Prosopagnosia in the context of right handedness, left hemisphere perinatal stroke, epileptogenic cyst, and focal epilepsy: A pre-surgical case report

Highlights

• •Prosopagnosia, or face blindness, is commonly associated with the right hemisphere.
• •This case of prosopagnosia is associated with a lesion in the left hemisphere only.
• •Providers should evaluate face recognition regardless of hemisphere of the lesion.
• •Neuroimaging and neuropsychological evaluation were performed.

Abstract

Prosopagnosia, a neurological condition affecting perception and differentiation of faces, is categorized as either acquired or developmental (present since birth). Acquired cases of prosopagnosia are usually caused by right hemisphere or bilateral damage. We present a right-handed 17-year-old male with a history of focal epilepsy and a new diagnosis of prosopagnosia due to a perinatal stroke affecting the left lingual gyrus, a structure in close proximity to the fusiform face area. In addition to showing that early acquired cases of prosopagnosia may go unrecognized, this case shows that left hemisphere lesions may also affect facial recognition. It is important to screen for prosopagnosia via comprehensive neuropsychological evaluation in patients with lesions proximal to the fusiform face area.

1. Background

Prosopagnosia is a specific neurological condition in which patients are unable to accurately perceive and differentiate faces [1]. Prosopagnosia is generally categorized as either acquired or developmental; acquired cases are usually caused by a structural lesion acquired later in life. In contrast, developmental cases are present since birth and are usually not associated with a structural abnormality [2], [3]. Acquired injury early in life can present similarly to developmental prosopagnosia and some have described early acquired injuries as developmental prosopagnosia [4], but patients with clear structural abnormality and prosopagnosia are now viewed as acquired prosopagnosics. Patients with acquired prosopagnosia are typically aware of their deficit in facial perception, whereas patients with developmental prosopagnosia are often unaware of difficulty processing faces.

Acquired prosopagnosia may follow brain trauma, stroke, or epilepsy surgery [1], [5], [6], [7], [8]. Acquired prosopagnosia is usually associated with lesions in the bilateral or right fusiform face area (FFA) [9], [10], [11]. Here, we present a unique case of acquired prosopagnosia due to a perinatal injury affecting the left fusiform face area that went unrecognized by the patient and his family until epilepsy surgery workup.

2. Case report

2.1. Initial history

A right-handed white 17-year-old male with drug resistant focal epilepsy due to neonatal intraventricular hemorrhage and venous stroke underwent phase one evaluation for epilepsy surgery. He was born full term with no pregnancy or delivery complications. Neonatal seizures with apnea, cyanosis, and right fist clenching in the newborn nursery prompted transfer to a tertiary care center, where MRI showed grade IV intraventricular hemorrhage and EEG showed additional neonatal seizures. Seizures were controlled with phenobarbital and he was discharged after two weeks. Right sided hemiparesis in the NICU subsequently resolved. Motor and speech milestones were met within expected time frames. He had a typical academic course and is now a high school senior performing at grade-level. He has never needed formal or informal supports at school.

2.2. Epilepsy history and evaluation

Phenobarbital was weaned one year after birth. At age two, he began to have occasional episodes of headaches with head nod to the right, initially thought to be a combination of migraines and tics. At 12 years of age, these episodes evolved into his current semiology, characterized by sharp pain above the left eye followed by colored static in his right hemifield. After ten seconds, bilateral hands clench, eyes dart back and forth, and with some seizures his right arm extends and he has subtle clonic head jerking to the right. Some seizures then progress to inability to speak and right hemibody clonic jerking, lasting up to four minutes with postictal confusion and difficulty walking for 12 to 48 hours. Seizures generally occur one to two times monthly; they were refractory to treatment with levetiracetam and carbamazepine but improved after titration of valproic acid. Between seizures, neurologic exam was unremarkable, with no visual field deficits.

The patient was admitted for phase one epilepsy surgery workup. MRI was conducted one day before neuropsychological evaluation, and showed a large cystic space (30.9 mm x 36.2 mm) centered in the lingual gyrus of left occipital lobe immediately superior to the FFA (Fig. 1), atrophy of the left hippocampus, parahippocampal gyrus, and mammillary body, and hemosiderin staining along the hippocampal sulcus. Prior to functional MRI, the Edinburgh Handedness Inventory [12] was administered, confirming right handedness. Diffusion Tensor Imaging showed the cyst interrupts the medial left occipital subcortical white matter and outwardly displaces lateral fibers including the left inferior longitudinal fasciculus (ILF), inferior fronto-occipital fasciculus (IFOF), and optic radiations. Task-based functional MRI showed displacement of typical blood-oxygen-level-dependent (BOLD) activation in response to right visual field stimuli to the occipital cortex posterolateral to the cyst and confirms left hemisphere dominance for language tasks. Fluorodeoxyglucose − positron emission tomography (PET) was normal aside from hypometabolism in the distribution of the cystic space.

Fig. 1.

Coronal (A) and Axial (B) T2-weighted images show a large cystic space in the lingual gyrus and left fusiform face area (shown in the right hemisphere in yellow). Coronal (C) and Axial (D) diffusion tensor images show lateral displacement of the optic radiations.

Video EEG captured four brief typical seizures with head pain, head twitch, and behavioral arrest, associated with one to five second bursts of bilateral posterior polyspikes at 1–2 Hz with left hemisphere predominance on EEG. Interictal scalp EEG showed sharply contoured left temporo-occipital slowing with intermixed spikes, and high-definition EEG (128-electrode cap) with source localization showed that epileptiform discharges arose from the posterior margin of the left occipital cyst.

2.3. Neuropsychological evaluation

For neuropsychological evaluation, his parent’s primary concern was difficulty with emotion regulation. Parent and self-report checklists were not supportive of any mental health diagnoses. Results showed low average intellectual functioning (Table 1). His cognitive profile was notable only for a few specific impairments. Despite being able to accurately perceive and recall objects, an exceptionally low performance on the Weschler Memory Scales-III Faces recognition task prompted us to investigate facial processing more in depth. He was able to identify emotions. Spontaneous recognition of famous faces was low (2/10) and improved with semantic cuing (7/10; i.e., occupation) [13]. He did not recognize the three unidentified faces. No pediatric-normed assessment of famous faces exists [14], so we created our own for clinical use (see supplemental material for details). The Prosopagnosia Index-20 (PI-20) [15] was added strictly for academic purposes and was rated as within normal limits, although no pediatric norms exist. Endorsements on the PI-20 contradicted clinical report of difficulty recognizing faces of familiar people and famous individuals. For example, the patient stated he “never pictures individual faces in his mind,” but provided a rating of ‘3’ on the 1–5 “strongly disagree” to “strongly agree” Likert scale for the item “I find it easy to picture individual faces in my mind.” Similarly, he “strongly disagreed” with “my friends and family think I have bad face recognition or bad face memory” but his parents stated this skill is problematic. He reported difficulty “putting a name to a face,” recognizing himself in photographs, differentiating people wearing similar clothing, and difficulty recognizing celebrities in “before-they-were-famous” photos. He explained he often uses movement cues [16], voice, and clothing as identifiers.

Table 1.

Neuropsychological test results.

Test	Score	Percentile
Wechsler Adult Intelligence Scale – 4th Edition
Verbal Comprehension Index	SS=87	19
Perceptual Reasoning Index	SS=81	10
Working Memory Index	SS=89	23
Processing Speed Index	SS=86	18
Full Scale IQ	SS=82	12
Similarities	ss = 8	25
Vocabulary	ss = 9	37
Information	ss = 6	9
Block Design	ss = 6	9
Matrix Reasoning	ss = 7	16
Visual Puzzles	ss = 7	16
Digit Span	ss = 7	16
Arithmetic	ss = 9	37
Symbol Search	ss = 7	16
Coding	ss = 8	25
Conners’ Continuous Performance Test – 3rd Edition
Detectability	T=53	62
Omissions	T=47	38
Commissions	T=54	66
Perseverations	T=48	42
Hit Reaction Time (HRT)	T=56	73
HRT Standard Deviation	T=50	50
Variability	T=55	69
HRT Block Change	T=46	34
HRT Inter-Stimulus Interval Change	T=42	21
Delis Kaplan Executive Function System (DKEFS) – Verbal Fluency
Letter Fluency	ss = 14	91
Category Fluency	ss = 8	25
Category Switching Total Correct	ss = 8	51
Category Switching Totally Switching Accuracy	ss = 7	16
Set-Loss Errors	ss = 6	9
Repetition Errors	ss = 8	25
DKEFS Trail Making Test
Visual Scanning	ss = 11	63
Number Sequencing	ss = 9	37
Letter Sequencing	ss = 8	25
Number Letter Sequencing	ss = 9	37
Motor Speed	ss = 12	75
California Verbal Learning Test – 3rd Edition
List A Total Trials 1–5	SS=96	39
List A Trial 1 Free Recall	ss = 10	50
List A Trial 5 Free Recall	ss = 8	25
List B Free Recall	ss = 11	63
List A Short Delay Free Recall	ss = 7	16
List A Short Delay Cued Recall	ss = 8	25
List A Long Delay Free Recall	ss = 7	16
List A Long Delay Cued Recall	ss = 7	16
Recognition Hits	ss = 6	9
Total Recognition Discriminability	ss = 4	2
Wechsler Memory Scale – 4th Edition
Visual Reproduction I	ss = 8	25
Visual Reproduction II	ss = 7	16
Visual Reproduction Recognition	−	26–50
Judgment of Line Orientation
Total Correct	Raw = 24	40
Hooper Visual Organization Task
Total Correct	T=52	58
Wide Range Assessment of Memory and Learning – 2
Finger Windows	ss = 2	<1
Beery-Buktenica Developmental Test of Visual Motor Integration (VMI) – 6th Edition
VMI	SS=72	3
Grooved Pegboard
Dominant Hand (right) Speed	Z=-2.98	<1
Non-Dominant Hand (left) Speed	Z=-2.30	1
Dominant Hand Drops	Raw = 3	−
Non-Dominant Hand Drops	Raw = 3	−
Rey Complex Figure Test
Copy	Raw = 26	<1
Pyramids and Palm Trees
Total Correct	Raw = 47 out of 52
Famous Faces
Total Correct − Named	Raw = 2 out of 10
Recognition	Raw = 7 out of 10
Emotion Faces (FACES database [18])
Total Correct	Raw = 5 out of 6
Wechsler Memory Scale – 3
Faces Recognition I	ss = 3	1
Faces Recognition II	ss = 3	1
Benton Facial Recognition
Total Correct	Raw = 36	<1
Wechsler Individual Achievement Test – 3
Word Reading	SS=83	13
Numerical Operations	SS=99	47
Spelling	SS=82	12

Note. Standard scores (SS): M=100, SD=15; scaled scores (ss): M=10, SD=3; T-scores: M=50, SD=10; z-scores M=0, SD=1.

Crawford et al. [17] describe methods for testing for deficits in a single case compared to a control sample. Using Crawford’s singlims_ES.exe program, we confirmed our patient’s facial recognition is impaired on the Benton Face Recognition Test (BFRT) compared to controls (two-tailed t(2 8 6) = -2.37, p = 0.018, 95% CI [-2.6 to -2.15]). Using Crawford’s point estimate calculation, his score on the BFRT is below the second percentile. Applying Crawford’s point estimate calculation, his object recognition on the Hooper Visual Organization Task is better than 83 percent of controls (two-tailed t(1 6 6) = 0.22, p = 0.83, 95% CI [0.06 to 0.37]); note these percentiles do not correspond to the percentiles of his standard scores reported in Table 1. The selective impairment in face recognition in the context of intact and even strong object recognition supports a diagnosis of prosopagnosia.

3. Discussion

This patient’s prosopagnosia is presumably due to his intraventricular hemorrhage, and this is best categorized as an acquired prosopagnosia. This pre-surgical epilepsy evaluation case is remarkable in that prosopagnosia is caused by a unilateral lesion in the left (dominant) hemisphere. Right handedness and left hemisphere activation on fMRI with language tasks support left hemisphere dominance, and clear localization of epilepsy to the left hemisphere with absence of right hemisphere imaging abnormalities on MRI, fMRI, and PET make an alternate cause of prosopagnosia in the right hemisphere unlikely, although undetected right hemisphere dysfunction resulting from the patient’s epilepsy cannot be totally excluded. It also seems likely that face recognition deficits are not caused by poor perceptual organization, given many of his perceptual skills (i.e., line orientation perception, mental rotation, combining puzzle pieces to identify an object) were strong.

Acquired prosopagnosia is most often seen with bilateral lesions in or around the FFA; unilateral lesions to the FFA resulting in prosopagnosia are generally reported in the right hemisphere [2], [9], [11]. Facial encoding has been traditionally conceptualized as a type of visual-spatial processing that lateralizes to the right hemisphere [19]; however, this case clearly defies that conceptualization. It is possible that other cases of unilateral left hemisphere acquired prosopagnosia have been overlooked [2] and that facial encoding is more weakly lateralized to the right hemisphere, as a recent fMRI study suggests [20]. Another fascinating possibility is that left hemisphere lesions may have a greater impact on development of facial recognition early in life [21]. In developmental prosopagnosia, decreased activation of the left hemisphere across many structures is seen with fMRI facial recognition tasks, but not with recognition of other visual stimuli such as words or lines [21], supporting a role for the left lingual gyrus in facial encoding and recognition.

Although the lingual gyrus was not identified as a “face preferential” region by Joseph et al. [22], younger children use the left hemisphere for face processing more than older children and adults do, and over time, the right hemisphere becomes more specialized for facial recognition [22], [23]. It may be that the left hemisphere becomes less critical for facial recognition as development progresses, which Gerlach et al. [21] theorize may explain milder deficits in developmental or early acquired prosopagnosia.

These points have implications for clinical management. Patients with an epileptogenic zone involving the FFA and nearby structures should be counseled about the potential for acquired prosopagnosia, even if the lesion is in the left hemisphere. Additionally, patients with lesions near the fusiform gyrus should be screened for prosopagnosia, even if they do not complain of symptoms. While usually associated with developmental prosopagnosia, this patient experienced notable lack of insight into the difficulty he had identifying faces. Identification of prosopagnosia allows earlier identification of deficits in facial processing, which in turn allows for earlier intervention. Individuals with prosopagnosia may improve recognition of others using compensatory strategies (i.e., using non-facial information to identify others) and remedial strategies [24], [25]. These strategies show promise in remediating facial recognition. In general, early interventions are known to improve outcomes for children with disorders and disabilities. Studies indicate remedial strategies specifically for facial perception are associated with improved facial recognition [24]. Promptly identifying children with prosopagnosia and equipping them with these strategies may improve their quality of life.

Ethical Statement

This project was completed with informed, written consent from the patient and the legal guardian. Consent was provided free from coercion. Written consent was obtained on 3/15/2023.

CRediT authorship contribution statement

Alanna Kessler-Jones: Writing – review & editing, Writing – original draft, Supervision, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Tayler M. Cieminski: Writing – review & editing, Writing – original draft, Validation, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Aaron Field: Writing – review & editing, Resources. Andrew Knox: Writing – review & editing, Writing – original draft, Validation, Supervision, Methodology, Data curation, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: