The Utility of Quantitative Magnetic Resonance Angiography in the Assessment of Intracranial In-Stent Stenosis
Background and Purpose— Noninvasive screening for intracranial in-stent stenosis is often limited by artifact because of the stent or associated coils. We aimed to determine the utility of quantitative MRA (QMRA) as a screening tool for detecting intracranial in-stent stenosis.
Methods— We reviewed 14 patients who had intracranial stent placement with follow-up QMRA and conventional angiography at our institution. Socio-demographic, medical, clinical, and imaging data were abstracted from medical charts. A blinded interventional neurologist reviewed all angiograms for presence of >50% in-stent stenosis. We tested QMRA (mL/min) at varying thresholds as a predictor of angiographic results.
Results— Among 14 patients (mean age, 62 years; 12 female, 2 male), 13 patients had Neuroform stents placed for wide-neck cerebral aneurysms and 1 patient had a Wingspan stent placement for atherosclerotic stenosis. Lesions were located in the intracranial internal carotid artery in 57.2% (n=8), the middle cerebral artery in 14.3% (n=2), and vertebrobasilar arteries in 28.6% (n=4). On follow-up angiography, 2 patients (14.3%) had >50% in-stent stenosis on angiography. Time-of-flight MRA was nondiagnostic in each case because of artifact from the stent or coils. A >20% reduction in vessel-specific blood flow by QMRA was associated with presence of >50% in-stent stenosis on angiography (P=0.033). As a screening tool to predict >50% angiographic in-stent stenosis, the sensitivity, specificity, positive predictive value, and negative predictive value of QMRA were 100%, 92%, 67%, and 100%, respectively.
Conclusion— We found that QMRA is a promising screening tool to detect intracranial in-stent stenosis. Future prospective studies should focus on whether QMRA has a role in the detection of radiographic restenosis and prediction of clinical events.
In-stent stenosis or thrombosis is not an infrequent complication after intracranial stent placement.1,2 Because noninvasive imaging such as magnetic resonance angiography (MRA) and computerized tomography angiography are often degraded by artifact caused by the stent itself or associated coils, digital subtraction angiography (DSA) continues to be utilized routinely after stent placement. DSA, however, carries a 1% risk of neurological and non-neurological procedural complications.3 Thus, the development of a reliable noninvasive screening test would help reduce these risks and provide accurate detection of in-stent stenosis in this high-risk patient population.
Quantitative MRA (QMRA) uses traditional time-of-flight and phase-contrast MRI to visualize vascular anatomy and measure blood flow.4 Recent studies have established this technology’s utility in the measurement of arterial blood flow in various cerebrovascular conditions such vertebrobasilar insufficiency and carotid bypass surgery.5,6 A recent report also found it was useful in assessing blood flow after extracranial vertebral artery Wingspan stent placement.7
However, there are no studies of QMRA in the detection of intracranial in-stent stenosis. In a single-center retrospective study, we postulated that angiographic in-stent stenosis could be detected noninvasively with QMRA.
Materials and Methods
We conducted a retrospective review to identify all patients who met the following inclusion criteria: (1) stent placement for cerebral aneurysm or intracranial stenosis; (2) QMRA scan performed within 1 year after stent placement; (3) follow-up diagnostic angiography study performed within 1 month of the QMRA scan; and (4) high-quality QMRA images without motion artifact.
Fourteen patients met these criteria. Socio-demographic, medical, clinical, and imaging data were abstracted from medical charts. A blinded neuroradiologist (M.D.J.) reviewed the QMRA data and a blinded interventional neurologist (M.C.) reviewed angiograms for >50% in-stent stenosis. Noninvasive optimal vessel analysis (NOVA; Vassol, Inc) was performed as follows: using routine 3-dimensional time-of-flight MRA, the vessel segment distal to the stent was identified and a slice placed perpendicular to the vessel axis; volumetric flow rate was obtained using cine phase contrast MRI and compared to published data from controls.4,5 We assessed 2 cutoffs to define “low flow” by QMRA (>15% and >20% reduced flow). The Fisher exact test was used to determine the association between the 2 cutoffs for “low flow” by QMRA and >50% stenosis by DSA. Significance was set at P<0.05. We calculated the specificity, sensitivity, and positive and negative predictive values of QMRA for both cutoffs. Statistical analyses were performed using SPSS 14.0.
Median age of the cohort was 59.5 years and 12 were female. Thirteen patients had Neuroform stents placed for wide-neck cerebral aneurysms and 1 patient had a Wingspan stent placement for atherosclerotic stenosis. Lesions were located in the intracranial internal carotid artery in 57.2% (n=8), the middle cerebral artery in 14.3% (n=2), and vertebrobasilar arteries in 28.6% (n=4). Table 1 summarizes results of QMRA and DSA in each patient.
The median time from stent to follow-up QMRA was 192 days (interquartile range, 151–215 days). The median time between follow-up angiography and QMRA was 0.5 days (interquartile range, 0–11 days, with 6 completed on the same day). On follow-up diagnostic DSA, 2 patients (14.3%) had >50% in-stent stenosis. Time of flight MRA was nondiagnostic in each case because of artifact from the stent or coils (Figure). A >15% decrease in QMRA vessel flow was not significantly associated with >50% in-stent stenosis (P=0.110). However, presence of >20% reduction in QMRA vessel flow was significantly associated with stenosis by DSA (P=0.033).
Using the threshold of 15% decrease in vessel flow, sensitivity of QMRA to detect >50% angiographic in-stent stenosis was100%, with a specificity of 75%; the positive predictive value was 40% and negative predictive value was 100%. Using the threshold of 20% decrease in vessel flow, the sensitivity and negative predictive value remained 100%, whereas specificity and positive predictive value increased to 92% and 67%, respectively (Table 2).
We found that low blood flow as measured by QMRA at sites of intracranial stent placement was significantly associated with in-stent stenosis by catheter-based angiography. Using the 20% decrease in QMRA vessel flow, the sensitivity, specificity, and negative predictive value were excellent whereas the positive predictive value was fair. Given the potential clinical consequences of in-stent stenosis, a test with a high sensitivity and an acceptably low false-positive rate (33%) are desirable. In addition to identifying all true cases of in-stent stenosis, the high negative predictive value of the test might be used to defer follow-up DSA in those with normal QMRA results.
The limitations of this study include its small sample size and its retrospective nature. Our findings, therefore, require confirmation in larger prospective studies. Although contrast-enhanced MRA and computerized tomography angiography provide some anatomic visualization at sites of stent placement,8,9 QMRA has the advantage of also providing physiological data. However, the best value that defines abnormality by QMRA is yet to be determined. Comparing absolute or relative reductions to normative data are one approach, whereas intraindividual change over time may be another useful approach. An intraindividual ratio of vessel blood flow to total cranial flow may be the best measure in diseased states and allow adjustment for confounding factors such as hypotension or congestive heart failure and measurement of regional or collateral blood flow.4 We were unable to calculate this ratio for all patients because of incomplete measurements for all cranial arteries.
We found that QMRA is a promising screening test to detect intracranial in-stent stenosis with high sensitivity and specificity. Further prospective study is warranted.
Dr Lopes owns shares in and is on the advisory board for Vassol, Inc.
- Received April 5, 2008.
- Revision received May 23, 2008.
- Accepted June 11, 2008.
Levy EI, Turk AS, Albuquerque FC, Niemann DB, Aagaard-Kienitz B, Pride L, Purdy P, Welch B, Woo H, Rasmussen PA, Hopkins LN, Masaryk TJ, McDougall CG, Fiorella DJ. Wingspan in-stent restenosis and thrombosis: Incidence, clinical presentation, and management. Neurosurgery. 2007; 61: 644–650.
Zhao M, Amin-Hanjani S, Ruland S, Curcio AP, Ostergren L, Charbel FT. Regional cerebral blood flow using quantitative MR angiography. AJNR Am J Neuroradiol. 2007; 28: 1470–1473.
Amin-Hanjani S, Du X, Zhao M, Walsh K, Malisch TW, Charbel FT. Use of quantitative magnetic resonance angiography to stratify stroke risk in symptomatic vertebrobasilar disease. Stroke. 2005; 36: 1140–1145.
Turk AS, Rowley HA, Niemann DB, Fiorella D, Aagaard-Kienitz B, Pulfer K, Strother CM. CT angiographic appearance of in-stent restenosis of intracranial arteries treated with the Wingspan stent. AJNR Am J Neuroradiol. 2007; 28: 1752–1754.