Neurocognitive Performance After Cerebral Revascularization in Adult Moyamoya Disease
Background and Purpose—Cerebral revascularization using EC-IC bypass is widely used to treat moyamoya disease, but the effects of surgery on cognition are unknown. We compared performance on formal neurocognitive testing in adults with moyamoya disease before and after undergoing direct EC-IC bypass.
Methods—We performed a structured battery of 13 neurocognitive tests on 84 adults with moyamoya disease before and 6 months after EC-IC bypass. The results were analyzed using reliable change indices for each test, to minimize test–retest variability and practice effects.
Results—Twelve patients (14%) showed significant decline postoperatively, 9 patients (11%) improved, and 63 patients (75%) were unchanged. Similar results were obtained when the analysis was confined to those who underwent unilateral (33) or bilateral (51) revascularization.
Conclusions—The majority of patients showed neither significant decline nor improvement in neurocognitive performance after EC-IC bypass surgery. Uncomplicated EC-IC bypass seems not to be a risk factor for cognitive decline in this patient population.
Moyamoya disease (MM) is a rare cerebrovascular disorder of unknown cause, characterized by progressive occlusion of the internal carotid arteries and their branches, with the development of an abnormal vascular network in the areas of occlusion. The disease was first identified in Asian children but is now known to be more widespread and to affect also adults into the third and fourth decades of life.1,2 The clinical manifestations are predominantly ischemic—particularly in North America2—although hemorrhagic complications may also occur. Treatment usually involves surgical revascularization of the cerebrum, either directly by artery-to-artery anastomosis such as extracranial-to-intracranial (EC-IC) bypass or indirectly by arteriosynangiosis.3 Case series have suggested more favorable long-term outcome after surgery with respect to stroke or death.4,5 A prospective, randomized study in MM presenting with hemorrhage6 demonstrated efficacy of surgical revascularization compared with medical therapy.
A large study7 involving 329 treated patients indicated a low risk of serious complications from EC-IC bypass, defined in terms of mortality, surgical morbidity (new and persisting neurological deficits), and 5-year risk of perioperative or subsequent stroke or death. However, other kinds of potential complications—including effects on cognition and mood8—have yet to be systematically evaluated. We have previously described the incidence and severity of cognitive impairment in adults with MM.9,10 We report here the short-term cognitive outcomes after direct EC-IC bypass for MM in adults.
We studied 84 consecutive adults with MM who underwent EC-IC bypass by a single neurosurgeon at our institution and who did not meet any of the exclusion criteria. The average age of the patients was 38.3 years ±11.3 (SD). Fifty-eight were women. Their mean education was 15.2±2.2 years.
Patients were excluded for the following reasons: age <18 years, lack of proficiency in English (Wechsler Adult Intelligence Scale-Third Edition Vocabulary score <7), presence of major psychopathology (psychosis, major depression, or substance abuse), presence of a motor deficit that precluded administration of some tests, presence of neurological comorbidity (congenital malformations or traumatic brain injury), major surgical complication (stroke or hemorrhage), or low intelligence (Full-Scale IQ <70).
This study was approved by the Stanford Committee on the Use of Human Subjects in Research, which is the institutional review board of Stanford Medical Center.
Intelligence was evaluated with the Wechsler Adult Intelligence Scale-Third Edition.11 Memory was assessed with the California Verbal Learning Test-Second Edition, Total Words and Long Delay Free Recall components12 and the Wechsler Memory Test—Revised, Visual Reproduction Immediate and Delayed components.13 Tests of executive functioning included Letter and Category Fluency Tests14,15 and the Trail Making Test-Part B.16 Motor function was assessed with Grooved Pegboard.17 Expressive language was measured with the Boston Naming Test18 and speed of information processing with the Trail Making Test-Part A.16 The Beck Depression Inventory-II19 was used as a self-report measure of depression.
The neuropsychological tests were administered to all patients in accord with manual instructions. Patients were evaluated in a single 3-hour session. Patients were tested just before surgery and again ≈6 months post-surgery. The mean test–retest interval was 6.8±1.7 (SD) months.
To determine whether significant change in cognition occurred between pre- and postsurgical neuropsychological assessments, reliable change indices (RCIs) were obtained for each test from published research.20–25 The RCI is an indicator used to determine whether a change in test score with repeated administration of a test to an individual is large enough to indicate true change in the test score, as opposed to test–retest variability, measurement error, or practice effects. As applied, the RCI is a significance test of the statistical probability that the observed difference between 2 sequential scores can be attributed to variability/error/practice (ie, the null hypothesis). When this probability is low (P<0.05), one can attribute change to other factors of interest, such as the effect of an intervention (in this case, revascularization).
RCIs were chosen to optimally match the current study design regarding relevant factors such as test–retest interval and were applied to the pre/postsurgery change value for each test. On the basis of this procedure, each patient was defined as having had a decline in cognitive functioning, no change, or improvement, for each neuropsychological test administered. These results were aggregated across patients and tests. Individual patients were classified as decliners or improvers if they showed significant net decline or improvement, respectively, on ≥2 neuropsychological tests.26 Significant change in level of depression was defined in terms of a consensus RCI value for the Beck Depression Inventory-II.27
Table 1 presents the results of RCI analysis for each test. Overall, 8% of the pre/postneuropsychological test comparisons indicated a significant decline in cognition, and significant improvement was found in 6%. In 86% of comparisons, the observed difference did not exceed the RCI (no change).
Considering the individual patients, on cognitive testing 12 patients (14%) were defined as decliners, 9 patients (11%) as improvers, and 63 patients (75%) as unchanged. On the Beck Depression Inventory-II, 16 patients (19%) had significantly improved depression and 9 patients (11%) significantly worsened, with 59 patients (70%) unchanged.
Of the 84 patients studied, 33 underwent unilateral surgery and 51 bilateral. These 2 subgroups were analyzed to compare the effects of single versus double surgery, and the results are presented in Table 2. To determine whether the difference in the proportion of improvers in the 2 groups was significant, the decliners and unchanged groups were combined. Application of the Wald test for significance of the difference between 2 proportions28 indicated that the proportion of improvers in the 2 groups was not significantly different (Z=1.82; P>0.05).
The 33 unilateral cases were further analyzed to determine whether there was a difference between those patients who underwent unilateral surgery on the dominant left hemisphere (19) versus the right (14). The number of decliners, unchanged, and improvers in the left subgroup was 3, 13, and 3, respectively. On the right, the corresponding numbers were 3, 8, and 3, respectively. These numbers were considered too small for meaningful application of a significance test but seem to indicate no important difference between the sides.
This is the first large study to examine prospectively the changes in objective measures of brain function in adults with MM undergoing EC-IC bypass. Previous studies have shown that there is an increased incidence of cognitive impairment in adult MM, even in the absence of clinical or radiological evidence of stroke.9,10 Presumably, this represents a manifestation of subclinical hypoxic–ischemic injury to the brain or a manifestation of ongoing hypoperfusion. In the former case, no beneficial effect of revascularization on neurocognitive function could be expected, whereas in the latter case, the possibility of some improvement could at least theoretically be anticipated.
Earlier studies have attempted to evaluate the outcome from EC-IC bypass in terms of recurrent stroke or hemorrhage,4–6 but there has never been a randomized controlled trial of EC-IC bypass in this disorder when the manifestations are predominantly ischemic—as in the majority of our patients—nor a randomized controlled comparison of different methods of revascularization. The available evidence tends to suggest a diminished rate of stroke, hemorrhage, or death after successful bypass, but there is a nontrivial incidence of surgical complications including perioperative stroke or hemorrhage, seizures, and hyperperfusion syndrome.3–5 The net effect of EC-IC bypass on the brain may, therefore, be considered to represent the sum of multiple positive and negative potential influences, including (1) correction of hypoperfusion, (2) prevention of further ischemic injury, (3) surgical trauma and complications, and (4) altered intracranial and pericranial hemodynamics.
We have administered objective tests of neurocognitive function to assess whether the net effect of revascularization in these individuals was overall more positive or more negative. Our findings indicate that neither was the case: the majority of patients (≥70%) showed no significant change compared with their own preoperative performance. A small proportion of patients (14%) showed a decline in performance, and a similar small proportion (11%) showed apparent improvement. These findings do not confirm a previous report29 of postoperative cognitive improvement in a small number of patients with similar demographic. Our findings also do not confirm previous reports of cognitive improvement after EC-IC bypass in small groups of patients with atherosclerotic occlusive cerebrovascular disease involving the anterior circulation.30,31 It should be noted that all of these previous studies29–31 failed to account for practice effects, thereby biasing the results toward positivity and possibly generating spurious conclusions.
To better understand the observed outcomes, we retrospectively analyzed some clinical features of the patients who showed either decline or improvement on serial testing, according to whether they had undergone a unilateral or bilateral procedure (corresponding to the extent of the MM disease). This analysis indicated that when the intervention was unilateral versus bilateral, no difference in the degree of neurocognitive change was found. The proportion of decliners, unchanged, and improvers was the same in both groups, statistically. Similarly, when those undergoing only right or left unilateral intervention were compared, the number of decliners and improvers was again similar, with the majority showing no significant change.
Although it was beyond our capacity to fully analyze the complex clinical and radiological features in this large group of patients, we selected the preoperative modified Rankin Scale scores of those patients who showed improvement or decline, in order to explore whether the cognitive outcome was related to clinical severity. This analysis showed that the improved group had modified Rankin Scale scores of 1.44±0.73, whereas the decliners scored 2.08±1.00. The difference failed to reach statistical significance (P=0.12) in this probably underpowered comparison but suggests an hypothesis that deserves further investigation, namely, that more severely affected individuals fare less well cognitively after bypass surgery.
These findings suggest that there should be no expectation of cognitive improvement in adults with MM undergoing revascularization. More importantly, perhaps, they also show that—absent perioperative complications—EC-IC bypass per se does not seem to cause cognitive decline in these patients.
Test–retest comparisons such as we have performed can be complicated by problems such as practice effects, which can make interpretation of the findings difficult.20–23 To a certain extent, all patients may be expected to perform a little better on the second round of testing because of some learning that carries forward from the initial testing. Additionally, the test measures themselves are not perfectly reliable, the human organism is subject to intrinsic fluctuations, and the degree of practice effect varies across tests.25 For these reasons, we have analyzed our data using a methodology that specifically tries to accommodate such practice effects and ordinary test–retest variability in biological measurements. It is possible that the use of a different method of measuring change, or different criteria to define significant change, might conceivably yield different results.
Another potential weakness of our study is that we have not endeavored to document and assess all possible aspects of our patients’ health circumstances, such as imaging findings, measures of cerebral blood flow, and medication use, which might have affected their performance on neurocognitive testing, but which were beyond the scope of this study. We did note that many patients were taking antianxiety, antidepressant, and sedative medications, but we adopted the view that these were in most cases optimized for the particular patients, so that our findings likely reflect their real-world performance as manifested in their daily lives.
Another possible limitation of our study is that all patients were treated by a single surgeon using one surgical strategy (direct EC-IC bypass). Other approaches to revascularization might conceivably give different results. It is possible that a 6-month test–retest interval is too short for optimum evaluation of long-term neurocognitive functioning after major neurosurgical intervention in a chronic disorder that extends for decades. For this reason, we are currently in the process of restudying this same group of patients yet another time, at 3 years post-surgery, and hope to report these follow-up findings in the near future.
We performed detailed neurocognitive testing on 84 adults with MM, before and 6 months after cerebral revascularization by EC-IC bypass. Our findings indicate that the majority of patients did not show a significant change in neurocognitive performance after this intervention. Uncomplicated EC-IC bypass seems not to be a risk factor for cognitive decline in this patient population.
Sources of Funding
This study was funded through unrestricted support from the Department of Neurology and Neurological Sciences and Department of Neurosurgery (The Huber Family Moyamoya Fund; Stanley and Alexis Shin), Stanford University School of Medicine; and from Stanford Health Care.
Guest Editor for this article was Harold P. Adams, MD.
- Received November 9, 2016.
- Revision received March 1, 2017.
- Accepted March 7, 2017.
- © 2017 American Heart Association, Inc.
- Abla AA,
- Gandhoke G,
- Clark JC,
- Oppenlander ME,
- Velat GJ,
- Zabramski JM,
- et al
- Miyamoto S,
- Yoshimoto T,
- Hashimoto N,
- Okada Y,
- Tsuji I,
- Tominaga T,
- et al
- 11.↵Psychological Corp. Wechsler Adult Intelligence Scale. 3rd ed. San Antonio, TX: The Psychological Corp; 1997.
- Delis DC,
- Kramer JH,
- Kaplan E,
- Ober BA
- Wechsler D
- Delis DC,
- Kaplan E,
- Kramer JH
- Reitan RM,
- Wolfson D
- Reitan RM,
- Davison D
- Kaplan E,
- Goodglass H,
- Weintraub S
- Beck AT,
- Steer RA,
- Brown GK
- Woods SP,
- Childers M,
- Ellis RJ,
- Guaman S,
- Grant I,
- Heaton RK
- Baek HJ,
- Chung SY,
- Park MS,
- Kim SM,
- Park KS,
- Son HU