwood Pkwy, Atlanta, GA, USA), with comparison to age-matched normal, z scores of regional hypometabolism were obtained in 10 regions of interest and a z score of 2.0 or greater regarded as statistically significant. A General Electric Brilliance LS camera (General Electric Company Corporate Office & Headquarters, Wildwood Pkwy, Atlanta, GA, USA) was used.

Results

Of the 161 patients presenting with dysmemory and cognitive impairment, 31 of 161 (130 of 161; 80.7% excluded) were evaluated according to the protocol including a PET brain scan. Most exclusions were due to the inability to complete (inattention, impersistence, abulia) the cognitive testing, and others included uninterpretable test results or inability to undergo the neuroimaging protocols. Demographic characteristics included 12 women and 19 men, with 11 patients, each, in FTLD and AD groups and 9 in the CVD group. The mean age in years; AD 71.3 (SD 7.1), CVD 62.3 (SD 9.5), FTLD 66 (standard deviation [SD] 9.6), (analysis of variance [ANOVA] F value 2.1260, P value .319) and the mean education in years; AD 13.8 (SD 2.1), CVD 14.6 (SD 3.1), FTLD 14.8 (SD 3.2), (ANOVA F value 0.303, P value .5612).

Neuropsychological Testing

Behavioral Neurological and Neuropsychiatric Measurements

The mean T scores (normal 50 ±10) for the 3 principal frontal system syndromes for the patients with AD were abnormal for apathy 80 ± 19, executive function 75.7± 18, and total score (76 ±18) but normal for disinhibition 55 ± 12. For the CV group, the scores were all abnormal; apathy 78 ± 17, disinhibition 83 ± 27, executive function 87 ± 16 and total 90 ± 17. For the FTLD group, all the scores were abnormal: apathy 91± 21, disinhibition 75 ± 34, executive function 88 ± 22 and total 92 ± 23. The ANOVA testing revealed that disinhibition differed significantly between the AD and FTD group (P = .02) in that the latter score was abnormal, T score mean of 55 ± 12 in the AD group, which was in the normal range, and 75 ± 34 in the FTLD group.

For between-group (the 3 principal dementia syndromes) analysis, disinhibition was the only component revealing significant differences (Table 1, Figure 1) with apathy (ANOVA, P = .3650), executive function (ANOVA,P = .2937), and total score (ANOVA,P = .1797) showing no intergroup differences.

Table 1.

Disinhibition Test (FRSBE) Versus the 3 Major Dementia Syndromes.^a

Disinhibition test
Disease	N Patients	Mean	SD	Median	N Missing
AD	11	55	11	59	3
CV	9	83	17	80	1
FTLD	11	86	33	84	1

Abbreviations: AD, Alzheimer Disease; CV, Cognitive Vascular Disease; FTLD, Frontotemporal Lobe Disorder; ANOVA, analysis of variance; FRSBE, Frontal Systems Behavior Scale; SD, standard deviation.

^a Interpretation: ANOVA test shows there is a significant difference in the disinhibition test score among the 3 groups (F ratio 4.35, P = .02).

Figure 1.

One-way analysis of disinhibition by dementia subtype Box-and-Whisker Plots (red) and Quartiles (green).

Cognitive Measures (MMSE, MOCA, and FAS Test)

With regard to cognitive test, there were no differences among the 3 dementia groups in terms of MMSE scores, with ANOVA test revealing no significant difference in the MMSE test score among the 3 groups. (P = .3627). Similarly Orientation testing by the Fisher exact test indicated no significant association between the orientation score and the disease (P = .5610). Serial-7 calculation testing by the Fisher exact test indicated no significant association between the serial 7 score and the disease (P = .4831). However, the memory testing using 5 words at 5 minutes revealed a significant association with AD (Table 2). With respect to the WLT (FAS Test), ANOVA testing showed significant difference in the FAS test score among the 3 groups with scores in AD and CVD significantly higher than the one in FTLD (Table 3, Figure 2).

Table 2.

Memory for 5 Words Recalled at 5 Minutes Test Versus the 3 Major Dementia Syndromes.^a

Table of Memory 5 Words by Disease
Memory 5 Words	Disease
Frequency	AD]	CV	FTLD	Total
0	3	0	1	5
1	4	0	0	4
2	1	1	3	5
3	3	8	1	12
4	0	0	1	1
5	0	0	1	1
Total	11	9	7	28
Frequency Missing = 4

Abbreviations: AD, Alzheimer disease; CV, cognitive vascular disease; FTLD, frontotemporal lobe disorder.

^a Interpretation: Fisher exact test indicates that there is a strong association between the 5 word memory test and the disease (P = .002). Among the 12 patients with AD, 8 (67%) of their memory 5 score was either 0 or 1, while only 1 patient with FTLD memory score was 0.

Table 3.

FAS Test Versus the 3 Major Dementia Syndromes.^a

FAS Test
Disease	N Patients	Mean	SD	Median	N Missing
AD	11	10	6	12	0
CV	9	13	4	12	0
FTLD	11	6	5	5	0

Abbreviations: AD, Alzheimer disease; ANOVA, analysis of variance; CVD, cognitive vascular disease; FTLD, frontotemporal lobe disorder; SD, standard deviation.

^a Interpretation: ANOVA test shows there is significant difference in the FAS test score among the 3 groups. (P = .02). Tukey studentized range test indicates the FAS scores in AD and CVD are significantly higher than the one in FTLD (P < .05).

Figure 2.

The FAS Box and Whisker plots of the FAS (word list generation task; “how many different words with the letter “F” can you recite in one minute?”)

The results suggest that the behavioral measure of disinhibition, as measured by the FRSBE test, is a frequent accompanying symptom in both FTLD and CVD. The cognitive measures of episodic memory (5-word memory test) is poor in AD and CVD and the executive measure (FAS test) is also much more impaired in FTLD than both AD and CVD. Based on the variables of the FAS test, receiver–operator characteristic curve analyses were performed and revealed good associations between FTLD and CVD (good) and FTLD and AD (good) and less so for AD and CVD (fair; Figures 3 –5).Only two pairs were compared (FTLD vs CVD; AD vs FTLD; AD vs CVD). The results were therefore based on each of the 2 subgroups. The value of these comparisons were to determine how well the test score can discriminate the subgroups and therefore comparisons would be FTLD versus no-FTLD, AD versus no-AD, and CVD versus no-CVD. The area under the curve AUC and P values appear in Table 4. Based on Bonferroni correction, the new significant level is 0.05/3 = 0.0167, indicating that FAS can be a useful marker to distinguish FTLD versus no-FTLD.

Figure 3.

Receiver–operator characteristic (ROC) curves for pairwise comparison; Alzheimer’s disease versus cognitive vascular disorder (area under the curve AUC value 0.61).

Figure 4.

Receiver–operator characteristic (ROC) curves for pairwise comparison; frontotemporal lobe disorder versus cognitive vascular disorder (area under the curve AUC value 0.854).

Figure 5.

Receiver–operator characteristic (ROC) curves for pairwise comparison; ADAlzheimer’s disease versus frontotemporal lobe disorder (area under the curve AUC value 0.722).

Table 4.

Receiver–Operator Curves: AUC Table.

	Based on FAS test
Outcome	AUC	P value
AD	0.5556	.6161
CVD	0.7196	.0203
FTLD	0.7831	.0052

Abbreviations: Abbreviations: AD, Alzheimer’s disease; AUC, area under the curve; CVD, cognitive vascular disease; FTLD, frontotemporal lobe disorder.

The hierarchical clinical assessment stage 3 comprising neuropsychological testing of duration of several hours was possible only in a minority of our patients (less than 25%, due to inattention, impersistance, language, and other impairments) hence data from these were not further considered. The Luria Motor Sequence test component results were similarly considered noninterpretable due to a high proportion (approximately 66%) encountering difficulty with the test.

Neuroimaging

Structural Neuroimaging

There was a strong association between the MRI brain scan and the disease as calculated by the Fisher exact test (P = .0049). The MRI brain scans of all the11 patients with AD were abnormal, but only 5 of the MRI brains scans of the 11 patients with FTLD were abnormal which entailed significant leukoaraiosis, atrophy, or infarcts (Table 5).

Table 5.

MRI Brain Scan Result Versus the 3 Major Dementia Syndromes.^a

Table of MRI by disease
MRI	Disease
Frequency	AD	CV	FTLD	Total
Abnormal	11	7	5	23
Normal	0	2	6	8
Total	11	9	11	31

Abbreviations: AD, Alzheimer’s disease; CVD, cognitive vascular disease; FTLD, frontotemporal lobe disorder; MRI, magnetic resonance imaging.

^a Interpretation: Fisher exact test indicates there is a strong association between the MRI brain scan and the disease (P = .005). MRI brain scans of all the 12 patients were abnormal, while only 4 MRI brains scans of the 9 patients with FTLD were abnormal (significant leukoaraiosis, atrophy, or infarcts).

Functional Neuroimaging

Fisher exact test indicates there is a strong association between the FDG-PET result and the disease (P < .0001; Table 6). Among the 11 patients with AD who had FDG-PET, 10 (91%) of the 11 revealed bilateral temporoparietal hypometabolism, while the FDG-PET results of 7 (100%) of the 7 patients with FTD revealed bifrontal and/or temporal hypometabolism.

Table 6.

PET Brain Scan Versus the 3 Major Dementia Syndrome.^a

	PET result
Disease status	TP	Global	FT
AD	10	1	0
FTLD	0	0	7
CV	1	5	3

Abbreviations: AD, Alzheimer’s disease; FTLD, frontotemporal lobe disorder; CVD, cerebrovascular dementia; TP, temporoparietal hypometabolism; FT, frontotemporal hypometabolism; MMSE, Mini Mental State Examination; PET, positron emmision tomography.

^a Interpretation: Fisher exact test indicates there is a strong association between the PET result and the disease (P < .0001). Among the 11 patients with AD who had PET test, 10 (91%) of the results are TP, while the PET results of 100% of the patients are FT.

Discussion

Neuropsychological testing usually evaluates 5 principal domains of cerebral functioning. These include executive function, attention, intelligence, language, and memory. Behavioral syndromes (apathy, abulia, disinhibition, loss of social graces, and empathy) are not typically evaluated by standardized testing. In this study, disinhibition, an important component of FNSs was found to be significantly associated with FTLD and CVD but not AD.

Behavioral abnormalities dominate completely the FTLD, the most common dementia under the age of 60 years. On the other hand, cognitive deficits, such as executive dysfunction, typically do not feature in the disease during the initial years at a time when the disease is subclinical or classified as mild cognitive impairment. During this phase, however, the behavioral components may dramatically impact their occupation, family, and interpersonal relationships and may cause fiscal disasters. For these reasons, earlier detection is paramount.

The method of clinical detection remains challenging. In the last few years, there has been the more widespread adoption of the MOCA test as a screening test for higher cognitive functions in favor of the MMSE, the mainstay test for this purpose for the last several decades. The main advantage of the MOCA versus MMSE, is the sampling of FNSs including executive function, which is not addressed by the MMSE. However a major shortcoming of the MOCA and in fact of neuropsychological assessments in general, is the paucity of behavioral assessments, such as disinhibition, apathy, abulia, gambling tendency, promiscuity, irritability, rage attacks as well as the so-called neuropsychiatric syndromes such as obsessive behavior, compulsive behavior, and content-specific delusional behavior. These entities dominate the early years of FTLD and these behavioral abnormalities may also be the dominant features of other conditions such as stroke, multiple sclerosis, and traumatic brain injury.

With respect to memory testing, the common application of using 3 (MMSE) or 5-word (MOCA) or more (California Verbal Learning Test, Wechsler Memory Scale, and others) does not do justice to the contemporary understanding of the differing dysmemory phenotypical categories. The dementias may present with various memory disorders,including deficits in working memory (short term, localized to the frontoparietal network), episodic (long term, medial hippocampal), semantic (lateral hippocampal), and procedural (cerebellum, basal ganglia). A clinical approach to memory loss, frontal subcortical (affecting working memory, procedural), and medial temporal (episodic memory) as the 2 principal may be more useful, as these differ clinically, radiologically, and in terms of prognosis. ²⁴ However, neither working memory nor procedural memory processes are adequately tested by our current screening tests, and attention to these may improve our clinical assessment of dementias.

The WLTs have long been considered a good bedside executive measure. In this study, subgroup comparisons of 2 pairs (FTLD vs CV; AD vs FTLD; AD vs CVD) were performed to determine how well the test score can discriminate the subgroups (Figures 3 –5 and Table 4). The FAS test was found to be a useful marker to distinguish FTLD versus no-FTLD.

In addition to the clinical evaluations that were included in this study, disinhibition, word list generation and 5-word memory testing, PET brain imaging may help distinguish the 3 most common dementia subtypes. Although many different neuropsychological tests as well as a variety of behavioral inventories (FRSBE, BRIEF, Frontal Behavioral Inventory) exist, people with dementia or cognitive impairment due to stroke, traumatic brain injury, or other brain injury are rarely able to concentrate for long. Furthermore, certain disease states such as stroke mandate rapid evaluation of patients within minutes because time is brain and in other common illness states such as traumatic brain injury, markedly reduced attention and volition are major factors in the preference for quick, yet informative cognitive/behavioral testing. Finally, restricted caregiver–patient interaction time in the clinic forced by low reimbursement rates all conspire to give us distressingly little time to perform adequate testing. The disinhibition tests, word list generation, 5-word memory test that were found to significantly differentiate the disease categories in the foregoing and are relatively rapidly administered, at least within 20-30 minutes.

Looking to the future, the recent advent of diagnostically accurate functional brain imaging and cerebrospinal fluid (CSF) biomarkers afford clinicians a more comprehensive spectrum of clinical, neurocognitive, laboratory, and neuroimaging armamentarium tetrad that will likely lead to improved diagnostic acumen in this complex conundrum of dementing conditions. There is increasing evidence from clinical, functional MRI, and PET brain scan studies supporting what has been termed the cognitive reserve hypothesis. ²⁵ Briefly, normality or subclinical disease may paradoxically be associated with extensive disease such as dementia. Therefore, a combination of cognitive evaluations, metabolic brain scanning, and CSF biomarkers (phosphorylated tau and A Beta amyloid 1-42) will most likely yield the most accurate assessment for the complex dementia syndromes. ^{26 , 27} Not only will it be important to ascertain the degree of cognitive reserve but also the degree of compensation.

In this study, PET brain imaging was employed because of its established use in differentiating brain disorders especially in the context of normal anatomical brain imaging by MR scanning. In fact, no direct relationship exists between the extent of pathology and clinical manifestation of the underlying disease or damage for that matter. In our study, the PET brain imaging results were profound and correctly classified 7 of 7 patients with FTLD and 10 of 11 patients with AD and excluded FTLD or AD in 5 of 9 patients with CVD.

Functional imaging studies support the neural reserve and neural compensation reflecting individual compensatory differences to pathology. For example, 2 people with the same cognitive impairment may have markedly different degrees of underlying AD pathology. This is clearly important for the diagnosis of preclinical AD, as patients mild cognitive impairment may have both minimal pathology or more extensive pathology. The cognitive reserve hypothesis, is used to describe this variability and is considered an important part of the assessment therefore. Clinical evaluation alone cannot be relied on and biomarkers will need to be part of the workup. ²⁸ In this study using PET scanning to establish whether significant hypometabolism existed in the context of the so-called normal cognitive functioning was not found. However, the nature of patient recruitment depended on some form of cognitive complaint in the first place. Clearly, we may be missing a proportion of the so-called normal people with already mild or even moderate disease.Functional MRI shows promising results regarding the imaging of the default mode network and other recently appreciated network such as the salience network. This network-opathy approach remains under evaluation, at present, in context of mild cognitive impairment diagnosis. ²⁹

Potential criticisms of this study include the relatively small sample size of the groups, which impacts the generalizability of the results. The many variables in each disease category are also of potential concern, and it is conceded that the diagnoses in these dementia categories remain in the probable range.

Conclusion

Evaluation for disinhibition, word list generation, 5-word memory testing, and PET brain imaging may help distinguish the 3 most common dementia subtypes. Despite the compounding influence of cognitive reserve, it appears that these simple, quickly executed bedside tests may be robust enough to alert the clinician to an impending brain failure.This research supports the use of relatively simple and rapidly administered bedside type cognitive and behavioral testing even for complex dementia syndromes. Many people have neither the residual cognitive faculties nor the necessary attentional capacities required for neuropsychological testing. The important concept of considering cognitive status in the context of cognitive reserve was also supported in this research.