Published online before print September 21, 2007, 10.1148/radiol.2452061096
(Radiology 2007;245:532-540.)
© RSNA, 2007
Differentiation between Symptomatic Chiari I Malformation and Asymptomatic Tonsilar Ectopia by Using Cerebrospinal Fluid Flow Imaging: Initial Estimate of Imaging Accuracy1
Shawn K. Hofkes, MD,
Bermans J. Iskandar, MD,
Patrick A. Turski, MD,
Lindell R. Gentry, MD,
Jeremy B. McCue, MD, and
Victor M. Haughton, MD
1 From the Departments of Radiology (S.K.H., P.A.T., L.R.G., J.B.M., V.M.H.) and Neurosurgery (B.J.I.), University of Wisconsin School of Medicine and Public Health, Clinical Science Center-E3/311, 600 Highland Ave, Madison, WI 53792-3252. Received July 16, 2006; revision requested September 20; revision received December 18; accepted January 19, 2007; final version accepted March 15.
Address correspondence to V.M.H. (e-mail: vm.haughton{at}hosp.wisc.edu).
 |
ABSTRACT
|
|---|
Purpose: To determine the sensitivity and specificity of cerebrospinal fluid (CSF) flow imaging in distinguishing between patients with symptomatic Chiari I malformation and those with asymptomatic tonsilar ectopia by using a neurosurgeon's overall clinical determination as the reference standard.
Materials and Methods: The institutional review board of the University of Wisconsin Hospital and Clinics approved our HIPAA-compliant retrospective study and granted a waiver for informed consent. Seventeen patients (five male, 12 female; aged 4–43 years) with tonsils extending more than 5 mm below the foramen magnum were classified by the neurosurgeon as symptomatic for Chiari I malformation or asymptomatic for tonsilar ectopia. The CSF flow images of the two groups were read independently in blinded fashion by four neuroradiologists. Reader agreement was calculated as percentage of readings in each patient that agreed with the neurosurgeon's classification. Sensitivity and specificity were respectively calculated as the percentage of abnormal readings in symptomatic patients and the percentage of normal readings in asymptomatic patients.
Results: Of 17 patients, nine were classified by the neurosurgeon as symptomatic Chiari I malformation and eight as asymptomatic tonsilar ectopia. Agreement between pairs of readers was 63%–44%. For sagittal and transverse images, reader sensitivity for finding abnormal flow in symptomatic Chiari I malformation patients averaged 76% and specificity for normal flow in patients with asymptomatic tonsilar ectopia averaged 62%. The number of positive readings in the symptomatic patient group was significantly greater than that in the asymptomatic group (P < .02).
Conclusion: Readers detected an abnormal CSF flow pattern significantly more often in patients with symptomatic Chiari I malformation than in patients with asymptomatic tonsilar ectopia.
Supplemental material: http://radiology.rsnajnls.org/cgi/content/full/2452061096/DC1
© RSNA, 2007
|
INTRODUCTION
|
|---|
The Chiari I malformation, characterized by downward displacement of the cerebellar tonsils, may be associated with syringomyelia and/or a variety of neurologic signs and symptoms. While tonsilar ectopia can be identified and measured effectively with magnetic resonance (MR) imaging (1,2), the presence and magnitude of the ectopia do not correlate with the presence and severity of signs or symptoms. Up to 30% of individuals with more than 5 mm of tonsilar ectopia lack signs or symptoms that are typically attributed to the Chiari I malformation (3,4). Additionally, patients with minimal or no tonsilar ectopia may have symptoms that are similar to those found in patients with symptomatic Chiari I malformation. Symptomatic Chiari I patients may benefit from surgical decompression (5,6) while patients without cerebrospinal fluid (CSF) flow abnormalities may not benefit.
Phase-contrast magnetic resonance (MR) imaging is often used to evaluate the effect of tonsilar ectopia on CSF flow and to assess its possible clinical importance. The validity of the phase-contrast MR CSF flow study has been investigated by comparing CSF flow patterns in Chiari I patients with those in healthy patients (7–16). Although quantitative methods of assessing CSF flow have been reported, nonquantitative visual assessment of CSF flow has been used in clinical studies and in many published reports (17,18). In those reports, reduced flow in the posterior subarachnoid space in phase-contrast MR images is generally considered a sign of abnormal CSF flow (19). The accuracy of this and other imaging features to distinguish abnormal flow has not been adequately assessed, in our opinion.
To our knowledge, no published studies compare the CSF flow patterns in patients who are symptomatic for Chiari I malformation with those who are asymptomatic for tonsilar ectopia. Since the main objective of CSF flow imaging is to select patients for surgical treatment, the differences between asymptomatic and symptomatic posterior fossa imaging findings warrant investigation. Thus, the purpose of our study was to retrospectively determine the sensitivity and specificity of CSF flow imaging in distinguishing patients with symptomatic Chiari I malformation from those with asymptomatic tonsilar ectopia, using the neurosurgeon's overall clinical determination as the reference standard.
|
MATERIALS AND METHODS
|
|---|
The institutional review board of the University of Wisconsin Hospital and Clinics approved our Health Insurance Portability and Accountability–compliant retrospective study and granted a waiver of informed consent.
Patients and Reference Standard
The medical records of patients treated in the Neurosurgery Specialty Clinic were reviewed to identify patients as having symptomatic Chiari I malformation or asymptomatic tonsilar ectopia (B.J.I.). In this clinic, patients were evaluated if tonsilar ectopia was discovered at computed tomography or MR, for example, for head trauma or for a known Chiari I malformation that was being considered for surgical management. Each patient evaluated in this clinic gave a thorough history and underwent physical and neurologic examinations, on the basis of which the diagnosis and classification were made. Patients were classified as symptomatic or asymptomatic before CSF flow studies were obtained.
To be classified as symptomatic for Chiari I malformation, the patient had to exhibit a combination of the following signs and symptoms: cervical syringomyelia, occipital headaches, Valsalva maneuver–induced headaches, back pain, dysphagia, dysarthria, truncal ataxia, nystagmus, lower cranial nerve dysfunction, motor or sensory deficit in the upper or lower extremities (particularly the hands), abnormal reflexes, or scoliosis (20).
For the symptomatic Chiari I group, an additional inclusion criterion was a positive clinical response to subsequent craniocervical decompression. Inclusion criteria for both groups were tonsils that extended 5 mm or more below the foramen magnum on a sagittal MR image with the neck in neutral position, findings during history and physical examination that permitted unambiguous classification as asymptomatic tonsilar ectopia or symptomatic Chiari I malformation, and a phase-contrast MR CSF flow study with images obtained in sagittal and transverse planes with the neck in neutral, flexed, and extended positions.
Seventeen patients who met the inclusion criteria were identified, including four male and 13 female patients ranging in age from 4 to 43 years (mean age, 19 years).
Phase-Contrast MR CSF Flow Imaging
Transverse and sagittal phase-contrast MR images were acquired in each patient with similar techniques. The images were electrocardiographically gated to the cardiac cycle. The plane of acquisition was selected according to a standard protocol. For sagittal images, the plane that optimally demonstrated the dens was chosen. For transverse images, a plane transverse to the axis of the spinal canal immediately below the tonsilar tips was selected. For the phase-contrast flow measurement performed with phase-contrast MR, the acquisition parameters were flip angle, 20°; repetition time msec/echo time msec, 20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; and encoding velocity, 10 cm · sec–1. For each patient, the CSF flow images were acquired, in order, with the patient's neck in neutral, flexed, and extended positions. Flow images were stored on the institution's picture archiving and communication system.
Blinded Readings of CSF Flow Images
Four readers independently and retrospectively reviewed the images. Two of the readers (P.A.T., L.R.G., each with 10 years experience reviewing CSF flow studies) were neuroradiologists, and two (J.B.M. and another fellow, each with experience reviewing 10–20 CSF flow studies under staff supervision) were neuroradiology fellows. Readers were initially shown two flow studies (Figs 1, 2). Figure 1 shows little difference between the flow in the anterior and posterior subarachnoid spaces in a patient with Chiari I malformation. Figure 2 shows flow only in the anterior subarachnoid space in another patient with Chiari I malformation. These flow studies were not included in the blinded readings. Each reader classified Figure 1 as normal flow and Figure 2 as abnormal.
View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a: (a) Sagittal T2-weighted and (b, c) sagittal and (d, e) transverse gated phase-contrast MR images in Chiari I patient show little disparity in CSF flow anterior and posterior to the spinal cord during (b, d) diastolic and (c, e) systolic phases. Readers classified study as normal at the foramen magnum (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1) Figure 1 was not included in reader agreement study.
|
|
View larger version (95K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b: (a) Sagittal T2-weighted and (b, c) sagittal and (d, e) transverse gated phase-contrast MR images in Chiari I patient show little disparity in CSF flow anterior and posterior to the spinal cord during (b, d) diastolic and (c, e) systolic phases. Readers classified study as normal at the foramen magnum (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1) Figure 1 was not included in reader agreement study.
|
|
View larger version (102K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1c: (a) Sagittal T2-weighted and (b, c) sagittal and (d, e) transverse gated phase-contrast MR images in Chiari I patient show little disparity in CSF flow anterior and posterior to the spinal cord during (b, d) diastolic and (c, e) systolic phases. Readers classified study as normal at the foramen magnum (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1) Figure 1 was not included in reader agreement study.
|
|
View larger version (66K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1d: (a) Sagittal T2-weighted and (b, c) sagittal and (d, e) transverse gated phase-contrast MR images in Chiari I patient show little disparity in CSF flow anterior and posterior to the spinal cord during (b, d) diastolic and (c, e) systolic phases. Readers classified study as normal at the foramen magnum (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1) Figure 1 was not included in reader agreement study.
|
|
View larger version (85K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1e: (a) Sagittal T2-weighted and (b, c) sagittal and (d, e) transverse gated phase-contrast MR images in Chiari I patient show little disparity in CSF flow anterior and posterior to the spinal cord during (b, d) diastolic and (c, e) systolic phases. Readers classified study as normal at the foramen magnum (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1) Figure 1 was not included in reader agreement study.
|
|
View larger version (85K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a: (a, b) Sagittal and (c, d) transverse phase-contrast MR images of Chiari I patient show restricted CSF flow in posterior subarachnoid space during (a, d) systolic and (b, c) diastolic flow. Readers classified study as abnormal (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm/sec). Figure 2 was not included in reader agreement study.
|
|
View larger version (86K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b: (a, b) Sagittal and (c, d) transverse phase-contrast MR images of Chiari I patient show restricted CSF flow in posterior subarachnoid space during (a, d) systolic and (b, c) diastolic flow. Readers classified study as abnormal (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm/sec). Figure 2 was not included in reader agreement study.
|
|
View larger version (125K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2c: (a, b) Sagittal and (c, d) transverse phase-contrast MR images of Chiari I patient show restricted CSF flow in posterior subarachnoid space during (a, d) systolic and (b, c) diastolic flow. Readers classified study as abnormal (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm/sec). Figure 2 was not included in reader agreement study.
|
|
View larger version (113K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2d: (a, b) Sagittal and (c, d) transverse phase-contrast MR images of Chiari I patient show restricted CSF flow in posterior subarachnoid space during (a, d) systolic and (b, c) diastolic flow. Readers classified study as abnormal (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm/sec). Figure 2 was not included in reader agreement study.
|
|
For the blinded readings, CSF flow images were retrieved and displayed by one investigator (S.K.H.) on the picture archiving and communication system. The readers were not provided specific criteria for diagnosing the abnormal flow. The four readers knew they would review images in patients who were either symptomatic for a Chiari I malformation or asymptomatic for tonsilar ectopia. The investigator varied the order in which readers saw the patient images and the sequence in which they reviewed the transverse and sagittal images. The readers were not given the patient's identification or clinical information or the neurosurgeon's final classification, and they were not shown anatomic head or cervical spine images.
The readers were first asked to determine if the images were technically adequate for interpretation. If the reader classified the image as inadequate for interpretation, no further information was requested. They were then asked to classify the study as showing normal flow (flow pattern not likely associated with symptoms) or abnormal flow (flow pattern probably associated with symptoms). They were asked to classify the CSF flow abnormality on a scale of 0–3: a score of 0 = no abnormality, a score of 1 = questionable abnormality, a score of 2 = definite abnormality, and a score of 3 = extreme abnormality. The average severity was calculated for the symptomatic and asymptomatic groups. Each reader was asked to describe the findings that were considered diagnostic of a craniocervical pathophysiologic CSF flow problem.
At a subsequent session, readers were asked to review, in the same manner, another set of images in the second plane. The readers were not given the patient's identification, clinical information, or information from the first reading. The same questions were asked each time. Each reader reviewed transverse images first in some patients and sagittal images first in others. The time elapsed between readings ranged from 4 to 8 weeks. If the reader diagnosed the CSF flow as abnormal, he or she was also asked to specify in which position (neutral, flexion, or extension) the most abnormal CSF flow was demonstrated.
Statistical Analysis
For each reader, sensitivity was calculated as the percentage of patients classified as abnormal in the symptomatic Chiari I patients; specificity, as the percentage of patients classified as normal among the asymptomatic tonsilar ectopia patients. The sensitivities and specificities for all readers were averaged by one investigator (S.K.H.).
Reader agreement was calculated for each patient as the percentage of readings agreeing with the reference standard. Overall agreement was calculated as the average of all readers and was calculated separately for the transverse and sagittal images. Overall agreement was also calculated separately for the pairs of more experienced and less experienced readers.
Statistical significance was measured by using a two-tailed t test with a P value of .05 indicating a significant difference. An Excel (Microsoft, Redmond, Wash) spreadsheet was used to calculate the P value (V.M.H.). The difference between the average abnormal interpretations by the readers in the symptomatic and asymptomatic groups was tested for significance. The average severity of flow abnormality was calculated for the symptomatic and asymptomatic groups, and the differences were tested for significance, as were differences in the sensitivity and specificity for more and less experienced readers. Agreement among all readers for the sagittal and transverse images was calculated and tested for significance. Agreement between the more and less experienced readers was also calculated and tested for significance.
|
RESULTS
|
|---|
Patients
Nine of the patients (two male, seven female; aged 6–30 years; mean, 14 years) were classified as having symptomatic Chiari I malformation; eight patients (two male, six female; aged 16–43 years; mean, 26 years), as having asymptomatic tonsilar ectopia. All nine symptomatic patients showed marked improvement from craniocervical decompression, which was performed subsequent to the CSF flow studies. Two underwent surgical decompression twice before experiencing clinical improvement, two had undergone previous fusion procedures to treat scoliosis, four had cervical syringomyelia, and one had neither of the foregoing.
Readings of CSF Flow Images
With the exception of three flow studies in the transverse plane considered suboptimal by one reader (J.B.M.), readers scored each of the transverse and sagittal flow studies (Table). For the 68 interpretations of the 17 sagittal flow studies, 41 were abnormal and 27 were normal flow patterns (Fig 3). Of 41 abnormal flow patterns, 27 were in patients assigned by the neurosurgeon to the symptomatic Chiari I group and 14 were in patients assigned to the asymptomatic tonsilar ectopia group. Of the 27 normal readings, nine were in patients assigned by the neurosurgeon to the symptomatic group and 18 were in patients assigned to the asymptomatic group.
In the symptomatic group, readers interpreted 67%–100% of sagittal and 44%–100% of transverse flow studies as abnormal in the symptomatic patients. The readers interpreted 50%–63% of sagittal and 43%–100% of transverse flow studies as normal in the asymptomatic group of patients. On average, the readers classified 75% of sagittal and 72% of transverse flow studies as abnormal in the symptomatic patients and 56% of sagittal and 68% of transverse flow studies as normal in the asymptomatic patients. The difference in the number of positive readings in the symptomatic compared with the asymptomatic group was significant (P = .02 for sagittal and P = .002 for transverse images).
The most common finding used by readers to determine normal CSF flow was moderate preponderance of flow anterior to the spinal cord. For patients classified as having abnormal flow, reduced flow in the posterior subarachnoid space, hyperactive flow, and flow jets were noted. Absence of posterior flow and the presence of anterior flow jets in the subarachnoid space were designated in patients falsely classified as having abnormal flow by readers (Figs 4, 5; Movies [http://radiology.rsnajnls.org/cgi/content/full/2452061096/DC1]).
View larger version (109K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a: (a, b) Sagittal phase-contrast MR images in patient with symptomatic Chiari I malformation. Images obtained during (a) diastolic and (b) systolic phases of CSF flow show much greater flow anterior than posterior to the spinal cord. (c–g) Transverse images show preponderance of CSF flow anterior to spinal cord, large jets in anterior subarachnoid space, and appearance of simultaneous flow in (d) superior-to-inferior and (e) inferior-to-superior directions. Three readers classified a and b as abnormal; all readers classified c–g as abnormal (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1).
|
|
View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b: (a, b) Sagittal phase-contrast MR images in patient with symptomatic Chiari I malformation. Images obtained during (a) diastolic and (b) systolic phases of CSF flow show much greater flow anterior than posterior to the spinal cord. (c–g) Transverse images show preponderance of CSF flow anterior to spinal cord, large jets in anterior subarachnoid space, and appearance of simultaneous flow in (d) superior-to-inferior and (e) inferior-to-superior directions. Three readers classified a and b as abnormal; all readers classified c–g as abnormal (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1).
|
|
View larger version (110K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4c: (a, b) Sagittal phase-contrast MR images in patient with symptomatic Chiari I malformation. Images obtained during (a) diastolic and (b) systolic phases of CSF flow show much greater flow anterior than posterior to the spinal cord. (c–g) Transverse images show preponderance of CSF flow anterior to spinal cord, large jets in anterior subarachnoid space, and appearance of simultaneous flow in (d) superior-to-inferior and (e) inferior-to-superior directions. Three readers classified a and b as abnormal; all readers classified c–g as abnormal (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1).
|
|
View larger version (104K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4d: (a, b) Sagittal phase-contrast MR images in patient with symptomatic Chiari I malformation. Images obtained during (a) diastolic and (b) systolic phases of CSF flow show much greater flow anterior than posterior to the spinal cord. (c–g) Transverse images show preponderance of CSF flow anterior to spinal cord, large jets in anterior subarachnoid space, and appearance of simultaneous flow in (d) superior-to-inferior and (e) inferior-to-superior directions. Three readers classified a and b as abnormal; all readers classified c–g as abnormal (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1).
|
|
View larger version (104K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4e: (a, b) Sagittal phase-contrast MR images in patient with symptomatic Chiari I malformation. Images obtained during (a) diastolic and (b) systolic phases of CSF flow show much greater flow anterior than posterior to the spinal cord. (c–g) Transverse images show preponderance of CSF flow anterior to spinal cord, large jets in anterior subarachnoid space, and appearance of simultaneous flow in (d) superior-to-inferior and (e) inferior-to-superior directions. Three readers classified a and b as abnormal; all readers classified c–g as abnormal (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1).
|
|
View larger version (101K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4f: (a, b) Sagittal phase-contrast MR images in patient with symptomatic Chiari I malformation. Images obtained during (a) diastolic and (b) systolic phases of CSF flow show much greater flow anterior than posterior to the spinal cord. (c–g) Transverse images show preponderance of CSF flow anterior to spinal cord, large jets in anterior subarachnoid space, and appearance of simultaneous flow in (d) superior-to-inferior and (e) inferior-to-superior directions. Three readers classified a and b as abnormal; all readers classified c–g as abnormal (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1).
|
|
View larger version (106K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4g: (a, b) Sagittal phase-contrast MR images in patient with symptomatic Chiari I malformation. Images obtained during (a) diastolic and (b) systolic phases of CSF flow show much greater flow anterior than posterior to the spinal cord. (c–g) Transverse images show preponderance of CSF flow anterior to spinal cord, large jets in anterior subarachnoid space, and appearance of simultaneous flow in (d) superior-to-inferior and (e) inferior-to-superior directions. Three readers classified a and b as abnormal; all readers classified c–g as abnormal (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1).
|
|
View larger version (103K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a: (a) Sagittal T2-weighted and (b, c) transverse and (d, e) sagittal phase-contrast flow MR images in patient with asymptomatic Chiari I malformation. Two readers interpreted images as having normal flow pattern; two other readers, as having abnormal flow patterns (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1). Diastolic flow is shown in b and e; systolic flow, in c and d.
|
|
View larger version (62K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b: (a) Sagittal T2-weighted and (b, c) transverse and (d, e) sagittal phase-contrast flow MR images in patient with asymptomatic Chiari I malformation. Two readers interpreted images as having normal flow pattern; two other readers, as having abnormal flow patterns (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1). Diastolic flow is shown in b and e; systolic flow, in c and d.
|
|
View larger version (63K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5c: (a) Sagittal T2-weighted and (b, c) transverse and (d, e) sagittal phase-contrast flow MR images in patient with asymptomatic Chiari I malformation. Two readers interpreted images as having normal flow pattern; two other readers, as having abnormal flow patterns (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1). Diastolic flow is shown in b and e; systolic flow, in c and d.
|
|
View larger version (96K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5d: (a) Sagittal T2-weighted and (b, c) transverse and (d, e) sagittal phase-contrast flow MR images in patient with asymptomatic Chiari I malformation. Two readers interpreted images as having normal flow pattern; two other readers, as having abnormal flow patterns (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1). Diastolic flow is shown in b and e; systolic flow, in c and d.
|
|
View larger version (86K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5e: (a) Sagittal T2-weighted and (b, c) transverse and (d, e) sagittal phase-contrast flow MR images in patient with asymptomatic Chiari I malformation. Two readers interpreted images as having normal flow pattern; two other readers, as having abnormal flow patterns (20/5; section thickness, 5 mm; field of view, 180 mm; matrix, 256 x 256; encoding velocity, 10 cm · sec–1). Diastolic flow is shown in b and e; systolic flow, in c and d.
|
|
The ratings of CSF flow abnormality ranged from 0 to 3. For the reader with the highest number of correct scores, the average severity score was 1.6 for the symptomatic Chiari I malformation patients and 0.8 for the asymptomatic tonsilar ectopia patients, indicating a significant difference (P = .004). For the reader with the lowest number of correct scores, the average severity score was 1.2 for the symptomatic group and 0.7 for the asymptomatic group, indicating a significant difference (P = .048). Images were identified as having definite or severe flow abnormality in 55% of symptomatic patients and as questionable in the remainder. Images were identified as having no or questionable abnormality in 75% of asymptomatic patients and as a severe abnormality in the remainder.
Readers found that CSF flow was more abnormal when the neck was in flexion in 46% of patients and when it was in neutral or extension positions in 16% of patients. Position did not appear to affect the flow pattern substantially in the remainder of the patients.
Reader Agreement
Agreement with the neurosurgeon's diagnosis ranged from 25%–100% for individual patients. Agreement averaged 67% on sagittal and 70% on transverse images, but the difference was not significant (P = .166). Agreement averaged 74% for more experienced readers and 62% for less experienced readers, but the difference was not significant (P = .096).
Sensitivity and Specificity
For sagittal and transverse images, the respective sensitivity for the four readers to diagnose abnormal flow in the symptomatic patients averaged 75% and 72% and the respective specificity to identify the asymptomatic patients averaged 56% and 68%. The combination of sensitivity and specificity between more and less experienced readers was not significantly different (P = .08).
|
DISCUSSION
|
|---|
CSF flow patterns were significantly different in the symptomatic Chiari I patients compared with the asymptomatic tonsilar ectopia group. Sensitivity ranged from 64% to 88% for readers diagnosing a symptomatic Chiari I malformation in our study, with a small trend toward better accuracy for more experienced readers. Specificity ranged from 59% to 64% for the lack of visible flow abnormalities in the group with no symptoms attributable to tonsilar ectopia.
Reader agreement in differentiating between symptomatic and asymptomatic CSF flow patterns in patients with Chiari I malformation and tonsilar ectopia, respectively, has not been reported to our knowledge, while differences in CSF flow patterns between Chiari I patients and healthy patients have been extensively reported (7–11,14,16). The sensitivity in our study approached 75%, a level often reported for accuracy in imaging (21). The sensitivity and specificity in our study may be lower than in published studies because abnormal tonsil location was present in both of our groups. We calculated reader agreement in comparison with the actual diagnosis, rather than compute the number of pairs of readers who agreed and disagreed, which may be the more common method.
Our study included a small number of patients because the enrollment criteria were strict. Thus, the significance must be considered conservatively. One of the limitations of our study was related to the lack of universally accepted criteria for diagnosing the symptomatic Chiari I malformation. No single clinical finding or group of findings differentiates between symptomatic Chiari I and asymptomatic tonsilar ectopia patients (4,20). Even the characteristic symptom of Valsalva maneuver–induced occipital headaches is absent in about 25% of patients with Chiari I malformation (20).
To enable the classification of patients as symptomatic or asymptomatic, the judgment of an experienced neurosurgeon was used as a reference standard (3,4). Since the neurosurgeon in our study had access to all information on the CSF flow studies, the effect of this knowledge on his classification cannot be excluded. The neurosurgeon differentiated the patients in symptomatic and asymptomatic groups entirely on the basis of the clinical findings and requested flow studies only after clinical assessment and review of static MR images.
The criteria by which readers distinguished abnormal flow patterns included decreased flow posterior to the spinal cord and flow jets in the foramen magnum. One source of reader bias may have been the presence of a syrinx in four of the symptomatic patients, although the syrinx was not readily visible on the flow images that were provided for interpretation. Whether the identification of abnormal CSF flow requires one section, multiple sections, or a volume acquisition remains undetermined. Whether sagittal or transverse images better demonstrate abnormal flow needs further evaluation. Our study outcome suggests that quantitative techniques warrant further evaluation.
Our study has shown that nonquantitative techniques can distinguish symptomatic and asymptomatic Chiari I malformations in a majority of patients. However, with the present nonquantitative image interpretation criteria, reader agreement, sensitivity, and specificity do not achieve the same levels as do those of other imaging studies. Further study is warranted, particularly for quantitative techniques.
In summary, readers detected differences in CSF flow patterns between symptomatic Chiari I malformation and asymptomatic tonsilar ectopia patients. For sagittal and transverse images, average reader sensitivity was 76% for finding abnormal flow in symptomatic patients and average specificity was 62% for finding normal flow in patients with asymptomatic tonsilar ectopia.
|
ADVANCE IN KNOWLEDGE
|
|---|
- Readings of cerebrospinal fluid (CSF) flow images based on current criteria have an average sensitivity of 76% and an average specificity of 62% in distinguishing between symptomatic Chiari I malformations and asymptomatic tonsilar ectopia.
|
IMPLICATION FOR PATIENT CARE
|
|---|
- Phase-contrast MR flow studies have a role in distinguishing between abnormal CSF flow patterns in the evaluation of patients with symptomatic Chiari I malformation and asymptomatic tonsilar ectopia to determine which may benefit from surgical intervention.
|
ACKNOWLEDGMENTS
|
|---|
Jason Fine, PhD, Department of Biostatistics and Medical Informatics, University of Wisconsin School of Medicine and Public Health, provided suggestions on the statistical protocols.
|
FOOTNOTES
|
|---|
Abbreviations: CSF = cerebrospinal fluid
Guarantors of integrity of entire study, J.B.M., V.M.H.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, S.K.H., B.J.I., V.M.H.; clinical studies, S.K.H., B.J.I., P.A.T., L.R.G., J.B.M.; statistical analysis, S.K.H., V.M.H.; and manuscript editing, S.K.H., B.J.I., L.R.G., J.B.M., V.M.H.
Authors stated no financial relationship to disclose.
|
References
|
|---|
- Barkovich AJ, Wippold FJ, Sherman JL, et al. Significance of cerebellar tonsilar position on MR. AJNR Am J Neuroradiol 1986;7:795–799. [Abstract]
- Elster AD, Chen MY. Chiari I malformations: clinical and radiologic reappraisal. Radiology 1992;183:347–353. [Abstract/Free Full Text]
- Meadows J, Kraut M, Guarnieri M, et al. Asymptomatic Chiari type I malformations identified on magnetic resonance imaging. J Neurosurg 2000;92:920–926. [CrossRef][Medline]
- Milhorat TH, Chou MW, Trinidad EM, et al. Chiari I malformation redefined: clinical and radiographic findings for 364 symptomatic patients. Neurosurgery 1999;44:1005–1017. [CrossRef][Medline]
- Iskandar BJ, Hedlund GL, Grabb PA, et al. The resolution of syringohydromyelia without hindbrain herniation after posterior fossa decompression. J Neurosurg 1998;89:212–216. [Medline]
- Tubbs RS, Elton S, Grabb P, Dockery SE, Bartolucci AA, Oakes WJ. Analysis of the posterior fossa in children with the Chiari 0 malformation. Neurosurgery 2001;48:1050–1054. [CrossRef][Medline]
- Heiss JD, Patronas N, DeVroom HL, et al. Elucidating the pathophysiology of syringomyelia. J Neurosurg 1999;91:553–562. [Medline]
- Oldfield EH, Muraszko K, Shawker TH, et al. Pathophysiology of syringomyelia associate with Chiari I malformation of the cerebellar tonsils: implications for diagnosis and treatment. J Neurosurg 1994;80:3–15. [Medline]
- Wolpert SM, Bhadelia RA, Bogdan AR, et al. Chiari I malformations: assessment with phase-contrast velocity MR. AJNR Am J Neuroradiol 1994;15:1299–1308. [Abstract]
- Hofmann E, Warmuth-Metz M, Bendzus M, Solymosi L. Phase-contrast MR imaging of the cervical CSF and spinal cord: volumetric motion analysis in patients with Chiari I malformation. AJNR Am J Neuroradiol 2000;21:151–158. [Abstract/Free Full Text]
- Armonda RA, Citrin CM, Foley KT, et al. Quantitative cinemode magnetic resonance imaging of Chiari I malformations: an analysis of cerebrospinal fluid dynamics. Neurosurgery 1994;35:214–224. [Medline]
- Levy LM. Magnetic resonance techniques in the evaluation of cerebrospinal fluid flow and spinal cord motion in neurologic disorders. In: Latchaw R, Kucharczyk J, Moseley M, eds. Imaging the nervous system: diagnostic and therapeutic applications. Philadelphia, Pa: Elsevier Mosby, 2004; 1093.
- Levy LM. MR imaging of cerebrospinal fluid flow and spinal cord motion in neurologic disorders of the spine. Magn Reson Imaging Clin N Am 1999;7:573–587. [Medline]
- Quigley MF, Iskandar B, Quigley ME, Nicosia M, Haughton V. Cerebrospinal fluid flow in foramen magnum: temporal and spatial patterns at MR imaging in volunteers and in patients with Chiari I malformation. Radiology 2004;232:229–236. [Abstract/Free Full Text]
- Greitz D, Franck A, Nordell B. On the pulsatile nature of intracranial and spinal CSF-circulation demonstrated by MR imaging. Acta Radiol 1993;34:321–328. [Medline]
- Bhadelia RA, Bogdan AR, Wolpert SM. Cerebrospinal fluid flow waveforms: effect of altered cranial venous outflow—a phase-contrast MR flow imaging study. Neuroradiology 1998;40:283–292. [CrossRef][Medline]
- McGirt MJ, Nimjee SM, Floyd J, Bulsara KR, George TM. Correlation of cerebrospinal fluid flow dynamics and headache in Chiari I malformation. Neurosurgery 2005;56:716–721. [CrossRef][Medline]
- Ventureyra EC, Aziz HA, Vassilyadi M. The role of cine flow MRI in children with Chiari I malformation. Childs Nerv Syst 2003;19:109–113. [Medline]
- Barkovitch JA, Pediatric neuroradiology, 4th ed. Philadelphia, Pa: Lippincott Williams & Wilkins, 2005; 374–378.
- Iskandar BJ, Oakes WJ. Chiari malformations. In: Albright L, Pollack I, Adelson D, eds. Principles and practice of pediatric neurosurgery. New York, NY: Thieme Medical, 1999; 165–187.
- Kundel HL, Polansky M. Measurement of observer agreement. Radiology 2003;228:303–308. [Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
A. F. Struck and V. M. Haughton
Idiopathic Syringomyelia: Phase-Contrast MR of Cerebrospinal Fluid Flow Dynamics at Level of Foramen Magnum
Radiology,
June 30, 2009;
(2009)
2531082135.
[Abstract]
[Full Text]
|
|
|
|
|
|
|
D. L CALDWELL, C. O DUBOSE, and T. B WHITE
Chiari Malformations
Radiol. Technol.,
March 1, 2009;
80(4):
340MR - 354MR.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
