Cognitive and Psychological Functioning in Chiari Malformation Type I Before and After Surgical Decompression - A Prospective Cohort Study

Abstract

BACKGROUND

Chiari Malformation Type I (CM-I) is defined as cerebellar tonsil displacement more than 5 mm below the foramen magnum. This displacement can alter cerebrospinal fluid flow at the cervicomedullary junction resulting in Valsalva-induced headaches and syringomyelia and compress the brainstem resulting in bulbar symptoms. However, little is known about cognitive and psychological changes in CM-I.

OBJECTIVE

To prospectively assess cognitive and psychological performance in CM-I and determine whether changes occur after surgical decompression.

METHODS

Blinded evaluators assessed symptomatic CM-I patients ages ≥18 with a battery of neuropsychological and psychological tests. Testing was conducted preoperatively and 6 to 18 mo postoperatively. Data were converted to Z-scores based on normative data, and t-tests were used to analyze pre-post changes.

RESULTS

A total of 26 patients were included, with 19 completing both pre- and post-op cognitive assessments. All patients had resolution of Valsalva-induced headaches and there was improvement in swallowing dysfunction (P < .0001), ataxia (P = .008), and sleep apnea (P = .021). Baseline performances in visual perception and construction (z = −1.11, P = .001) and visuospatial memory (z = −0.93, P = .002) were below average. Pre-post comparisons showed that CM-I patients had stable cognitive and psychological functioning after surgery, without significant changes from preoperative levels.

CONCLUSION

CM-I patients had below average performance in visuospatial and visuoconstructional abilities preoperatively. Prospective longitudinal data following surgery demonstrated improved neurologic status without any decline in cognition or psychological functioning. Routine pre- and postoperative formal neuropsychological assessment in CM-I patients help quantify cognitive and behavioral changes associated with surgical decompression.

Graphical Abstract

Graphical Abstract.

ABBREVIATIONS

BAI

beck anxiety inventory

BDI

Beck depression inventory

BPI

brief pain inventory

CM-I

Chiari Malformation Type I

CFT

complex figure test

GPT

grooved pegboard test

JOLO

judgment of line orientation test

PICA

posterior inferior cerebellar artery

SCA

superior cerebellar artery

WAIS

Wechsler adult intelligence scale

WMS

Wechsler memory scale

WRAT

wide range achievement test

Chiari Malformation Type I (CM-I) is defined as cerebellar tonsil displacement more than 5 mm below the foramen magnum. This displacement can alter cerebrospinal fluid (CSF) flow at the cervicomedullary junction and lead to Valsalva-induced headaches, syringomyelia, and bulbar symptoms. Increasing evidence implicates the cerebellum's involvement in cognition.^1-4 Recently, researchers have begun to evaluate how cognition is affected in patients with CM-I. For example, a few studies on CM-I have identified difficulties in memory, visuoconstruction, and decision-making that affect quality of life.^5-9 Compared to controls, nonsurgically treated CM-I patients have displayed deficits in attention, visuoperceptual processing, and executive functions¹⁰ in addition to problems with processing speed and facial recognition.¹¹ Additional deficits in attention have been attributed to chronic pain and cervicomedullary compression.¹² However, retrospective designs, subjective evaluations, small sample sizes, and/or narrow neuropsychological assessments limit these studies.

Though surgery has been shown to improve Valsalva-induced headache and neurologic function,¹³ surgical effects on cognitive performance have not been definitively characterized.¹⁴ Studies comparing controls to a mix of operated and unoperated CM-I patients have found deficits in similar domains in the CM-I group¹⁵ with little difference between CM-I groups¹⁶; interestingly, patients with CM-I have been noted to display disorganized complex figure copy performances (a task of visuospatial construction).¹⁷ However, a prospective longitudinal study with comprehensive analysis of CM-I before and after surgery defining the domains, magnitude and direction of cognitive changes is lacking. Therefore, we sought to characterize the cognitive and psychological profiles in patients with CM-I and to assess how posterior fossa decompression affects performance on a wide range of cognitive measures.

METHODS

Study Design

Participants were CM-I surgical candidates based on previously published criteria.¹⁸ Inclusion/exclusion criteria, enrollment process, excluded patients, and timeline of testing are listed in Figure 1. The University of Iowa Institutional Review Board approved this study, and all patients gave informed consent.

FIGURE 1.

Flow diagram detailing the enrollment and screening process with accompanying inclusion/exclusion criteria, study design and testing timeline.

Neuropsychological and Psychological Measures

Each participant was assessed with a battery of standard neuropsychological and psychological tests administered by technicians supervised by a board-certified clinical neuropsychologist blinded to the clinical status of the subjects. The assessment battery was designed to probe the major domains of cognitive functioning (attention, visuospatial, memory, verbal, processing speed, and executive function)¹⁹ using established clinical testing protocols.²⁰ The 26 neuropsychological and psychological tests used for analysis are included in Table 1.

TABLE 1.

Neuropsychology and Psychology Tests and Functions Assessed

Test/Task	Function
WAIS IV
Similarities	Abstract thinking
Vocabulary	Word knowledge
Block design	Visuoconstructional processing
Matrix reasoning	Inductive reasoning
Digit span	Working memory
Symbol search	Processing speed
Digit-symbol coding	Processing speed
WRAT:
Reading test	IQ proxy
Spelling test	Spelling aptitude
Rey auditory verbal learning test (AVLT):
Immediate (AVLT 6)	Verbal memory
30-min delayed recall (AVLT 30)	Delayed verbal memory
Rey-Osterrieth CFT:
Copy	Visuospatial construction
30-min delayed recall	Visuospatial construction and memory
WMS spatial memory subtest (Spatial Span)	Spatial memory
Controlled Oral Word Association Test (COWA)	Verbal fluency
JOLO	Visual perception
Lafayette GPT:
Dominant (dom)	Motor coordination
Nondominant (nondom) hand	Motor coordination
Trail making test:
Part A (TMT A)	Processing speed/visual search
Part B (TMT B)	Divided attention and set shifting
Stroop test:
Color	Visual perception/color naming
Word	Speed reading
Color/word	Color/reading congruency
Interference	Response inhibition
BAI	Anxiety symptoms
BDI	Depressive symptoms
BPI	Pain interference in daily living

Italicized tasks represent those using alternate forms at postoperative evaluation to minimize practice effects

Operative Intervention Technique

All surgeries were performed by 2 attending neurosurgeons (B.D. or A.M.), consisting of a standard prone suboccipital craniectomy. After osseous decompression, excision of the epidural compression band and external durotomy, intraoperative ultrasonography was performed to evaluate the adequacy of decompression assessing tonsillar configuration, tonsillar pulsation, brainstem decompression, and CSF cleft posterior to the brainstem. If further decompression was warranted, an intradural exploration with tonsillar reduction via low-power pia-arachnoid coagulation, intradural lysis of adhesions, opening of fourth ventricle, and autologous cervical fascia duraplasty was performed.²¹

Data Analysis

Statistical analysis was conducted with SPSS (IBM SPSS Statistics version 26 for Mac, IBM Corp., Armonk, NY, USA). Graphical illustrations were constructed using Graphpad Prism (GraphPad Prism version 7.0c for Mac, GraphPad Software, La Jolla, CA, USA).

For comparisons to healthy populations, neuropsychological data were transformed into z-scores accounting for age, years of education and, where appropriate, sex. Means and standard deviations for z-score calculation were obtained from testing manuals. Beck depression inventory (BDI), beck anxiety inventory (BAI), and brief pain inventory (BPI) are reported as raw scores.^22-26 Two cohorts were assessed: participants that completed only preoperative testing (pre-op cohort) and participants that completed testing at both time points (pre-post cohort).

To draw statistical inferences to normative data from a healthy population, a one-samples t-test was performed for each neuropsychological assessment. Paired-samples t-tests were performed to identify statistical change between time points. A Benjamini-Hochberg procedure was performed to account for multiple comparisons with false discovery rate set at 25%. The effect of a syrinx on motor related tasks and approach related differences was assessed with independent samples t-test, and the impact of motor performance as assessed by the grooved pegboard test (GPT)-dom task on motor related tasks was assessed with Pearson correlation statistic. To assess the impact of mood on neuropsychological performance, a Pearson correlation statistic with the BDI score was performed for each assessment. Statistical significance was defined as P < .05 and a statistical trend as P < .1. Clinical outcome variables were assessed with two-tailed t-tests and Chi-Square or Fisher Exact tests.

RESULTS

Demographics

A total of 26 patients met appropriate inclusion and exclusion criteria and consented to participate in research. Of those, 19 completed both preoperative and postoperative testing with mean follow-up testing at 12.7 mo; the remaining 7 had preoperative testing only (Table 2, Figure 1).

TABLE 2.

Demographics of the Total Preoperative CM-I Cohort and the CM-I Cohort That Completed Both Preoperative and Postoperative Neuropsychology and Psychology Testing

Characteristic	Total preoperative CM-I cohort value	Preop/postop CM-I cohort value
n	26	19
Mean age (range)	25.5 (18.4-56.2)	36.6 (18.4–56.2)
Mean education (range)	14.1 (9-20)	14.2 (9-20)
Female gender (%)	23 (88.5%)	19 (100.0%)
Handedness right/left/mixed	20/3/3	15/2/2
Mean (SD) follow-up testing time (mo)	–	12.7 (6.4)

Values are expressed as the number of patients (%) or as the mean (range).

Clinical Features and Outcomes

Clinical features and outcomes are presented in Table 3. The mean tonsillar descent was 15.2 mm (SD 4.8mm). Valsalva-induced headache was present in all cases. Most patients presented with bulbar signs or symptoms. All patients had routine clinical postsurgical follow-up with median follow-up of 12 mo (range 6-42). All patients had resolution of Valsalva-induced headaches and most saw improvement in brainstem dysfunction, particularly swallowing dysfunction and ataxia.

TABLE 3.

Clinical Presentation Of The Total Preoperative CM-I Cohort and CM-I Cohort That Completed Both Preoperative and Postoperative Neuropsychology and Psychology Testing

	Total preoperative CM-I cohort, n = 26			Preop/postop CM-I cohort, n = 19
Clinical feature	pre-op	post-op	P-value	pre-op	post-op	P-value
Mean TonsillarDescent (SD), mm	15.2 (4.8)Intradural (n = 9): 18.4Extradural (n = 17): 13.5	–	.010	15.0 (4.8)Intradural (n = 6): 18.3Extradural (n = 13): 13.5	–	.037
Valsalva-induced headache	26 (100%)	0 (0%)	–	19 (100%)	0 (0%)	–
Swallowing dysfunction	20 (77%)	4(15%)	<.0001	13 (68%)	4 (21%)	.003
Altered facial Sensation	4 (15%)	1 (4%)	.158	4 (21%)	1 (5%)	.340
Ataxia	10 (38%)	2 (8%)	.008	7 (37%)	2 (11%)	.124
Sleep apnea	7 (27%)	1 (4%)	.021	6 (32%)	1 (5%)	.090
Nystagmus	5 (19%)	2 (8%)	.385	3 (16%)	1 (5%)	.604
Syringomyelia	11 (42%)	4R/3I/3S	–	7 (37%)	3R/2I/2S	–
Scoliosis	2 (8%)	–	–	2 (11%)	–	–

Values are expressed as the number of patients (%) or as the mean (standard deviation). R - Resolved; I - Improved; S - Stable; SD - Standard Deviation. Statistical measures included 2 samples T-test for scalar variables and Chi-Square or Fisher's Exact Test for categorical variables. Bolded values indicate statistical significance (P < .05) and italicized for statistical trend (<.1).

No patients required readmission postsurgery for any cause, and no patients developed aseptic meningitis, CSF leak, pseudomeningocele, infection, development of hydrocephalus, need for ventriculoperitoneal shunt, or need for reoperation.

Preoperative Cognitive Findings

Comparison to Normative Data

CM-I patients performed significantly lower than normal on complex figure test (CFT)-Copy, a task measuring visual perception and visuospatial construction (z = −1.11, P = .001). CM-I patients also performed significantly lower than normal on CFT—Recall, a task that assesses both construction and memory (z = −0.93, P = .002) (Figure 2, Table 4). Patients also performed worse on motor tasks in both the dominant (GPT-dom, z = −1.35, P = .025) and nondominant hands (GPT-nondom, z = −1.38, P = .040). The remaining tests showed normal performance. Syringomyelia had no impact on CFT-Copy (P = .300) or either GPT conditions (dominant [P = .659] and nondominant [P = .588] hands), indicating the significant differences in these tests were not spinal cord/syrinx related. Interestingly, syrinx did affect the performance on CFT-Recall (−1.46 vs −0.47, P = .035). There was a moderate correlation in the performance on the GPT-dom tasks and CFT – Copy (Pearson r = 0.493, P = .023) and CFT – Recall (r = .658, P = .028).

FIGURE 2.

Mean preoperative z-scores on a wide range of clinical neuropsychological tests in patients with CM-I. Error bars indicate ± 2 standard error of the mean. Red color indicates cognitive performance that was significantly different from the population mean of 0. CM-I patients show decreased scores in CFT, both copy and delayed recall conditions, as well as GPT in the dominant hand. CFT - Recall, CFT- Copy, TMT-A and TMT-B z scores were adjusted for performance on GPT dominant.

TABLE 4.

Neuropsychology Test Performance in the Total Preoperative CM-I Cohort Compared to Normative Data and Adjusted for Motor Function Using GPT

Test	n	z-score	P-value
CFT - copy	26	−1.11	.001
CFT – delayed recall	15	−0.93	.002
GPT - dominant	21	−1.35	.025
GPT - nondominant	21	−1.38	.040
TMT - B	23	0.02	.959
TMT - A	23	0.15	.578
AVLT - delayed recall	25	−0.40	.102
COWA	25	−0.22	.331
AVLT - immediate recall	25	−0.42	.106
WRAT - reading	18	−0.32	.619
Stroop - word	21	−0.18	.490
WAIS - vocabulary	14	−0.43	.096
WAIS - digit span	24	−0.09	.73
WAIS - similarities	23	−0.13	.78
Stroop - color	21	0.20	.513
WAIS - block design	19	−0.05	.835
WRAT - spelling	14	−0.10	.614
WMS - spatial span	16	0.27	.186
JOLO	10	0.16	.804
Stroop - color/word	21	0.34	.165
Stroop - interference	21	0.21	.236
WAIS - matrix	21	0.21	.236
WAIS - coding	23	0.28	.174

Boldface type indicates statistical significance where the P-value < Benjamini-Hochberg (i/m)*Q value (I = rank, m = total tests, Q = false discovery rate), as derived from the one-sample t-test comparing the mean Z-score deviation from 0. A negative z-score indicate lower performance compared to normative data.

Postop Data compared to Preop Data

Paired samples t-tests showed no significant changes in performance following surgical intervention (Table 5). The pre-post cohort profiles at both time points are shown in Figure 3.

TABLE 5.

Mean Z-Scores of the Neuropsychology and Psychology Tests Before and After Surgery for the Paired-Samples Analysis

Assessment	n	Preoperative mean (SD)	Postoperative mean (SD)	Mean difference
WAIS – similarities	15	0.09 (0.87)	0.11 (0.77)	−0.02
WAIS – vocabulary	4	−0.82 (1.12)	−0.75 (1.11)	0.07
WAIS – block design	10	−0.01 (1.16)	−0.03 (1.10)	−0.02
WAIS – matrix reasoning	13	0.28 (0.72)	0.52 (0.91)	0.24
WAIS – digit span	15	−0.03 (0.76)	0.25 (0.87)	0.27
WAIS - coding	16	0.54 (0.88)	0.66 (0.84)	0.12
WRAT - reading	5	−0.24 (0.84)	−0.12 (1.03)	0.12
WRAT - spelling	1	−0.10 (n/a)	0.10 (n/a)	0.20
AVLT – immediate recall	13	−0.17 (0.87)	0.04 (1.31)	0.22
AVLT – delayed recall	18	−0.01 (1.07)	−0.36 (1.14)	−0.35
CFT - copy	18	−0.78 (1.49)	−1.14 (1.73)	−0.36
CFT – recall	9	−0.50 (0.84)	−0.33 (1.18)	0.16
WMS – spatial span	7	−0.39 (1.11)	0.06 (0.83)	0.44
COWA	18	−0.27 (1.23)	−0.34 (1.19)	−0.08
JOLO	4	−0.80 (1.65)	−1.30 3.12)	−0.50
GPT -dominant	12	−0.69 (1.91)	−0.60 (1.41)	0.09
GPT - nondominant	12	−0.55 (1.37)	−0.68 (1.48)	−0.13
TMT - A	15	0.41 (0.95)	0.52 (0.95)	0.11
TMT - B	15	0.16 (1.32)	−0.41 (2.98)	−0.57
Stroop - word	17	−0.22 (0.97)	−0.49 (1.21)	0.27
Stroop - color	17	0.01 (1.36)	0.23 1.04)	0.22
Stroop – color/word	17	0.15 (0.96)	0.61 (1.19)	0.31
Stroop - interference	17	0.13 (0.79)	0.33 (0.91)	0.20
*BDI	14	17.57 (11.72)	14.50 (12.26)	3.07
*BAI	14	17.07 (11.63)	14.00 (11.23)	3.07
*BPI	6	5.41 (4.02)	3.41 (3.79)	2.01

*: Raw Scores.

Matched pair mean performance for patients who completed pre and postsurgery testing.

FIGURE 3.

Summary of A, preoperative and B, postoperative performances on a wide range of neuropsychological tests and psychological measures in patients with CM-I in the pre-post longitudinal cohort. Patients have relatively similar profiles with normalization of the CFT-recall score postoperatively.

Psychological Findings

Preoperatively, in the pre-post cohort of 19 patients, worsened depression scores strongly correlated with worse performance in Wechsler adult intelligence scale (WAIS)-similarities, wide range achievement test (WRAT)-Reading, WRAT-Spelling, Wechsler memory scale (WMS)-Spatial Span, judgment of line orientation test (JOLO), GPT-dom, and trail-making task B, as well as worsened anxiety and pain scores. Worsened depression scores weakly correlated with worse performance in WAIS-Digit Span, CFT-Copy, and Stroop Interference (all P < .05, Supplemental Table 1). Postoperatively, worsened mood scores strongly correlated with worse performance in CFT-Copy and anxiety and pain scores.

Differences in Performance by Surgical Approach

The relationship between approach selected (extradural vs intradural) and cognitive performance was also assessed. Prior to surgery, the extradural group had a statistically worse performance than the intradural group in WAIS-Similarities, WAIS-digit span, WRAT-reading, AVLT6, and BDI, with a trend in worse performance in WAIS-vocab, AVLT30, CFT-Copy, WMS-Spatial Span, BAI, and BPI (all P < .05, Supplemental Table 2).

Postoperatively, both the extradural and intradural groups performed similarly compared to before surgery in all cognitive and psychological tests. Thus, neither group showed a significant improvement over the other in any test.

Qualitative Patient Report

At the postoperative clinical and neuropsychology follow-up, qualitative assessment was obtained from the patients with a general open-ended question (ie “How are you doing” or “How are you feeling?”) No patient reported worsening of physical or cognitive symptoms after surgery, and the majority stated that they had improvement with representative comments included in Supplemental Table 3.

DISCUSSION

Lower Cognitive Performance in CM-I

Before surgery, patients with CM-I showed lower performance in tasks requiring visuospatial perception, construction, and memory (CFT copy and recall). These findings are consistent with previous reports that CM-I patients had worse preoperative performance in visuospatial and executive function tasks.^6,15,7,8,27 Additionally, performance on the Grooved Pegboard Test, a task of psychomotor function, also revealed poorer performance, although it did not correlate with syringomyelia. Psychomotor deficit did not account for the poor performance in CFT-Copy, a task that requires motor function to execute. Interestingly, syrinx did contribute to worsened CFT-Recall performance; this may be attributed to smaller sample size given the lack of differences in the CFT-Copy condition.

The finding that the extradural group showed worse preoperative cognitive performance than the intradural group across multiple measures is interesting considering the intradural group had worse preoperative radiographic features. However, worsened mood scores correlated with poorer performance in many of these neuropsychological tests. As such, the differences between approach groups seem to be at least in part explained by worse mood scores in the extradural group than the intradural group. These findings warrant further study.

CM-I Cognitive and Psychological Functioning After Surgery

Only a few studies have evaluated cognitive and psychological functioning in CM-I patients before and after surgery. These studies found some differences postoperatively in tasks of verbal learning, executive function, and language function or response inhibition.^8,9,28 Our study suggests that in an adult CM-I population, patients had improvement in their neurologic function without any cognitive or behavioral decline after major surgery. The subjective comments postoperatively, although informally collected, corroborate these findings. The lack of performance improvement postoperatively may reflect that CM-I is a congenital-developmental abnormality with largely fixed neuropsychological profiles. That is, once deficits occur in CM-I, not reversible with surgical decompression. This may be due to fixed cerebellar dysfunction as measured by preoperative resting-state fMRI in Chiari-I compared to controls.²⁹ Alternatively, neuropathic pain pathways may form and contribute to central sensitization of pain and ongoing cognitive interference¹⁵ despite our patients having improved headaches and pain postoperatively. What preoperative deficits exist in a younger CM-I population including children, if any, and how they change postoperatively, is an area of ongoing study.

Interestingly, mood correlated with many of the measures we assessed, including CFT-Copy which was below average compared to the normative data, a finding that has been noted previously in nonoperated CM-I.³⁰ While mood is known to impact performance on neuropsychological assessments and changes in mood may reflect expectancy effect,¹⁵ mood has also been implicated to be affected by abnormalities in cerebellar function.² This will require further study.

Possible Neuroanatomical Mechanisms

Increasing evidence indicates the cerebellum plays a critical role in cognitive function.¹ Historically, the cerebellum was thought to only modulate motor function. However, Schmahmann and Sherman² described disturbances of executive function, impaired spatial cognition, personality changes, and linguistic difficulties in a heterogenous group of patients with cerebellar disease. He coined these disturbances the cerebellar cognitive affective syndrome.³¹

Since this, studies examining cerebellar contributions to cognition have grown.³² Patients with right cerebellar lesions have deficits in memory, executive function, verbal fluency, and attention and affect, whereas patients with left sided lesions have deficits in visuospatial functions, a pattern that is reversed from cerebral lesions and attributed to the crossed cerebellar diaschisis phenomenon.³³ Furthermore, episodic memory is worse in patients with strokes in the posterior inferior cerebellar artery (PICA) distribution rather than in superior cerebellar artery (SCA) distribution,³⁴ with lesion mapping demonstrating predictable patterns of cognitive dysfunction depending on location.³⁵ Retrograde viral tracing found that the dentate nucleus has relevant functional domains, with the dorsal dentate projecting to motor and premotor areas and ventral dentate nucleus projecting to prefrontal and posterior parietal cortex.³⁶ Additionally, the dorsal dentate nucleus is primarily supplied by SCA perforators while the ventral dentate nucleus is supplied by PICA perforators.³⁷ These findings provide compelling evidence for the involvement of cerebellum in modulating cognitive tasks and a potential mechanism for dysfunction in disease states such as CM-I, where cerebellar and brainstem compression and distortion can disrupt afferent and efferent cerebellar networks.³⁸ Moreover, chronic vascular effects may occur due to compression of branches of the PICA between the tonsils¹⁸ or from distortion on other vasculature which may lead to scarring and gliosis that has been observed intraoperatively.^18,39

Implications

Our findings reported here may reflect cerebellum or brainstem compression that has altered the normal function of the cerebellar projections to cognitive networks. It is unknown whether these projections truly participate in the cognitive network or function to regulate and pace the normal operations of these networks. The latter would support Schmahmann's “dysmetria of thought” concept outlined in the cerebellar cognitive affective syndrome.^1,2

Some obvious questions arise from this work. What is the role of neuropsychological testing and performance at baseline and at follow-up, and can it be used to assess recovery? Does earlier diagnosis and intervention portend better cognitive outcomes and at what age? Do all types of posterior fossa decompressions result in similar postoperative profiles? What changes in resting-state fMRI may occur before and after surgery, and can this be used to predict or understand recovery or rehabilitation? Such questions warrant further study.

Limitations

Caution must be maintained regarding generalizing these results broadly in CM-I. The CM-I cohort here had classic Valsalva headaches and tonsillar descent beyond the 5 mm criteria with many exhibiting signs and symptoms of neurologic dysfunction on detailed neurologic exam. The generalizability to patients with an asymptomatic CM-I or patients with cerebellar ectopia <5 mm and Valsalva induced headaches is not well understood. These populations therefore ought to undergo proper study before counseling any such patients regarding cognitive findings or cognitive benefits of surgery.

Though 7 of the patients did not complete postoperative testing, this is a prospective study with continued active enrollment. As the goal was to be as comprehensive as possible, the testing battery was both broad and detailed requiring approximately 1.5 to 2 h to complete. Due to time constraints and/or subject tolerance, not every patient completed every measure. Practice effects can affect repeated neuropsychological measurements, and incorporation of appropriate normative test-retest data would help to quantify and interpret such effects. However, we took specific steps to reduce such effects, eg, by using a relatively long test-retest interval and using alternate forms when possible.

Patients who underwent testing were subject to selection bias, as in any surgical cohort. Although we found no differences between preoperative and postoperative performance, these analyses are subject to type I and type II error and warrant follow-up observations, particularly in light of the subjective patient comments that suggest improvement. While this study lacks a traditional control group per se, the conversion of the patient data to Z-scores for comparison to normative data of similar age, education, and sex in many ways accomplishes the goal of a control group. This represents a prospective preliminary study comprehensively assessing the neuropsychological performance preoperatively and postoperatively in a reasonably large cohort of CM-I patients; however it is not a clinical trial and the level of evidence should not be regarded as such.

CONCLUSION

This study defines the cognitive and psychological functioning of CM-I patients both preoperatively and at follow-up after surgical decompression. CM-I patients performed below normal on several tests and cognitive domains before surgical decompression. Moreover, CM-I patients demonstrate neurologic improvement without any decline in cognitive performance or psychological functioning following surgery. Of note, the current population within our prospective registry does not suggest an intradural decompression or an extradural decompressive surgery is better in regards to postoperative neuropsychological function. This represents the largest formal cognitive assessment of CM-I patients. If replicated and extended in larger numbers of CM-I patients, these results may prove critical to understanding cognitive and psychological functioning in CM-I patients.

Funding

This work was funded by NIH P50 MH0942581, NIH U01 NS103780, and the Kiwanis International Neuroscience Research Foundation.

Disclosures

The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.

Notes

Previously presented in part at the AANS Pediatric Section Meeting as an oral presentation, Dec 4th 2020 and at the Society for Neuroscience as a poster presentation, January 13 2021.

Supplemental Table 1. Approach Findings. Comparison of significant or near-significant differences between the intradural and extradural groups.

Supplemental Table 2. Mood Correlations. Association of significant or near significant findings between scores on the Beck Depression Inventory and various neuropsychological and psychological assessments.

Supplemental Table 3. Patient Subjective Comments.