When real life conditions do not follow textbook diagnosis: A diagnostic challenge from a neuropsychiatric clinic

Introduction

Neurodegenerative disorder is a general term for insidious and progressive decline in cognition often accompanied by emotional and behavioural disturbances. The diagnosis of neurodegenerative disorders is often challenging in the context of anxiety disorders and an overlay of functional somatic dramatization. Accurate estimate of premorbid functioning has a pivotal value in understanding patient’s current level of functioning and ultimate diagnosis. Neurodegenerative disorders encompass a number of common diagnoses including Alzheimer’s Disease (AD), Vascular Cognitive Impairment, Dementia with Lewy Bodies (DLB), and Frontotemporal Lobar Degeneration (FTLB) among other causes. Patients with cognitive deficit leading to a decline from their previous level of functioning, corroborated by history and neuropsychological examination, would be considered for a diagnosis of neurodegenerative disorder. If their change in functionality does not interfere with instrumental activities of daily living (IADLS) and activities of daily living (ADL), the diagnosis of mild cognitive impairment (MCI) (equivalent terminology of mild neurocognitive disorder from the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition) would be considered. ¹ When cognitive decline interferes with independent living and functioning, patients meet criteria for dementia (major neurocognitive disorder). ² Neurodegenerative disorders involve a wide range of cognitive domains including memory and learning, executive functioning, language, and visuospatial with varying degrees. ³ Some neurodegenerative disorders may have symptoms that overlap. Imaging and CSF biomarkers are helpful to narrow the diagnosis. However, these advanced modalities are often unavailable or practically difficult to pursue in certain subpopulations.

This case covers the multi-layered neuropsychiatric manifestations of brain diseases and questions the boundaries of clear-cut DSM chart list diagnoses. It also highlights the practical challenges that patients with co-morbid cognitive and severe anxiety disorders pose for the clinician, who needs to disentangle the individual and complicating role of each disorder against the complex influence of premorbid functioning and learning disabilities in cognitive decline.

Case presentation

We present the case of a 61-year-old single, African American, retired, catholic priest with a 4-year history of cognitive decline, initially involving the domains of memory and language, and later affecting his executive function and visuospatial abilities. He complained of forgetting conversations, repeating questions, asking for clarifications, and misplacing objects often. At least once he got lost driving around his hometown in Western Massachusetts. He reported word finding difficulties consisting of being unable to remember the ending of words as he was pronouncing them. He endorsed frustration related to his cognitive problems, which had greatly affected his ability to hold his job and being independent.

His birth and early developmental history are unknown, since he was adopted at a very young age. His parents report no difficulty with developmental milestones. Throughout school, he was an average C student and struggled in math. He was not known to have intellectual or learning disability, particularly dyslexia. After finishing high school, he moved to California and attended seminary college. He then worked for over 20 years in different churches. He was a popular priest and his sermons were well attended.

At age 46, he left his parish and came back to live at home with his elder parents. About 10 years prior to his retirement, he started to experience memory gaps such as forgetting his sermon notes and word finding difficulties while preaching. At age 57, he sought medical attention for his cognitive problems. His mother was also concerned by the impact of his anxiety in his day-to-day life. A note of a previous neurological assessment indicated that the patient could not repeat a story and could not draw a clock while the rest of the neurological exam was unremarkable.

He underwent a neuropsychological testing in 2015 prior to establishing care at our clinic. The results of American National Adult Reading Test (ANART) showed average to high average (60th to 80th percentile) premorbid performance. He performed high average (75th %) on verbal fluency. He performed average on measures of oral world reading (WRAT-IV) (60th %), spelling (45th %), general information (WAIS-IV) (63rd %), verbal reasoning (similarities) (50th%), vocabulary (SILS) (42nd %), and basic construction skills (CERAD praxis) (37th %). Patient’s performance fell within Low average on abstract sequence completion (SILS) (16th %), Non-verbal reasoning (WAIS-IV matrices) (9th %). His performance was noted to be below average on measures of auditory attention span (WAIS-IV, digit span; 6 forwards, 3 backward) (2nd %), visual reasoning (visual puzzles) (5th %), animal fluency (8th %), MMSE: 23/30 (0/3 on recall), sustained attention (CPT-II, inattentive, impulsive) (1st %), processing speed (TMT, PART A) (1st %), divided attention (TMT, PART B)(1st %), naming (BNT < 1st %), written arithmetic (WRAT-IV) (<1st %), verbal memory (WMS-IV) including logical memory (1st %) and recognition (2nd %), verbal list learning (CVLT-II) (<1st %), and visual memory (WMS-IV) including visual reproduction (1st %) and recognition (2nd %). No measures reported his effort during the exam.

A striking discrepancy between the neurocognitive assessment and his report of being able to live independently was highlighted. It was also noteworthy that he appeared to improve in neurocognitive assessments when he did not have to perform in front of others. These observations lead the treating physician to conclude that anxiety was likely the cause of his cognitive complains. A year later, he stopped delivering any sermon and subsequently was put in charge of nuns’ training on a part time basis.

He initially presented to our neuropsychiatry clinic for evaluation of cognitive decline in 2017. He endorsed worsening of memory since 2015. At that point, he was still able to handle his own finances and drive his car without having received a citation or having being involved in an accident. He also endorsed anxiety in the context of some psychosocial challenges. Patient had been placed on fluoxetine 40 mg and venlafaxine 37.5 mg for his anxiety, and depression. We stopped his venlafaxine and adjusted the dose of fluoxetine to 60 mg to avoid drug-drug interaction. Later, buspirone 30 mg daily was added to help augment the effects of fluoxetine. As the patient was unable to manage taking medication twice daily, buspirone was administered once a day. Also, as Alzheimer disease was one of the main differential diagnosis, donepezil was started and continued with undetermined efficacy.

In the ensuing months, he continued to be followed in the neuropsychiatry clinic. During these visits, we were able to elicit more information and to better observe his behavior. We did not notice signs of disinhibition or changes of social graces. He often appeared to be disoriented to day and date including month, and year. He knew the name of the hospital, city, state, and country. During conversations, he would shift topics or lose focus quickly. His performance and behavior during repeated attempts at testing working memory and sustained attention suggested poor effort and heightened anxiety. In the first trial of memory testing, he registered 3/3 but had 0/3 on spontaneous recall and refused to undergo recognition testing. On the following visit, he was able to register 5/5 but refused spontaneous recall, becoming tearful stating that he could not remember anything. His speech was overall fluent with normal volume, rate, and tone but was noted to have word-finding difficulties. He made occasional phonemic paraphasic errors as well as semantic paraphasic errors (e.g. “pen” for pencil cap, “phone” for speaker). He had difficulty understanding more complex questions. He was unable to carry out complex 3 step commands. He refused to get tested for praxis. On visuospatial skill assessment, he drew a cube in 3D albeit it’s orientation was not accurate. On the clock drawing test, he drew a circle but did not put hands or numbers. His neurological exam never evidenced signs of parkinsonism. He and his mother also denied a history of hallucinations or delusions.

On a follow up visit in 2018, he eventually agreed to perform a Montreal Cognitive Assessment (MoCA) after major hesitation. He scored 6/30 showing significant difficulties in executive functioning, registration, attention, and language. He was only able to register 1 word (out of 5) and was not able to recall it after 5 minutes. He was disoriented to the date including the year. His parents reported more decline in his function at home. He was rarely driving a car at that point. In a follow up visit in 2019, he agreed to undergo a more detailed office-based speech-language exam, given the involvement of language in the initial and follow up presentations. He did not appear to have significant difficulties with oral–motor mechanisms, but his speech was somewhat apractic as he struggled in sequential motion rates (sequential motor rates (SMRs): ability to move quickly and sequentially from one articulatory posture to another). He had difficulty with repetition of words worse for longer syllable words. He had difficulty with spelling (e.g.: “elphnt “for elephant, “asile” for aisle, and “yatch” for yacht). His reading of irregular words was impaired. He struggled in naming (9/14) and had difficulty with understanding semantic correlates. On phonemic fluency, he just produced 3 words and on categorical fluency he was only able to generate 2 animals, and 3 vegetables (with errors of inclusion as he included water). On picture description (see Figure 1), he was unable to appropriately describe the event. His description was empty with grammatical and spelling errors. He struggled to read his own writing. Figure 2 shows his low performance on clock drawing. In the visits spanning 2018–2019, it was noted that anxiety and depression continued to play a role despite modest improvement on a combination of fluoxetine 60 mg daily and buspirone 30 mg daily. From a functional standpoint, he became dependent on his parents for his IADLs but continued to be independent with his personal care. During his appointments, he was overall neatly dressed and had appropriate grooming. His social grace was appropriate. He was no longer driving and relied on his parents to accompany him to all of his appointments.

OPEN IN VIEWER

Figure 1.

Patient’s picture description, showing an empty description with poor organization of words, and many spelling and grammatical errors.

OPEN IN VIEWER

Figure 2.

(Clock drawing – showing dementia clock (i.e. the meaning of the clock is lost) – in addition to significant difficulties with visuoconstructive skills, organization of numbers, and executive functioning.

Laboratory work up of metabolic and infectious causes of dementia such as thyroid function test, folic acid, Vitamin B12, HIV, and RPR were within normal range.

MRI of brain was performed without contrast and demonstrated minimal changes related to chronic microangiopathy, without evidence of a significant burden of microvascular disease, prior infarcts, or evidence of microhemorrhages/intracranial hemorrhages. There was moderate prominence of the ventricular system with moderate widening of the biparietal and bifrontal lobe sulci consistent with moderate parenchymal volume loss (Figure 3).

OPEN IN VIEWER

Figure 3.

Brain MRI showing significant atrophy in different cortical areas including frontal, parietal, and temporal areas. There is a slight pattern of asymmetry with more involvement of left hemisphere.

Fluorodeoxyglucose-positron emission tomography (FDG-PET) (Figure 4) demonstrated markedly decreased metabolism in a nearly symmetric distribution involving the posterior cingulate gyri, precuneus regions, and lateral aspects of parietal lobes. Reduced metabolic activity was also noted anteriorly involving the anterior cingulate gyri, frontal lobes, and anterior temporal lobes, bilaterally. FDG avidity in the remaining portions of the cerebral hemispheres including the occipital lobes, dorsal striata, thalami, midbrain, pons and both cerebellar hemispheres appeared within physiologic limits.

OPEN IN VIEWER

Figure 4.

FDG-PET showing general hypometabolism in multiple cortical areas with relative preservation of occipital area.

Due to variability in presentation, rapid decline in cognition, and concerns over under-performance on cognitive examinations and overall functionality, the team decided to pursue the Test of Memory Malingering (TOMM) ⁴ questioning an overlay of affective dramatization of symptomatology as well as a lumbar puncture to check for biomarkers of neurocognitive disorders. Unfortunately, on his next follow up visit, the patient became agitated while waiting in the waiting area, pacing around without staff being able to redirect him; he became loud and stormed off the clinic with no intention to reschedule. He refused to pursue TOMM and LP for further evaluation.

Discussion

This case highlights the interplay of anxiety and cognitive dysfunction in an individual with history and functional decline concerning for neurodegenerative disorders. His presentation, in addition to the discrepancy between his functional performance and the initial neuropsychological reports, centred on the role of performance anxiety and functional neurological disorder. Functional cognitive disorder is an evolving concept that currently lacks objective imaging markers. Using PET, a cohort of patients with major depression with cognitive impairment were compared with a group of equally depressed non-cognitively impaired depressed patients. Regional cerebral blood flow abnormalities showed decreases in the left anterior medial prefrontal cortex and increases in the cerebellar vermis. While this finding needs to be replicated, the authors concluded that patient’s widespread multi-cortical hypometabolism in the FDG-PET imaging likely indicates a neurodegenerative process rather than what would be expected from functional cognitive impairment.

An accurate estimate of his premorbid childhood functioning and/or disability was never attempted. We also found out that he was never assigned to an Individualized Educational Program (IEP) in school. His adoption at a very young age with no knowledge of his family history rendered the search for genetic vulnerabilities difficult to pursue. The possibility of borderline intellectual functioning and/or undiagnosed childhood learning disability was considered, as the patient was a C student during school and struggled in some topics. Interestingly, he was successful in his career as a priest and often would address large congregations. American National Adult Reading Test (ANART), as an estimate of premorbid functioning, showed average to high average performance which would argue against borderline intellectual functioning.

While anxiety and depression could certainly cause and exacerbate cognitive problems, the pattern of memory and visuospatial decline as well as his language decline with semantic and phonemic paraphasic errors were not consistent with a primary mood or anxiety disorder. Given his overall poor performance in IADLs, full dependence on his parents, the poor performance on different cognitive domains in neuropsychological testing and the evidence of significant parietal >frontal atrophy with marked cerebral hypometabolism as well as the emerging behavioural disturbances, major neurocognitive disorder due to neurodegeneration was highest on our differential diagnoses list. While early onset Alzheimer’s disease (AD) is the most common early-onset neurodegenerative disorder, ³ given the relatively atypical presentation, we considered other phenotypic variants.

The initial involvement of both episodic memory and language could suggest a diagnosis of frontolobar degeneration (FTLD). Because of the absence of apathy, compulsions, and interpersonal behavioral issues, he would not fit the diagnosis of behavioural variant of FTD (BvFTD). ⁵ Additionally, the patient’s social interest and personal warmth were maintained throughout. Patient’s behavioural disturbances including agitation, and irritability thought to be more typical for AD rather than bvFTD. The possibility of logopenic variant of primary progressive aphasia (PPA) was entertained given his difficulty with single word retrieval, repetition, and phonological errors. However, his single word comprehension and object knowledge were also impacted rendering this diagnosis unlikely. Frank agrammatism in his written language, observed in the picture description (Figure 1), would also argue against logopenic PPA. ⁶ The lack of a clear asymmetry involving the temporal lobe, particularly the left temporal lobe on both MRI and FDG-PET made the logopenic PPA a much less likely differential consideration. ⁷

Unlike typical cases of Posterior Cortical Atrophy (PCA), another phenotypic variant of early-onset AD, this patient did not appear to have visuospatial deficits that predated and exceeded memory deficits at least in the initial phases of the condition. The preservation of metabolic activity within the occipital lobes and lack of the confounding clinical history of visual hallucinations made dementia of Lewy Body (DLB) much less likely.

The nearly symmetric pattern of severe hypometabolism in FDG-PET involving the frontal, temporal, and parietal lobes with involvement of the anterior and posterior cingulate gyri raised concern for an early-onset AD. ⁶ Studies have shown early-onset AD have a faster deterioration than late-onset AD which is compatible with this case with fairly aggressive clinical course.^8,9 While a relatively high percentage of patients with early-onset AD present with language, executive functioning, or visuospatial symptoms than usual amnestic variant, ¹⁰ difficulty with memory was a cardinal symptom from the beginning in this case. Additionally, patients with early-onset AD tend to show larger cortical volume loss in temporo-parietal lobes and relative sparing of hippocampal areas compared to late-onset AD^10,11 which appear to be reflected in this case as well (Figure 3). Based upon our clinical impression of early-onset AD, the dose of donepezil was adjusted and was augmented with memantine given the severity of cognitive decline. Patient’s behavioral disturbances, unfortunately, deteriorated to the level of violent behavior leading to acute hospitalization and nursing home placement.