- Email: [email protected]
Spinal Cord Infarction: Clinical and Radiological Features
ARTICLE IN PRESS
Spinal Cord Infarction: Clinical and Radiological Features Nishtha Yadav, MD,* Hima Pendharkar, DNB, DM,* and Girish Baburao Kulkarni, MD, DM† Introduction: Spinal cord infarction is an uncommon disease varying in its clinical presentation. This study describes the clinical and radiological presentation of spinal cord infarcts in 17 consecutive patients. Material and Methods: Clinical and MR imaging data of 17 patients were reviewed. Inclusion criteria were acute or subacute presentation (peak within 72 hours) and MRI showing typical signal changes on T2WI compatible with spinal cord infarct. Exclusion criteria were clinical or MRI findings suggesting other etiologies. Results: Clinical presentation included dissociative anesthesia, weakness of limbs, back or neck pain, and autonomic symptoms with symptom onset to peak time ranging from few minutes to 48 hours in patients with anterior spinal artery infarct (n = 16), and weakness and sensory loss in ipsilateral upper limb in patient with posterior spinal artery infarct (n = 1). One patient presented with “man-in-the-barrel syndrome (MIB).” MRI findings in anterior spinal artery infarcts included pencillike hyperintensities on T2 sagittal (n = 16, 100%) and “owl eye” appearance on T2 axial (n = 6, 37.5%) images. Diffusion restriction was noted in 8 cases and enhancement was noted in 2 cases. The posterior spinal artery infarct showed T2 hyperintensity in left posterior paramedian triangular distribution in cervical cord (C2-C7). Follow-up was available for 9 patients (period ranging from 15-41 months). Four patients had a favorable outcome who could walk independently, 1 patient could walk with support, and 2 patients were wheelchair bound. Two patients died. Conclusion: Spinal cord infarction is a rare but important cause of acute spinal syndrome. Typical distribution and appropriate imaging can help in timely diagnosis. Key Words: Spinal cord—ischemia—infarction—MRI—anterior spinal artery © 2018 National Stroke Association. Published by Elsevier Inc. All rights reserved.
Spinal cord infarction (SCI) is an uncommon disease that widely varies in its clinical presentation, severity, and outcome.1 Due to its rarity, reliable estimates of incidence are few; however, one study noted that SCI accounted for 1.2% of all strokes.2,3 Spinal cord infarcts can occur in the territories of the anterior spinal artery (ASA) or posterior spinal artery (PSA) or both. ASA syndrome typically presents with an abrupt onset of bilateral weakness, especially the lower limbs, sudden back pain, flaccid paraplegia, areflexia, loss of pain and, temperature sensations below the level of the lesion; sparing of proprioception and vibration sense (dissociative anesthesia) and autonomic
dysfunction involving the bladder and bowel.4 Infarcts in the territories of PSA are very rare.5 The patients with PSA infarct usually present with signs of involvement of bilateral posterior columns, bilateral posterior horns, and posterior segments of bilateral lateral columns.6 The diagnosis of SCI depends on clinical symptoms as mentioned above and are substantiated by Magnetic resonance Imaging (MRI) findings that includes focal cord swelling and “pencillike” hyperintensities on T2-weighted images.7,8 However, as is now known, conventional MRI may be normal in some patients especially in the acute phase.2,9,10 Diffusion weighted imaging (DWI) is
From the *Department of Neuroimaging & Interventional Radiology, National Institute of Mental Health and Neurosciences, Bangalore 560029, India; and †Department of Neurology, National Institute of Mental Health and Neurosciences, Bangalore 560029, India. Received March 21, 2018; revision received May 7, 2018; accepted June 8, 2018. Address correspondence to Hima Pendharkar, Department of Neuro Imaging & Interventional Radiology, National Institute of Mental Health and Neurosciences, Hosur Road, Bangalore 560029, India. E-mails: [email protected] [email protected] [email protected] 1052-3057/$ - see front matter © 2018 National Stroke Association. Published by Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jstrokecerebrovasdis.2018.06.008
Journal of Stroke and Cerebrovascular Diseases, Vol. &&, N0. && (&&), 2018: pp 1-12
1
ARTICLE IN PRESS N. YADAV ET AL.
2
recommended in all cases for detection of suspected SCI in the early stage, as is done for the brain. Our study describes the clinical and radiological presentation of spinal cord infarcts in 17 consecutive patients who presented in our institute over a period of 4 years.
Methods This retrospective observational study was performed at a tertiary referral center in southern India. Patients who presented to our institute with signs and symptoms of SCI over a period of 4 years (March 2013 to March 2017) were included.
Inclusion and Exclusion Criteria Inclusion criteria: (1) acute or subacute spinal cord syndrome like presentation (onset to peak within 72 hours); and (2) MRI scan showing signal changes on T2 weighted images in typical locations compatible with a spinal cord infarct. Exclusion criteria: Clinical or MRI findings suggesting other etiologies such as compressive myelopathy/ autoimmune/ infectious/parainfectious myelitis or systemic disease. MRI studies and the case records of 17 patients with typical sudden onset of neurological deficits due to SCI within the said period were evaluated. The images were reviewed by two neuroradiologists independently (H.S.P., N.Y.) and the clinical data for age, sex, and clinical symptoms at onset were reviewed by the neurologist (G.B.K.) and recorded. As per routine clinical protocol, all patients had blood tests, an electrocardiogram, and an evaluation for vascular risk factors: arterial hypertension, diabetes mellitus, previous stroke; or myocardial infarction. Tests for infectious and connective tissue diseases were performed only when considered necessary by the clinicians. All patients had an MRI scan with T2 weighted sagittal images visualizing the entire length of the spinal cord. T1 sagittal, T1 axial, and T2 axial images were acquired at the level of the lesion. A total of 22 MRI scans were done on 17 patients; 17 scans were done at the time of presentation of whom 15 were performed at our institute and 2 patients had initial scan done outside; 5 were follow-up scans done at our institute (follow-up ranging from 5 days to 1 year). 15 scans at presentation were performed at our institute in 1.5 T MAGNETOM Aera Siemens machine (n = 10 scans) and 3T MAGNETOM Skyra Siemens machine (n = 5 scans). The 5 follow-up scans were also performed in 1.5 T MAGNETOM Aera Siemens machine (n = 4 scans) and 3T MAGNETOM Skyra Siemens machine (n = 1 scan). Out of 22 available scans, DWI was available in 11 studies and postcontrast T1W images were available in 16 studies. Clinical and verbal follow-up was obtained.
Results There were 11 male and 6 female patients, age ranging from 12 years to 87 years (mean age 32.6 years). Symptom
onset to peak time ranged from few minutes to 48 hours. The clinical symptoms are summarized in Table 1. There were 16 patients with ASA infarct and 1 patient with PSA infarct. All 16 patients with ASA infarct presented dissociative anesthesia (n = 16, 100%), and weakness of limbs (n = 16, 100%) upper/lower was noted depending on the level of lesion in the cord. Back pain and neck pain was also noted (n = 10, 62.5%). Autonomic symptoms were noted in 10 patients (n = 10, 62.5%) in form of urinary retention (n = 6), urinary incontinence (n = 4), and constipation (n = 1, this patient additionally had urinary retention). Interestingly, a 12-year-old girl presented with sudden onset weakness of bilateral upper limbs, with onset to peak time of 5 minutes, without weakness in lower limbs or any bulbar symptoms. One patient had PSA infarct, who presented with weakness and loss of sensation in ipsilateral upper limb; he also had associated brachial plexus injury (posttraumatic). Among these 17 patients with SCI, associated cardiovascular risk factors included hypertension (n = 3, 17.6%), diabetes (n = 1, 5.8%), mitral valve prolapse (n = 1, 5.8%), and coarctation of aorta (n = 1, 5.8%). The underlying etiologies for SCI include atherosclerosis (n = 1, 5.8 %), vertebral artery thrombosis (n = 1, 5.8%), iatrogenic (n = 3, 17.6%, post superior cerebellar artery aneurysm coiling, postclipping of middle cerebral artery bifurcation aneurysm in patient with coarctation of aorta, and iatrogenic vertebral artery injury; posttraumatic (n = 1, 5.8%, patient with PSA infarct). The cause remained unknown in 11 cases. Management: Out of the 16 cases of ASA infarct, antiplatelet therapy (T. Aspirin) was given in all cases of SCI with vertebral artery thrombosis/atherosclerotic/idiopathic etiology (n = 13). Two patients had SCI due to hypotension and global hypoperfusion that was managed appropriately. One patient was on ventilatory support but he expired within 1 week. The patient with PSA infarct had associated brachial plexus injury and underwent intercosto-musculocutaneous neurotization with sural nerve graft, no separate treatment was given for PSA infarct. All patients underwent physiotherapy and functional ability training for rehabilitation. Follow-up and outcome: It was available in 9 (n = 52.9 %) patients (period ranging from 15 months to 41 months). Of these 9 cases, clinical follow-up was available for 8 patients while only verbal follow-up was obtained for 1 patient. Four patients had favorable outcome and could walk independently. One patient could walk with a stick at 15-month follow-up and 2 patients were wheelchair bound at 18- and 37-month follow-up. One patient had involvement of cord above C3 level (C2-C7), he developed respiratory failure and expired 1 week following ASA infarct. One patient died due to unknown cause.
MRI Features Patients with ASA infarct (n = 16): These 16 patients had initial scan done within 17 hours to 5 months
Symptoms Case no. Age/sex
Onset to peak
Back/neck pain
21/M
4h
Back pain present
2
14/M
1 hour
Back pain present
3
78/M
Within minutes
Back pain present
4
33/M
4h
Neck pain present
5
87/F
Within minutes
Back pain present
6
44/M
48 h
Absent
7
52/M
Postcoiling
Absent
8
19/F
30 min
Back pain present
9
17/F
2h
Neck pain present
10
15/M
30 min
Absent
11
16/M
6h
Absent
Paraplegia followed by distal upper limb weakness Bilateral upper limb weakness followed by lower limb Bilateral lower limb weakness
Sensory
Level of Cardiovascular ischemia risk factors
Etiology
Follow-up
Urinary Pain sensation retention impaired from T4 level
C6-T3
None
Unknown
Walk with stick
Pain sensation Urinary impaired from C4 incontinence level
C4-C7
None
Unknown
Walk independently
Urinary retention, constipation
T5-T12
Hypertension
Atherosclerosis
Dead
Absent
C3-C5
None
Unknown
Not available
Absent
D9 till conus
None
Unknown
Not available
Urinary incontinence
C5-C6
Hypertension, diabetes
Left vertebral Walk artery thrombosed independently
Absent
C2-C7
Hypertension
Post right SCA aneurysm coiling Unknown
Dead
Loss of pain temperature, preserved posterior column Weakness of upper Loss of pain, touch sensation, level limbs followed by T3-T4 lower limbs Loss of pain, temWeakness lower perature bilateral limbs followed by lower limbs upper limbs Decreased pain, Weakness lower touch in C3,C5 limbs and distal dermatomes upper limbs Quadriplegia Comatosed
Weakness upper limb followed by lower limb Weakness lower limb followed by upper limb Weakness bilateral upper limb followed by lower limbs Weakness bilateral lower limb
Autonomic
Pain, touch absent till D2
Urinary C2- D4 incontinence
Mitral valve prolapse
Walk independently
Decreased pain Urinary sensation, level incontinence C5 Dragging sensation, Urinary retention numbness bilateral upper limbs
C3-D3
None
Unknown
Wheelchair bound
C3-D1
None
Unknown
Not available
Loss of pain, touch Urinary sensation, sensory retention level T4
C5-T1
None
Unknown
Not available
3
(continued on next page)
ARTICLE IN PRESS
1
Motor
CLINICAL AND RADIOLOGICAL FEATURES
Table 1. Clinical features, etiology, and follow-up
4
Table 1 (Continued) Symptoms Case no. Age/sex
Onset to peak
Back/neck pain
21/M
5h
Absent
13
30/F
Postoperative Absent
14
48/M
Postoperative Absent
15
26/F
4h
Back pain present
16
12/F
5 min
17
22/M
Postop, posttrauma
Neck pain radiating to bilateral upper limbs Absent
Sensory
followed by upper limb Weakness upper limbs followed by lower limbs
Autonomic
Level of Cardiovascular ischemia risk factors
Etiology
Follow-up
Loss of pain, touch sensation, paraesthesia, sensory level T3 Bilateral lower Loss of pain, touch limb weakness sensation, sensory level D6 Bilateral lower Loss of pain, touch limb weakness sensation, sensory level D6 Paraesthesial bilatWeakness left eral lower limbs lower limb foland decreased lowed by right pain sensation, lower limb level D4 Weakness bilateral Decreased pain sensation, level upper limb (proxC3 imal > distal)
Urinary retention
C5-C6
None
Unknown
Walk independently
Absent
D4 to conus
Coarctation of aorta
Postoperative? Hypotension
Wheelchair bound
Absent
D4 till conus
None
Urinary retention
D4-D7
None
Postoperative injury Not available to left vertebral artery Unknown Not available
Absent
C3- C7
None
Unknown
Not available
Weakness left upper limb
Absent
C2-C7
None
Traumatic
Not available
Decreased sensation left upper limb
ARTICLE IN PRESS
12
Motor
N. YADAV ET AL.
ARTICLE IN PRESS CLINICAL AND RADIOLOGICAL FEATURES
(depending on patient's presentation to the hospital). One patient (post SCA coiling) had a scan done immediately after coiling. The location of infarct was: cervical cord (n = 6, 37.5%), cervicodorsal cord (n = 5, 31.3%), dorsal cord (n = 2, 12.5%), and dorsolumbar cord (n = 3, 18.7%). The involvement was noted in anterior two-thirds of cord involving gray and white matter in 6 patients (37.5%), only in anterior horns in 5 patients (31.25%), and in anterior horn and central gray matter in 5 patients (31.25%). The initial scan of 7 patients showed cord expansion (43.7%). Owl's eye pattern was noted in 6 patients (37.5%). None of these patients had involvement of peripheral territory of the ASA. Signal changes were hyperintense on T2 and isointense on T1 in all patients. DWI was available in 11 studies and restriction was noted in 8 studies: these studies were performed between 3 hours to 21 days after onset of symptoms. Post contrast T1 W images were available in 16 studies (immediate and follow-up MRI); enhancement was noted in 2 studies when MRI was performed within 8 days and 17 days after onset of symptoms in 2 different patients. One of our cases showed vertebral body marrow signal changes attributed to ischemia. Five patients had follow-up scans with period ranging from 4 days to 1 year; these scans showed T2 hyperintensity and T1 isointensity at their respective levels on follow-up. One patient showed cord atrophy in 12-month follow-up scan. One patient with PSA infarct had T2 hyperintense, T1 isointense signal change in left posterior paramedian triangular distribution in the cervical cord (C2-C7), without expansion/enhancement of cord. DWI was not available for this patient. All imaging features are summarized in Tables 2 and 3. Representative images of ASA and PSA infarcts are shown in Figure 1.
Discussion Spinal cord has an intricate network of blood supply. Various texts and articles describe the anatomy in exquisite details.11-15 Patients with SCI develop symptoms quickly with maximal symptomatology reached within 12 hours for 50% of patients and within 72 hours for most patients.16 The symptom onset to peak time ranged from few minutes to 48 hours in our study. All 16 patients with ASA infarct in our study presented with dissociative anesthesia and weakness of upper limbs/lower limbs depending on the level of lesion in the cord; back pain or neck pain was noted in 10 patients (n = 10, 62.5%). Additionally, autonomic symptoms were noted in 10 patients (n = 10, 62.5%). The clinical presentation is defined by the vascular territory involved with the severity of the impairment varying widely, from paraplegia to minor weakness.
5
Back pain often accompanies SCI and has been reported in as many as 70% of patients typically at the level of the lesion.16 Following are the clinical patterns of ASA infarct17
1. Typical ASA infarct pattern: with involvement of anterior two-thirds of cord involving gray and white matter, causing bilateral weakness, dissociative sensory loss with or without autonomic symptoms. 2. Incomplete pattern: a) Involvement localized at the level of the anterior horns. Clinical manifestation may be in form of: i) Acute paraplegia (pseudopoliomyelitic form) without sensory abnormalities and without sphincter dysfunction. ii) Painful bilateral brachial diplegia in the case of a cervical lesion (the MIB syndrome).18 iii) Progressive distal amyotrophy due to chronic lesions of the anterior horns; this form may be misdiagnosed as lateral amyotrophic sclerosis. b) Sulcocommissural syndrome presents as a partial Brown-Sequard syndrome with sparing of postural sensibility. The syndrome consists of hemiparesis with a contralateral spinothalamic sensory deficit. c) Central spinal infarct occurs after cardiac arrest or prolonged hypotension; its clinical presentation includes bilateral spinothalamic sensory deficit with sparing of the posterior columns. Motor deficit and sphincter dysfunction are usually absent. 3. Another rare presentation is transverse medullary infarction (full transverse lesions) presenting with sudden inability to walk due to paraplegia/paraparesis or tetraplegia/tetraparesis (in higher cord lesions). There may be complete sensory loss involving all modalities and pain reported as circumferential tightness. Autonomic dysfunction is noted. Usually it is of embolic origin. Interestingly 1 patient, a 12-year-old girl, presented with sudden onset weakness of bilateral upper limbs, with onset to peak time of 5 minutes. There was no weakness in lower limbs and no bulbar symptoms. This type of presentation is known as acute brachial diplegia (MIB syndrome). MIB syndrome has been ascribed to supratentorial brain lesions mostly in patients with lesions at the borderzone between anterior and middle cerebral artery
6
Table 2. MRI findings at presentation Time from onset to MRI S. no.
2 3
4 days 8 days
4 5
17 h 2 days
6 7
5 month 2h
8
36 h
9
30 h
10 11
19 h 35 h
12
36 h
13 14
4 days 30 h
15
21 days
16
48 h
17
5 months
C6-D4
Anterior two-thirds cord (gray matter + white matter) C4-C7 Anterior horn D5-D12 Anterior two-thirds cord (gray matter + white matter) C3-C5 Anterior horn D9 till conus Anterior horn + central gray matter C5-C6 Anterior horn C2-C7 Anterior horn + central gray matter C2- D4 Anterior two-thirds cord (gray matter + white matter) C3-D3 Anterior two-thirds cord (gray matter + white matter) C3-D1 Anterior horn C5-D1 Anterior two-thirds cord (gray matter + white matter) C4-C6 Anterior two-thirds cord (gray matter + white matter) D4 till conus Anterior horn D4 till conus Anterior horn + central gray matter D1-D5 Anterior horn + central gray matter C3- C7 Anterior horn + central gray matter C2-C7 Left paramedian triangular
T1
T2
DWI
Contrast
Isointense Hyperintense Not available
Not available
Yes
Yes
No
No
Isointense Hyperintense Not available Isointense Hyperintense Diffusion restriction present
Not available Enhancement present
No No
Yes Yes
Yes No
No No
Isointense Hyperintense Not available Isointense Hyperintense Diffusion restriction present Isointense Hyperintense Not available Isointense Hyperintense Diffusion restriction present Isointense Hyperintense Diffusion restriction present
No enhancement No enhancement
No No
Yes Yes
No No
No No
No enhancement Not available
No Yes
Yes Yes
Yes No
No No
No enhancement
Yes
Yes
Yes
No
Isointense Hyperintense Diffusion restriction present
No enhancement
Yes
Yes
Yes
No
Isointense Hyperintense Not available Isointense Hyperintense Diffusion restriction present
No enhancement No enhancement
Yes No
Yes Yes
Yes No
No No
Isointense Hyperintense No diffusion restriction
No enhancement
Yes
Yes
No
No
Isointense Hyperintense Not available Isointense Hyperintense Not available
Not available Not available
No Yes
Yes Yes
Yes No
No No
Isointense Hyperintense No diffusion restriction Isointense Hyperintense Diffusion restriction present Isointense Hyperintense Not available
No enhancement
No
Yes
No
Yes
No enhancement
No
Yes
No
No
No enhancement
No
Yes
No
No
ARTICLE IN PRESS
4 days
Location
Vertebral Cord Pencil thin Owl’s eye body signal changes expansion appearance appearance
N. YADAV ET AL.
1
Level of ischemia
ARTICLE IN PRESS
No Yes No
No
No Yes
Isointense Hyperintense Diffusion restriction present 5 days 14
D4 to conus 17 days 13
C4-C7 C3-D1 9 months 1 year 2 9
D4 till conus Anterior horn + central gray matter
Isointense Hyperintense No diffusion restriction
Enhancement present (enhancement of anterior horn D12-L1 with cauda equine anterior roots enhancement) No enhancement
No
Yes
No No Yes Yes Yes Yes No Atrophy Isointense Hyperintense Not available Not available Isointense Hyperintense Not available No enhancement
No No Yes No Isointense Hyperintense Not available No enhancement
Anterior two-thirds cord (gray matter + white matter) Anterior horn Anterior two-thirds cord (gray matter + white matter) Anterior horn 2 months 1
C7-D2
DWI T2 T1 Location Time from onset to MRI
Level of ischemia
7
Case no.
Table 3. Follow-up MRI findings
Contrast
Vertebral Cord Pencil thin Owl’s eye body signal changes expansion appearance appearance
CLINICAL AND RADIOLOGICAL FEATURES
territories including the prerolandic cortex; etiologies may be varied.18-21 Spinal cord infarct as a cause of MIB is much less common than cerebral lesions. The literature has less than 10 cases of MIB syndrome caused by cervical infarcts. The clinical pattern of PSA infarction: Because of presence of two posterior spinal arteries, this syndrome is usually unilateral and less severe. Based on the functional neuroanatomy, this syndrome leads to ipsilateral loss of light touch, vibration, and proprioception while mostly sparing the motor function. However, there can be segmental deep tendon areflexia due to posterior horn involvement and paresis below the level at which the posterior portion of the cord is affected. Among the 17 patients with SCI in our study, the underlying etiologies include atherosclerosis (n = 1), vertebral artery thrombosis (n = 1), iatrogenic (n = 3, post superior cerebellar artery aneurysm coiling, postclipping of middle cerebral artery bifurcation aneurysm in patient with coarctation of aorta, and iatrogenic vertebral artery injury), posttraumatic (n = 1, patient with PSA infarct). Causes of SCI differ in the pediatric and adult populations. In children, the most common causes are cardiac malformations and trauma. In adults, atherosclerosis is one of the principal causes, but thoracoabdominal aneurysms, aortic surgery, embolic disease, dissection, systemic hypotension, spinal arteriovenous malformations, diving, coagulopathies, cocaine abuse, and sickle cell disease are implicated (17). In many cases, the cause remains unknown and these are categorized as idiopathic; as was noted in 11 of our cases. Our institute does not have a vascular surgery department explaining the lack of cases attributed to aortic diseases as cause of SCI. One of our patient (case no. 13) developed SCI following clipping of middle cerebral artery bifurcation aneurysm; she had documented intraoperative hypotension complicated additionally by presence of coarctation of aorta, possibly making the patient more susceptible to the effect of hypotension (Fig 2). Another case of craniovertebral junction surgery (case no. 14) with iatrogenic vertebral artery injury also had documented intraoperative hypotension. Both these patients with hypotension as the etiology developed infarct in dorsolumbar cord (D4 till conus). Historically, a spinal cord “watershed zone” of ischemic vulnerability centered at the midthoracic level (D4D6 level) has been suggested.22-24 This assumption was mainly based on anatomic studies and case reports describing the relative hypovascularity of the midthoracic region (D4-D8),14-23 which was then equated with increased susceptibility of this area to ischemic injury. However, there have been studies suggesting that global ischemia may affect the low thoracic and lumbosacral cord to a greater extent than the other levels of the spinal cord.25-28
ARTICLE IN PRESS 8
N. YADAV ET AL.
Figure 1. Typical imaging appearance of anterior spinal artery (different patients) (a-f) and posterior spinal artery infarct (g, h). T2W sagittal image (a) shows pencil thin hyperintensity in anterior aspect of cervical cord. T2W axial images (b-d) show the patterns of T2 hyperintensity involving anterior horn cells (owl's eye) in (b), anterior two thirds of spinal cord (c), and anterior horn and central gray matter (d). Diffusion weighted image (e) and apparent diffusion coefficient (ADC) map (f) showing restricted diffusion involving cord in anterior spinal artery territory. Posterior spinal artery infarct: Left posterior paramedian triangular T2 hyperintensity on axial T2 image (g) and associated brachial plexus injury with pseudomeningocele formation on STIR coronal (h).
The findings in the study done by Duggal and Lach also indicated a greater vulnerability of lumbosacral neurons to ischemic insults from either cardiac arrest or a severe sustained episode of hypotension.29 They proposed that this is most likely a consequence of greater metabolic demands of the gray matter at this level of the spinal cord. Our findings in patient with SCI secondary to hypotension are also in concordance with this. In our study, the involvement was noted in cervical cord (n = 6, 37.5%), cervicodorsal cord (n = 5, 31.3%), dorsal cord (n=2, 12.5%), and dorsolumbar cord (n = 3, 18.7%). The ASA syndrome is divided into three clinical pictures according to the particular cord segment involved: cervicothoracic, middle thoracic, or thoracolumbar. In previously published studies, thoracolumbar infarction has been considered as the most frequent and the best clinically defined syndrome.30 However, we noted cervical and cervicodorsal regions as the more common location. This, as stated earlier, can be attributed to lack of aortic surgeries at our institute as a cause and hence the given distribution of lesions. The sensitivity of initial MRI of the cord is limited, nearly 17 45% of clinically suspected SCI cases have a normal scan. More importantly, MRI rules out other causes presenting as acute spinal syndrome, such as compressive myelopathies, vascular malformations, infective myelitis, demyelinating disorders, and tumor.31
Literature describes MRI patterns in patients with ASA infarcts as “owl eye” or “snake eye” appearance on T2 weighted (T2W) axial and pencillike hyperintensities on T2W sagittal images.32 The grey matter of the anterior horns exhibit the highest vulnerability to ischemia due to high metabolic demand that may lead to the typical ‘‘snake-eye’’ or “owl-eye” appearance of cord infarction on axial T2W images7 as noted in 6 (37.5%) of our patients. ‘‘Pencillike’’ hyperintensity in sagittal T2W images was seen in all 16 of our patients with ASA infarct, similar to the study results of Yuh et al33 and Stefan Weidauer et al.7 Seven (43.7%) of our patients showed cord expansion that has been described in early stages. In the acute or subacute stage, DWI was available in 11 studies and diffusion restriction was noted in 8 studies; the earliest study being performed 3 hours and the longest one was performed 21 days after onset of symptoms. Out of these 11 cases, 3 cases that did not show restriction were done on the 17th day and 21st day after onset of symptoms. The shortest time reported in the literature between the onset of clinical symptoms and abnormalities on DWI is 3 hours,8 which was also noted in our study in 1 patient (case no. 7). In the brain, DWI can demonstrate restriction as early as 20-30 minutes after the onset of symptoms. However, larger studies are needed to establish a time threshold for diffusion changes in patients with spinal cord ischemia/infarction. Similar to our study, in most of
ARTICLE IN PRESS CLINICAL AND RADIOLOGICAL FEATURES
the previously reported patients, signal abnormalities were demonstrated on T2-W images and on DWI. However, in previously described studies, conventional Magnetic resonance (MR) sequences did not show abnormality in 4 reported patients in 3, 4, 10, and 46 hours after the onset of spinal ischemia10,34,35; DWI might have helped in these cases. A number of technical limitations exist for DWI of spinal cord including field inhomogeneity, CSF and blood pulsations, and small size of infarcts.36 Though it seems to be established that SCI can be identified using DW-MRI, the dynamic evolution of diffusion abnormalities remains to be identified.8,37 Enhancement is usually noted in the subacute stage. In our study, enhancement was noted in 2 studies (case nos. 3 and 13), when MRI was performed within 8 days and 17 days after onset of symptoms. Interestingly, in case no. 13, the study performed on the 4th day following symptoms did not show enhancement. However, MRI performed on the same patient on the 17th day showed enhancement of anterior horn at D12 to L1 with enhancement of anterior roots of cauda equina (Fig 2). The concomitant enhancement of the cauda equina was first reported by Friedman and Flanders in 1992.38 This phenomenon was described as a characteristic finding in the
9
course of SCI that might involve the cord and the ventral cauda equina as well (which comprises of motor fiber bundles).39 It denotes disruption of the blood cord barrier as well as reactive hyperemia.38-40 One of our cases showed vertebral body marrow signal changes attributed to ischemia (Fig 3). Studies have demonstrated that a finding of vertebral body infarction adjacent to a cord signal change on MRI is a useful confirmatory sign41,42 although found in only 4%-35% of the patients (since arterial occlusion may be located distal to the anterior or posterior central arteries that supply the vertebral body).31 The vertebral bone marrow abnormality that tends to occur earlier is more exaggerated compared to the cord and is more commonly seen in ASA syndromes as compared to PSA syndromes.42 Appropriate clinical setting and cord signal changes are indicative of an ischemic etiology as cause of vertebral marrow signal change. In our study, 1 patient with PSA infarct had T2 hyperintense, T1 isointense signal change in left posterior paramedian triangular distribution in the cervical cord (C2-C7), without the presence of expansion/enhancement of the cord. Diffusion sequence was however not available for this patient.
Figure 2. Case no. 14. Thirty-year-old female who presented with sudden onset headache followed by generalized seizures. Non contrast CT head (a) shows the presence of Sub arachnoid hemorrhage (SAH) in basal cisterns, sylvian fissures, and interhemispheric fissure. During cerebral angiography, the catheter could not be negotiated beyond the thoracic aorta and a pigtail injection at this point (b) showed the presence of coarctation of aorta. A CT angiography (c) revealed the presence of left MCA bifurcation aneurysm. Surgery for clipping of aneurysm was complicated by intraoperative hypotension; postoperative paraplegia was noted. Dorsal spine T2 sagittal (d) and T2 axial images (e) show pencil thin hyperintensity involving cord from D4 till conus with hyperintensity involving anterior horn and central gray matter. Contrast enhanced T1W sagittal (f) and axial images (g, h) done on postoperative day 17 shows characteristic contrast enhancement involving anterior horns and ventral cauda equina. String of pearls appearance of the ventral nerve roots is noted in fig h.
ARTICLE IN PRESS N. YADAV ET AL.
10
Figure 3. Case no. 16. Twenty-six-year-old female presented with sudden onset back pain, followed by weakness of bilateral lower limbs (left followed by right) with onset to peak time of 4 hours. T2W sagittal (a) and axial images (b) show the presence of pencil thin hyperintensity in upper dorsal cord and hyperintensity involving anterior horns respectively. T2W image in parasagittal location (c) shows the presence of vertebral marrow hyperintensity involving D4 vertebra (white arrow). T2W axial image at D4 (d) shows the cord hyperintensity with associated vertebral marrow wedge shaped hyperintensity on left side. Corresponding diffusion weighted imaging (e) showing restriction in the cord. The marrow signal change was isointense on T1W axial image (f) and shows contrast enhancement (white arrow) on post contrast T1W axial image (g).
PSA syndrome is rare and to our knowledge only about 30 cases have been reported so far.5,43 Williamson first reported an autopsy case of PSA syndrome in 1895.44 Described causes of PSA infarct include syphilitic arteritis, cholesterol emboli from atheromatous aortic plaques, intrathecal injection of phenol, vertebral artery dissection, and plasmacytoma.6 In our patient, PSA infarct was attributed to trauma due to road traffic accident resulting in associated ipsilateral brachial plexus injury. The infarct was noted in the cervical cord spanning 6 segments (C2-C7). The majority of described PSA infarcts occur at the thoracolumbar level but they have been observed at the thoracic and cervical levels too.43 Longitudinal extension averages 2 vertebral segments but their span can range from 1-6 segments,43 as was also noted in our study (C2-C7). Appropriate clinical setting and cord signal changes will support the diagnosis of a PSA infarct.
Clues to Identification of SCI Include Clinical Sudden onset of symptoms with onset to peak duration less than 72 hours. Symptoms varying depending on territory, location and extent of infarction.
Imaging Different imaging modalities might give a clue to the diagnosis depending on the clinical scenario (eg Computed tomography (CT) angiography done in patient with abdominal pain and hypotension may reveal aortic dissection and CT spine done in case of cervical spine trauma may reveal vertebral artery injury). However, if SCI is suspected, MRI is essential for diagnosis and should include the following sequences: T2W Sagittal, T2W axial, Short tau inversion recovery (STIR) sagittal, DWI, contrast enhanced T1W sagittal, and axial sequences. Specifically we look for pencil thin hyperintensity, owl's eye appearance (ASA infarct), and posterior paramedian triangular hyperintensity (PSA infarct). Marrow signal changes/diffusion restriction/contrast enhancement can be noted in ASA/PSA infarct depending on time of imaging from onset. Follow-up of patients with ASA infarct was available for 9 patients. Of these, 4 could walk independently, 1 patient could walk with support, 2 patients were wheelchair bound, 1 patient died due to unknown cause, and 1 succumbed to respiratory failure with involvement of cord above C3 (C2-C7). Previous studies have advocated various factors in predicting prognosis of patients of SCI. Severe neurological deficits at onset and involvement of
ARTICLE IN PRESS CLINICAL AND RADIOLOGICAL FEATURES
proprioception were associated with poor outcome, whereas the presence of owl's eye sign was associated with a good outcome.2 Management largely depends on the underlying cause of infarction. Emergency surgery is needed if acute aortic event, vascular compression is the underlying cause of spinal ischaemia. In the event of global hypoperfusion, maintaining adequate blood pressure to maintain spinal perfusion is crucial. Antiplatelet therapy is needed to prevent vascular occlusions or embolism. Corticosteroids are indicated in cases of vasculitis or aortitis. Novel findings in our study include: depiction of spinal cord infarct on DWI at 3 hours after symptom onset (3 hours being the earliest in literature for DWI to demonstrate the signal changes). Next is a girl presenting with “(wo)man-in-the-barrel syndrome” due to cervical cord infarct; literature describes less than 10 such cases. Another observation is of concomitant enhancement of anterior horn and cauda equina noted in one of our patients; this adds to the limited literature on this characteristic finding in the course of spinal cord ischemia. Further, 2 patients developed cord ischemia secondary to hypotension in dorsolumbar cord, which in accordance with studies suggesting that global ischemia may affect the low thoracic and lumbosacral cord to a greater extent than the other levels of the spinal cord as opposed to historically described spinal cord “watershed zone” of ischemic vulnerability centered at the midthoracic level (D4-D6 level). However, our study also had certain limitations. Ours being only a neurosciences tertiary referral center, relative late presentation to the hospital was common. Diffusion sequence was only available for 11 cases: the imaging protocols have now been modified. Follow-up was available only for 9 patients. In conclusion, ASA and PSA infarcts are rare but important causes of acute spinal syndrome. A high index of suspicion is essential to identify these lesions. The earliest change is noted on DWI sequence while enhancement is noted in the subacute stage. Typical distribution and imaging appearance of lesions with signal changes evolving over time on various sequences can help in differentiating SCI from other etiologies. Spinal imaging with addition of DWI can thus help in diagnosing this uncommon entity early in the disease course.
References 1. Lynch K, Oster J, Apetauerova D, Hreib K. Spinal cord stroke: acute imaging and intervention. Case Rep Neurol Med 2012;2012:1-3. 2. Masson C. Spinal cord infarction: clinical and magnetic resonance imaging findings and short term outcome. J Neurol Neurosurg Psychiatry 2004;75:1431-1435. 3. de la Barrera SS, Barca-Buyo A, Montoto-Marques A, et al. Spinal cord infarction: prognosis and recovery in a series of 36 patients. Spinal Cord 2001;39:520.
11 4. Millichap JJ, Sy BT, Leacock RO. Spinal cord infarction with multiple etiologic factors. J Gen Intern Med 2007;22:151-154. 5. Kaneki M, Inoue K, Shimizu T, et al. Infarction of the unilateral posterior horn and lateral column of the spinal cord with sparing of posterior columns: demonstration by MRI. J Neurol Neurosurg Psychiatry 1994;57:629-631. 6. Matsubayashi J, Tsuchiya K, Shimizu S, et al. Posterior spinal artery syndrome showing marked swelling of the spinal cord: a clinico-pathological study. J Spinal Cord Med 2013;36:31-35. 7. Weidauer S, Nichtweiss M, Lanfermann H, et al. Spinal cord infarction: MR imaging and clinical features in 16 cases. Neuroradiology 2002;44:851-857. 8. Thurnher MM, Bammer R. Diffusion-weighted MR imaging (DWI) in spinal cord ischemia. Neuroradiology 2006;48:795-801. 9. Sibon I, Menegon P, Moonen CTW, et al. Early diagnosis of spinal cord infarct using magnetic resonance diffusion imaging. Neurology 2003;61. 1622 1622. 10. Fujikawa A, Tsuchiya K, Takeuchi S, et al. Diffusionweighted MR imaging in acute spinal cord ischemia. Eur Radiol 2004;14:2076-2078. 11. Amato ACM, Stolf NAG. Anatomia da circula¸c ~ ao medular. J Vasc Bras 2015;14:248-252. 12. Williams GM, Perler BA, Burdick JF, et al. Angiographic localization of spinal cord blood supply and its relationship to postoperative paraplegia. J Vasc Surg 1991;13:23-35. 13. Martirosyan NL, Feuerstein JS, Theodore N, et al. Blood supply and vascular reactivity of the spinal cord under normal and pathological conditions: a review. J Neurosurg Spine 2011;15:238-251. 14. Hassler O. Blood supply to human spinal cord. A microangiographic study. Arch Neurol 1966;15:302-307. 15. Prince E, Ahn S. Basic vascular neuroanatomy of the brain and spine: what the general interventional radiologist needs to know. Semin Interv Radiol 2013;30:234-239. 16. Novy J, Carruzzo A, Maeder P, et al. Spinal cord ischemia: clinical and imaging patterns, pathogenesis, and outcomes in 27 patients. Arch Neurol 2006;63:1113-1120. 17. Vargas MI, Gariani J, Sztajzel R, et al. Spinal cord ischemia: practical imaging tips, pearls, and pitfalls. Am J Neuroradiol 2015;36:825-830. 18. Antelo MJG, Facal TL, S anchez TP, et al. Man-in-the-barrel. A case of cervical spinal cord infarction and review of the literature. Open Neurol J 2013;7:7-10. 19. Berg D, Miillges W, Koltzenburg M, et al. Case report man-in-the-barrel syndrome caused by cervical spinal cord infarction. Acta Neurol Scand 1998;97:417-419. 20. Ben Sassi S, El Euch G, Regaieg A, et al. Man-in-the-barrel syndrome with combination of infarctions in the anterior spinal artery and posterior inferior cerebellar artery territories. Cerebrovasc Dis 2009;27:201-202. 21. Rouanet C, Reges D, Rocha E, et al. “Man in the barrel” syndrome with anterior spinal artery infarct due to vertebral artery dissection. J Stroke Cerebrovasc Dis 2017;26: e41-e42. 22. Bartsch W, Hopf HC. New observations on the relations between cardiac performance and spinal cord blood supply. Dtsch Z Nervenheilkd 1963;184:288-307. 23. Marcus ML, Heistad DD, Ehrhardt JC, et al. Regulation of total and regional spinal cord blood flow. Circ Res 1977;41:128-134. 24. Sliwa JA, Maclean IC. Ischemic myelopathy: a review of spinal vasculature and related clinical syndromes. Arch Phys Med Rehabil 1992;73:365-372.
ARTICLE IN PRESS 12 25. Imaizumi H, Ujike Y, Asai Y, et al. Spinal cord ischemia after cardiac arrest. J Emerg Med 1994;12:789-793. 26. Cheshire WP, Santos CC, Massey EW, et al. Spinal cord infarction: etiology and outcome. Neurology 1996;47: 321-330. 27. Gilles FH, Nag D. Vulnerability of human spinal cord in transient cardiac arrest. Neurology 1971;21:833-839. 28. Azzarelli B, Roessmann U. Diffuse “anoxic” myelopathy. Neurology 1977;27:1049-1052. 29. Duggal N, Lach B. Selective vulnerability of the lumbosacral spinal cord after cardiac arrest and hypotension. Stroke 2002;33:116-121. 30. Monteiro L, Leite I, Pinto JA, et al. Spontaneous thoracolumbar spinal cord infarction: report of six cases. Acta Neurol Scand 1992;86:563-566. 31. Alexander M, Patil A, Mani S, et al. Spectrum of clinicoradiological findings in spinal cord infarction: report of three cases and review of the literature. Ann Indian Acad Neurol 2013;16:190. 32. Cheng M-Y, Lyu R-K, Chang Y-J, et al. Spinal Cord Infarction in Chinese patients. Cerebrovasc Dis 2008;26:502-508. 33. Yuh WT, Marsh EE, Wang AK, et al. MR imaging of spinal cord and vertebral body infarction. Am J Neuroradiol 1992;13:145-154. 34. Weidauer S, Dettmann E, Krakow K, et al. Diffusionsgewichtete MRT bei spinalen Infarkten. Nervenarzt 2002;73:999-1003. 35. Loher TJ, Bassetti CL, Lovblad KO, et al. Diffusionweighted MRI in acute spinal cord ischaemia. Neuroradiology 2003;45:557-561.
N. YADAV ET AL. 36. Kuker W, Weller M, Klose U, et al. Diffusion-weighted MRI of spinal cord infarction: high resolution imaging and time course of diffusion abnormality. J Neurol 2004;251:818-824. Available from: http://link.springer. com/10.1007/s00415-004-0434-z. 37. Bammer R, Fazekas F, Augustin M, et al. Diffusionweighted MR imaging of the spinal cord. Am J Neuroradiol 2000;21:587-591. 38. Friedman DP, Flanders AE. Enhancement of gray matter in anterior spinal infarction. Am J Neuroradiol 1992;13:983-985. 39. Amano Y, Machida T, Kumazaki T. Spinal cord infarcts with contrast enhancement of the cauda equina: two cases. Neuroradiology 1998;40:669-672. 40. Mull M, Thron A. Spinal infarcts. In R€ udiger von Kummer and Tobias Back (Eds.). Magnetic Resonance Imaging in Ischemic Stroke, Springer-Verlag Berlin Heidelberg 2006, 251-267. 41. Faig J, Busse O, Salbeck R. Vertebral body infarction as a confirmatory sign of spinal cord ischemic stroke. Stroke 1998;29:239-243. 42. Suzuki T, Kawaguchi S, Takebayashi T, et al. Vertebral body ischemia in the posterior spinal artery syndrome: case report and review of the literature. Spine 2003;28: E260-E264. 43. Mascalchi M, Cosottini M, Ferrito G, et al. Posterior spinal artery infarct. Am J Neuroradiol 1998;19:361-363. 44. Williamson RT. Spinal softening limited to the parts supplied by the posterior arterial system of the cord. Lancet 1895;146:520-522.
Spinal Cord Infarction: Clinical and Radiological Features Nishtha Yadav, MD,* Hima Pendharkar, DNB, DM,* and Girish Baburao Kulkarni, MD, DM† Introduction: Spinal cord infarction is an uncommon disease varying in its clinical presentation. This study describes the clinical and radiological presentation of spinal cord infarcts in 17 consecutive patients. Material and Methods: Clinical and MR imaging data of 17 patients were reviewed. Inclusion criteria were acute or subacute presentation (peak within 72 hours) and MRI showing typical signal changes on T2WI compatible with spinal cord infarct. Exclusion criteria were clinical or MRI findings suggesting other etiologies. Results: Clinical presentation included dissociative anesthesia, weakness of limbs, back or neck pain, and autonomic symptoms with symptom onset to peak time ranging from few minutes to 48 hours in patients with anterior spinal artery infarct (n = 16), and weakness and sensory loss in ipsilateral upper limb in patient with posterior spinal artery infarct (n = 1). One patient presented with “man-in-the-barrel syndrome (MIB).” MRI findings in anterior spinal artery infarcts included pencillike hyperintensities on T2 sagittal (n = 16, 100%) and “owl eye” appearance on T2 axial (n = 6, 37.5%) images. Diffusion restriction was noted in 8 cases and enhancement was noted in 2 cases. The posterior spinal artery infarct showed T2 hyperintensity in left posterior paramedian triangular distribution in cervical cord (C2-C7). Follow-up was available for 9 patients (period ranging from 15-41 months). Four patients had a favorable outcome who could walk independently, 1 patient could walk with support, and 2 patients were wheelchair bound. Two patients died. Conclusion: Spinal cord infarction is a rare but important cause of acute spinal syndrome. Typical distribution and appropriate imaging can help in timely diagnosis. Key Words: Spinal cord—ischemia—infarction—MRI—anterior spinal artery © 2018 National Stroke Association. Published by Elsevier Inc. All rights reserved.
Spinal cord infarction (SCI) is an uncommon disease that widely varies in its clinical presentation, severity, and outcome.1 Due to its rarity, reliable estimates of incidence are few; however, one study noted that SCI accounted for 1.2% of all strokes.2,3 Spinal cord infarcts can occur in the territories of the anterior spinal artery (ASA) or posterior spinal artery (PSA) or both. ASA syndrome typically presents with an abrupt onset of bilateral weakness, especially the lower limbs, sudden back pain, flaccid paraplegia, areflexia, loss of pain and, temperature sensations below the level of the lesion; sparing of proprioception and vibration sense (dissociative anesthesia) and autonomic
dysfunction involving the bladder and bowel.4 Infarcts in the territories of PSA are very rare.5 The patients with PSA infarct usually present with signs of involvement of bilateral posterior columns, bilateral posterior horns, and posterior segments of bilateral lateral columns.6 The diagnosis of SCI depends on clinical symptoms as mentioned above and are substantiated by Magnetic resonance Imaging (MRI) findings that includes focal cord swelling and “pencillike” hyperintensities on T2-weighted images.7,8 However, as is now known, conventional MRI may be normal in some patients especially in the acute phase.2,9,10 Diffusion weighted imaging (DWI) is
From the *Department of Neuroimaging & Interventional Radiology, National Institute of Mental Health and Neurosciences, Bangalore 560029, India; and †Department of Neurology, National Institute of Mental Health and Neurosciences, Bangalore 560029, India. Received March 21, 2018; revision received May 7, 2018; accepted June 8, 2018. Address correspondence to Hima Pendharkar, Department of Neuro Imaging & Interventional Radiology, National Institute of Mental Health and Neurosciences, Hosur Road, Bangalore 560029, India. E-mails: [email protected] [email protected] [email protected] 1052-3057/$ - see front matter © 2018 National Stroke Association. Published by Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jstrokecerebrovasdis.2018.06.008
Journal of Stroke and Cerebrovascular Diseases, Vol. &&, N0. && (&&), 2018: pp 1-12
1
ARTICLE IN PRESS N. YADAV ET AL.
2
recommended in all cases for detection of suspected SCI in the early stage, as is done for the brain. Our study describes the clinical and radiological presentation of spinal cord infarcts in 17 consecutive patients who presented in our institute over a period of 4 years.
Methods This retrospective observational study was performed at a tertiary referral center in southern India. Patients who presented to our institute with signs and symptoms of SCI over a period of 4 years (March 2013 to March 2017) were included.
Inclusion and Exclusion Criteria Inclusion criteria: (1) acute or subacute spinal cord syndrome like presentation (onset to peak within 72 hours); and (2) MRI scan showing signal changes on T2 weighted images in typical locations compatible with a spinal cord infarct. Exclusion criteria: Clinical or MRI findings suggesting other etiologies such as compressive myelopathy/ autoimmune/ infectious/parainfectious myelitis or systemic disease. MRI studies and the case records of 17 patients with typical sudden onset of neurological deficits due to SCI within the said period were evaluated. The images were reviewed by two neuroradiologists independently (H.S.P., N.Y.) and the clinical data for age, sex, and clinical symptoms at onset were reviewed by the neurologist (G.B.K.) and recorded. As per routine clinical protocol, all patients had blood tests, an electrocardiogram, and an evaluation for vascular risk factors: arterial hypertension, diabetes mellitus, previous stroke; or myocardial infarction. Tests for infectious and connective tissue diseases were performed only when considered necessary by the clinicians. All patients had an MRI scan with T2 weighted sagittal images visualizing the entire length of the spinal cord. T1 sagittal, T1 axial, and T2 axial images were acquired at the level of the lesion. A total of 22 MRI scans were done on 17 patients; 17 scans were done at the time of presentation of whom 15 were performed at our institute and 2 patients had initial scan done outside; 5 were follow-up scans done at our institute (follow-up ranging from 5 days to 1 year). 15 scans at presentation were performed at our institute in 1.5 T MAGNETOM Aera Siemens machine (n = 10 scans) and 3T MAGNETOM Skyra Siemens machine (n = 5 scans). The 5 follow-up scans were also performed in 1.5 T MAGNETOM Aera Siemens machine (n = 4 scans) and 3T MAGNETOM Skyra Siemens machine (n = 1 scan). Out of 22 available scans, DWI was available in 11 studies and postcontrast T1W images were available in 16 studies. Clinical and verbal follow-up was obtained.
Results There were 11 male and 6 female patients, age ranging from 12 years to 87 years (mean age 32.6 years). Symptom
onset to peak time ranged from few minutes to 48 hours. The clinical symptoms are summarized in Table 1. There were 16 patients with ASA infarct and 1 patient with PSA infarct. All 16 patients with ASA infarct presented dissociative anesthesia (n = 16, 100%), and weakness of limbs (n = 16, 100%) upper/lower was noted depending on the level of lesion in the cord. Back pain and neck pain was also noted (n = 10, 62.5%). Autonomic symptoms were noted in 10 patients (n = 10, 62.5%) in form of urinary retention (n = 6), urinary incontinence (n = 4), and constipation (n = 1, this patient additionally had urinary retention). Interestingly, a 12-year-old girl presented with sudden onset weakness of bilateral upper limbs, with onset to peak time of 5 minutes, without weakness in lower limbs or any bulbar symptoms. One patient had PSA infarct, who presented with weakness and loss of sensation in ipsilateral upper limb; he also had associated brachial plexus injury (posttraumatic). Among these 17 patients with SCI, associated cardiovascular risk factors included hypertension (n = 3, 17.6%), diabetes (n = 1, 5.8%), mitral valve prolapse (n = 1, 5.8%), and coarctation of aorta (n = 1, 5.8%). The underlying etiologies for SCI include atherosclerosis (n = 1, 5.8 %), vertebral artery thrombosis (n = 1, 5.8%), iatrogenic (n = 3, 17.6%, post superior cerebellar artery aneurysm coiling, postclipping of middle cerebral artery bifurcation aneurysm in patient with coarctation of aorta, and iatrogenic vertebral artery injury; posttraumatic (n = 1, 5.8%, patient with PSA infarct). The cause remained unknown in 11 cases. Management: Out of the 16 cases of ASA infarct, antiplatelet therapy (T. Aspirin) was given in all cases of SCI with vertebral artery thrombosis/atherosclerotic/idiopathic etiology (n = 13). Two patients had SCI due to hypotension and global hypoperfusion that was managed appropriately. One patient was on ventilatory support but he expired within 1 week. The patient with PSA infarct had associated brachial plexus injury and underwent intercosto-musculocutaneous neurotization with sural nerve graft, no separate treatment was given for PSA infarct. All patients underwent physiotherapy and functional ability training for rehabilitation. Follow-up and outcome: It was available in 9 (n = 52.9 %) patients (period ranging from 15 months to 41 months). Of these 9 cases, clinical follow-up was available for 8 patients while only verbal follow-up was obtained for 1 patient. Four patients had favorable outcome and could walk independently. One patient could walk with a stick at 15-month follow-up and 2 patients were wheelchair bound at 18- and 37-month follow-up. One patient had involvement of cord above C3 level (C2-C7), he developed respiratory failure and expired 1 week following ASA infarct. One patient died due to unknown cause.
MRI Features Patients with ASA infarct (n = 16): These 16 patients had initial scan done within 17 hours to 5 months
Symptoms Case no. Age/sex
Onset to peak
Back/neck pain
21/M
4h
Back pain present
2
14/M
1 hour
Back pain present
3
78/M
Within minutes
Back pain present
4
33/M
4h
Neck pain present
5
87/F
Within minutes
Back pain present
6
44/M
48 h
Absent
7
52/M
Postcoiling
Absent
8
19/F
30 min
Back pain present
9
17/F
2h
Neck pain present
10
15/M
30 min
Absent
11
16/M
6h
Absent
Paraplegia followed by distal upper limb weakness Bilateral upper limb weakness followed by lower limb Bilateral lower limb weakness
Sensory
Level of Cardiovascular ischemia risk factors
Etiology
Follow-up
Urinary Pain sensation retention impaired from T4 level
C6-T3
None
Unknown
Walk with stick
Pain sensation Urinary impaired from C4 incontinence level
C4-C7
None
Unknown
Walk independently
Urinary retention, constipation
T5-T12
Hypertension
Atherosclerosis
Dead
Absent
C3-C5
None
Unknown
Not available
Absent
D9 till conus
None
Unknown
Not available
Urinary incontinence
C5-C6
Hypertension, diabetes
Left vertebral Walk artery thrombosed independently
Absent
C2-C7
Hypertension
Post right SCA aneurysm coiling Unknown
Dead
Loss of pain temperature, preserved posterior column Weakness of upper Loss of pain, touch sensation, level limbs followed by T3-T4 lower limbs Loss of pain, temWeakness lower perature bilateral limbs followed by lower limbs upper limbs Decreased pain, Weakness lower touch in C3,C5 limbs and distal dermatomes upper limbs Quadriplegia Comatosed
Weakness upper limb followed by lower limb Weakness lower limb followed by upper limb Weakness bilateral upper limb followed by lower limbs Weakness bilateral lower limb
Autonomic
Pain, touch absent till D2
Urinary C2- D4 incontinence
Mitral valve prolapse
Walk independently
Decreased pain Urinary sensation, level incontinence C5 Dragging sensation, Urinary retention numbness bilateral upper limbs
C3-D3
None
Unknown
Wheelchair bound
C3-D1
None
Unknown
Not available
Loss of pain, touch Urinary sensation, sensory retention level T4
C5-T1
None
Unknown
Not available
3
(continued on next page)
ARTICLE IN PRESS
1
Motor
CLINICAL AND RADIOLOGICAL FEATURES
Table 1. Clinical features, etiology, and follow-up
4
Table 1 (Continued) Symptoms Case no. Age/sex
Onset to peak
Back/neck pain
21/M
5h
Absent
13
30/F
Postoperative Absent
14
48/M
Postoperative Absent
15
26/F
4h
Back pain present
16
12/F
5 min
17
22/M
Postop, posttrauma
Neck pain radiating to bilateral upper limbs Absent
Sensory
followed by upper limb Weakness upper limbs followed by lower limbs
Autonomic
Level of Cardiovascular ischemia risk factors
Etiology
Follow-up
Loss of pain, touch sensation, paraesthesia, sensory level T3 Bilateral lower Loss of pain, touch limb weakness sensation, sensory level D6 Bilateral lower Loss of pain, touch limb weakness sensation, sensory level D6 Paraesthesial bilatWeakness left eral lower limbs lower limb foland decreased lowed by right pain sensation, lower limb level D4 Weakness bilateral Decreased pain sensation, level upper limb (proxC3 imal > distal)
Urinary retention
C5-C6
None
Unknown
Walk independently
Absent
D4 to conus
Coarctation of aorta
Postoperative? Hypotension
Wheelchair bound
Absent
D4 till conus
None
Urinary retention
D4-D7
None
Postoperative injury Not available to left vertebral artery Unknown Not available
Absent
C3- C7
None
Unknown
Not available
Weakness left upper limb
Absent
C2-C7
None
Traumatic
Not available
Decreased sensation left upper limb
ARTICLE IN PRESS
12
Motor
N. YADAV ET AL.
ARTICLE IN PRESS CLINICAL AND RADIOLOGICAL FEATURES
(depending on patient's presentation to the hospital). One patient (post SCA coiling) had a scan done immediately after coiling. The location of infarct was: cervical cord (n = 6, 37.5%), cervicodorsal cord (n = 5, 31.3%), dorsal cord (n = 2, 12.5%), and dorsolumbar cord (n = 3, 18.7%). The involvement was noted in anterior two-thirds of cord involving gray and white matter in 6 patients (37.5%), only in anterior horns in 5 patients (31.25%), and in anterior horn and central gray matter in 5 patients (31.25%). The initial scan of 7 patients showed cord expansion (43.7%). Owl's eye pattern was noted in 6 patients (37.5%). None of these patients had involvement of peripheral territory of the ASA. Signal changes were hyperintense on T2 and isointense on T1 in all patients. DWI was available in 11 studies and restriction was noted in 8 studies: these studies were performed between 3 hours to 21 days after onset of symptoms. Post contrast T1 W images were available in 16 studies (immediate and follow-up MRI); enhancement was noted in 2 studies when MRI was performed within 8 days and 17 days after onset of symptoms in 2 different patients. One of our cases showed vertebral body marrow signal changes attributed to ischemia. Five patients had follow-up scans with period ranging from 4 days to 1 year; these scans showed T2 hyperintensity and T1 isointensity at their respective levels on follow-up. One patient showed cord atrophy in 12-month follow-up scan. One patient with PSA infarct had T2 hyperintense, T1 isointense signal change in left posterior paramedian triangular distribution in the cervical cord (C2-C7), without expansion/enhancement of cord. DWI was not available for this patient. All imaging features are summarized in Tables 2 and 3. Representative images of ASA and PSA infarcts are shown in Figure 1.
Discussion Spinal cord has an intricate network of blood supply. Various texts and articles describe the anatomy in exquisite details.11-15 Patients with SCI develop symptoms quickly with maximal symptomatology reached within 12 hours for 50% of patients and within 72 hours for most patients.16 The symptom onset to peak time ranged from few minutes to 48 hours in our study. All 16 patients with ASA infarct in our study presented with dissociative anesthesia and weakness of upper limbs/lower limbs depending on the level of lesion in the cord; back pain or neck pain was noted in 10 patients (n = 10, 62.5%). Additionally, autonomic symptoms were noted in 10 patients (n = 10, 62.5%). The clinical presentation is defined by the vascular territory involved with the severity of the impairment varying widely, from paraplegia to minor weakness.
5
Back pain often accompanies SCI and has been reported in as many as 70% of patients typically at the level of the lesion.16 Following are the clinical patterns of ASA infarct17
1. Typical ASA infarct pattern: with involvement of anterior two-thirds of cord involving gray and white matter, causing bilateral weakness, dissociative sensory loss with or without autonomic symptoms. 2. Incomplete pattern: a) Involvement localized at the level of the anterior horns. Clinical manifestation may be in form of: i) Acute paraplegia (pseudopoliomyelitic form) without sensory abnormalities and without sphincter dysfunction. ii) Painful bilateral brachial diplegia in the case of a cervical lesion (the MIB syndrome).18 iii) Progressive distal amyotrophy due to chronic lesions of the anterior horns; this form may be misdiagnosed as lateral amyotrophic sclerosis. b) Sulcocommissural syndrome presents as a partial Brown-Sequard syndrome with sparing of postural sensibility. The syndrome consists of hemiparesis with a contralateral spinothalamic sensory deficit. c) Central spinal infarct occurs after cardiac arrest or prolonged hypotension; its clinical presentation includes bilateral spinothalamic sensory deficit with sparing of the posterior columns. Motor deficit and sphincter dysfunction are usually absent. 3. Another rare presentation is transverse medullary infarction (full transverse lesions) presenting with sudden inability to walk due to paraplegia/paraparesis or tetraplegia/tetraparesis (in higher cord lesions). There may be complete sensory loss involving all modalities and pain reported as circumferential tightness. Autonomic dysfunction is noted. Usually it is of embolic origin. Interestingly 1 patient, a 12-year-old girl, presented with sudden onset weakness of bilateral upper limbs, with onset to peak time of 5 minutes. There was no weakness in lower limbs and no bulbar symptoms. This type of presentation is known as acute brachial diplegia (MIB syndrome). MIB syndrome has been ascribed to supratentorial brain lesions mostly in patients with lesions at the borderzone between anterior and middle cerebral artery
6
Table 2. MRI findings at presentation Time from onset to MRI S. no.
2 3
4 days 8 days
4 5
17 h 2 days
6 7
5 month 2h
8
36 h
9
30 h
10 11
19 h 35 h
12
36 h
13 14
4 days 30 h
15
21 days
16
48 h
17
5 months
C6-D4
Anterior two-thirds cord (gray matter + white matter) C4-C7 Anterior horn D5-D12 Anterior two-thirds cord (gray matter + white matter) C3-C5 Anterior horn D9 till conus Anterior horn + central gray matter C5-C6 Anterior horn C2-C7 Anterior horn + central gray matter C2- D4 Anterior two-thirds cord (gray matter + white matter) C3-D3 Anterior two-thirds cord (gray matter + white matter) C3-D1 Anterior horn C5-D1 Anterior two-thirds cord (gray matter + white matter) C4-C6 Anterior two-thirds cord (gray matter + white matter) D4 till conus Anterior horn D4 till conus Anterior horn + central gray matter D1-D5 Anterior horn + central gray matter C3- C7 Anterior horn + central gray matter C2-C7 Left paramedian triangular
T1
T2
DWI
Contrast
Isointense Hyperintense Not available
Not available
Yes
Yes
No
No
Isointense Hyperintense Not available Isointense Hyperintense Diffusion restriction present
Not available Enhancement present
No No
Yes Yes
Yes No
No No
Isointense Hyperintense Not available Isointense Hyperintense Diffusion restriction present Isointense Hyperintense Not available Isointense Hyperintense Diffusion restriction present Isointense Hyperintense Diffusion restriction present
No enhancement No enhancement
No No
Yes Yes
No No
No No
No enhancement Not available
No Yes
Yes Yes
Yes No
No No
No enhancement
Yes
Yes
Yes
No
Isointense Hyperintense Diffusion restriction present
No enhancement
Yes
Yes
Yes
No
Isointense Hyperintense Not available Isointense Hyperintense Diffusion restriction present
No enhancement No enhancement
Yes No
Yes Yes
Yes No
No No
Isointense Hyperintense No diffusion restriction
No enhancement
Yes
Yes
No
No
Isointense Hyperintense Not available Isointense Hyperintense Not available
Not available Not available
No Yes
Yes Yes
Yes No
No No
Isointense Hyperintense No diffusion restriction Isointense Hyperintense Diffusion restriction present Isointense Hyperintense Not available
No enhancement
No
Yes
No
Yes
No enhancement
No
Yes
No
No
No enhancement
No
Yes
No
No
ARTICLE IN PRESS
4 days
Location
Vertebral Cord Pencil thin Owl’s eye body signal changes expansion appearance appearance
N. YADAV ET AL.
1
Level of ischemia
ARTICLE IN PRESS
No Yes No
No
No Yes
Isointense Hyperintense Diffusion restriction present 5 days 14
D4 to conus 17 days 13
C4-C7 C3-D1 9 months 1 year 2 9
D4 till conus Anterior horn + central gray matter
Isointense Hyperintense No diffusion restriction
Enhancement present (enhancement of anterior horn D12-L1 with cauda equine anterior roots enhancement) No enhancement
No
Yes
No No Yes Yes Yes Yes No Atrophy Isointense Hyperintense Not available Not available Isointense Hyperintense Not available No enhancement
No No Yes No Isointense Hyperintense Not available No enhancement
Anterior two-thirds cord (gray matter + white matter) Anterior horn Anterior two-thirds cord (gray matter + white matter) Anterior horn 2 months 1
C7-D2
DWI T2 T1 Location Time from onset to MRI
Level of ischemia
7
Case no.
Table 3. Follow-up MRI findings
Contrast
Vertebral Cord Pencil thin Owl’s eye body signal changes expansion appearance appearance
CLINICAL AND RADIOLOGICAL FEATURES
territories including the prerolandic cortex; etiologies may be varied.18-21 Spinal cord infarct as a cause of MIB is much less common than cerebral lesions. The literature has less than 10 cases of MIB syndrome caused by cervical infarcts. The clinical pattern of PSA infarction: Because of presence of two posterior spinal arteries, this syndrome is usually unilateral and less severe. Based on the functional neuroanatomy, this syndrome leads to ipsilateral loss of light touch, vibration, and proprioception while mostly sparing the motor function. However, there can be segmental deep tendon areflexia due to posterior horn involvement and paresis below the level at which the posterior portion of the cord is affected. Among the 17 patients with SCI in our study, the underlying etiologies include atherosclerosis (n = 1), vertebral artery thrombosis (n = 1), iatrogenic (n = 3, post superior cerebellar artery aneurysm coiling, postclipping of middle cerebral artery bifurcation aneurysm in patient with coarctation of aorta, and iatrogenic vertebral artery injury), posttraumatic (n = 1, patient with PSA infarct). Causes of SCI differ in the pediatric and adult populations. In children, the most common causes are cardiac malformations and trauma. In adults, atherosclerosis is one of the principal causes, but thoracoabdominal aneurysms, aortic surgery, embolic disease, dissection, systemic hypotension, spinal arteriovenous malformations, diving, coagulopathies, cocaine abuse, and sickle cell disease are implicated (17). In many cases, the cause remains unknown and these are categorized as idiopathic; as was noted in 11 of our cases. Our institute does not have a vascular surgery department explaining the lack of cases attributed to aortic diseases as cause of SCI. One of our patient (case no. 13) developed SCI following clipping of middle cerebral artery bifurcation aneurysm; she had documented intraoperative hypotension complicated additionally by presence of coarctation of aorta, possibly making the patient more susceptible to the effect of hypotension (Fig 2). Another case of craniovertebral junction surgery (case no. 14) with iatrogenic vertebral artery injury also had documented intraoperative hypotension. Both these patients with hypotension as the etiology developed infarct in dorsolumbar cord (D4 till conus). Historically, a spinal cord “watershed zone” of ischemic vulnerability centered at the midthoracic level (D4D6 level) has been suggested.22-24 This assumption was mainly based on anatomic studies and case reports describing the relative hypovascularity of the midthoracic region (D4-D8),14-23 which was then equated with increased susceptibility of this area to ischemic injury. However, there have been studies suggesting that global ischemia may affect the low thoracic and lumbosacral cord to a greater extent than the other levels of the spinal cord.25-28
ARTICLE IN PRESS 8
N. YADAV ET AL.
Figure 1. Typical imaging appearance of anterior spinal artery (different patients) (a-f) and posterior spinal artery infarct (g, h). T2W sagittal image (a) shows pencil thin hyperintensity in anterior aspect of cervical cord. T2W axial images (b-d) show the patterns of T2 hyperintensity involving anterior horn cells (owl's eye) in (b), anterior two thirds of spinal cord (c), and anterior horn and central gray matter (d). Diffusion weighted image (e) and apparent diffusion coefficient (ADC) map (f) showing restricted diffusion involving cord in anterior spinal artery territory. Posterior spinal artery infarct: Left posterior paramedian triangular T2 hyperintensity on axial T2 image (g) and associated brachial plexus injury with pseudomeningocele formation on STIR coronal (h).
The findings in the study done by Duggal and Lach also indicated a greater vulnerability of lumbosacral neurons to ischemic insults from either cardiac arrest or a severe sustained episode of hypotension.29 They proposed that this is most likely a consequence of greater metabolic demands of the gray matter at this level of the spinal cord. Our findings in patient with SCI secondary to hypotension are also in concordance with this. In our study, the involvement was noted in cervical cord (n = 6, 37.5%), cervicodorsal cord (n = 5, 31.3%), dorsal cord (n=2, 12.5%), and dorsolumbar cord (n = 3, 18.7%). The ASA syndrome is divided into three clinical pictures according to the particular cord segment involved: cervicothoracic, middle thoracic, or thoracolumbar. In previously published studies, thoracolumbar infarction has been considered as the most frequent and the best clinically defined syndrome.30 However, we noted cervical and cervicodorsal regions as the more common location. This, as stated earlier, can be attributed to lack of aortic surgeries at our institute as a cause and hence the given distribution of lesions. The sensitivity of initial MRI of the cord is limited, nearly 17 45% of clinically suspected SCI cases have a normal scan. More importantly, MRI rules out other causes presenting as acute spinal syndrome, such as compressive myelopathies, vascular malformations, infective myelitis, demyelinating disorders, and tumor.31
Literature describes MRI patterns in patients with ASA infarcts as “owl eye” or “snake eye” appearance on T2 weighted (T2W) axial and pencillike hyperintensities on T2W sagittal images.32 The grey matter of the anterior horns exhibit the highest vulnerability to ischemia due to high metabolic demand that may lead to the typical ‘‘snake-eye’’ or “owl-eye” appearance of cord infarction on axial T2W images7 as noted in 6 (37.5%) of our patients. ‘‘Pencillike’’ hyperintensity in sagittal T2W images was seen in all 16 of our patients with ASA infarct, similar to the study results of Yuh et al33 and Stefan Weidauer et al.7 Seven (43.7%) of our patients showed cord expansion that has been described in early stages. In the acute or subacute stage, DWI was available in 11 studies and diffusion restriction was noted in 8 studies; the earliest study being performed 3 hours and the longest one was performed 21 days after onset of symptoms. Out of these 11 cases, 3 cases that did not show restriction were done on the 17th day and 21st day after onset of symptoms. The shortest time reported in the literature between the onset of clinical symptoms and abnormalities on DWI is 3 hours,8 which was also noted in our study in 1 patient (case no. 7). In the brain, DWI can demonstrate restriction as early as 20-30 minutes after the onset of symptoms. However, larger studies are needed to establish a time threshold for diffusion changes in patients with spinal cord ischemia/infarction. Similar to our study, in most of
ARTICLE IN PRESS CLINICAL AND RADIOLOGICAL FEATURES
the previously reported patients, signal abnormalities were demonstrated on T2-W images and on DWI. However, in previously described studies, conventional Magnetic resonance (MR) sequences did not show abnormality in 4 reported patients in 3, 4, 10, and 46 hours after the onset of spinal ischemia10,34,35; DWI might have helped in these cases. A number of technical limitations exist for DWI of spinal cord including field inhomogeneity, CSF and blood pulsations, and small size of infarcts.36 Though it seems to be established that SCI can be identified using DW-MRI, the dynamic evolution of diffusion abnormalities remains to be identified.8,37 Enhancement is usually noted in the subacute stage. In our study, enhancement was noted in 2 studies (case nos. 3 and 13), when MRI was performed within 8 days and 17 days after onset of symptoms. Interestingly, in case no. 13, the study performed on the 4th day following symptoms did not show enhancement. However, MRI performed on the same patient on the 17th day showed enhancement of anterior horn at D12 to L1 with enhancement of anterior roots of cauda equina (Fig 2). The concomitant enhancement of the cauda equina was first reported by Friedman and Flanders in 1992.38 This phenomenon was described as a characteristic finding in the
9
course of SCI that might involve the cord and the ventral cauda equina as well (which comprises of motor fiber bundles).39 It denotes disruption of the blood cord barrier as well as reactive hyperemia.38-40 One of our cases showed vertebral body marrow signal changes attributed to ischemia (Fig 3). Studies have demonstrated that a finding of vertebral body infarction adjacent to a cord signal change on MRI is a useful confirmatory sign41,42 although found in only 4%-35% of the patients (since arterial occlusion may be located distal to the anterior or posterior central arteries that supply the vertebral body).31 The vertebral bone marrow abnormality that tends to occur earlier is more exaggerated compared to the cord and is more commonly seen in ASA syndromes as compared to PSA syndromes.42 Appropriate clinical setting and cord signal changes are indicative of an ischemic etiology as cause of vertebral marrow signal change. In our study, 1 patient with PSA infarct had T2 hyperintense, T1 isointense signal change in left posterior paramedian triangular distribution in the cervical cord (C2-C7), without the presence of expansion/enhancement of the cord. Diffusion sequence was however not available for this patient.
Figure 2. Case no. 14. Thirty-year-old female who presented with sudden onset headache followed by generalized seizures. Non contrast CT head (a) shows the presence of Sub arachnoid hemorrhage (SAH) in basal cisterns, sylvian fissures, and interhemispheric fissure. During cerebral angiography, the catheter could not be negotiated beyond the thoracic aorta and a pigtail injection at this point (b) showed the presence of coarctation of aorta. A CT angiography (c) revealed the presence of left MCA bifurcation aneurysm. Surgery for clipping of aneurysm was complicated by intraoperative hypotension; postoperative paraplegia was noted. Dorsal spine T2 sagittal (d) and T2 axial images (e) show pencil thin hyperintensity involving cord from D4 till conus with hyperintensity involving anterior horn and central gray matter. Contrast enhanced T1W sagittal (f) and axial images (g, h) done on postoperative day 17 shows characteristic contrast enhancement involving anterior horns and ventral cauda equina. String of pearls appearance of the ventral nerve roots is noted in fig h.
ARTICLE IN PRESS N. YADAV ET AL.
10
Figure 3. Case no. 16. Twenty-six-year-old female presented with sudden onset back pain, followed by weakness of bilateral lower limbs (left followed by right) with onset to peak time of 4 hours. T2W sagittal (a) and axial images (b) show the presence of pencil thin hyperintensity in upper dorsal cord and hyperintensity involving anterior horns respectively. T2W image in parasagittal location (c) shows the presence of vertebral marrow hyperintensity involving D4 vertebra (white arrow). T2W axial image at D4 (d) shows the cord hyperintensity with associated vertebral marrow wedge shaped hyperintensity on left side. Corresponding diffusion weighted imaging (e) showing restriction in the cord. The marrow signal change was isointense on T1W axial image (f) and shows contrast enhancement (white arrow) on post contrast T1W axial image (g).
PSA syndrome is rare and to our knowledge only about 30 cases have been reported so far.5,43 Williamson first reported an autopsy case of PSA syndrome in 1895.44 Described causes of PSA infarct include syphilitic arteritis, cholesterol emboli from atheromatous aortic plaques, intrathecal injection of phenol, vertebral artery dissection, and plasmacytoma.6 In our patient, PSA infarct was attributed to trauma due to road traffic accident resulting in associated ipsilateral brachial plexus injury. The infarct was noted in the cervical cord spanning 6 segments (C2-C7). The majority of described PSA infarcts occur at the thoracolumbar level but they have been observed at the thoracic and cervical levels too.43 Longitudinal extension averages 2 vertebral segments but their span can range from 1-6 segments,43 as was also noted in our study (C2-C7). Appropriate clinical setting and cord signal changes will support the diagnosis of a PSA infarct.
Clues to Identification of SCI Include Clinical Sudden onset of symptoms with onset to peak duration less than 72 hours. Symptoms varying depending on territory, location and extent of infarction.
Imaging Different imaging modalities might give a clue to the diagnosis depending on the clinical scenario (eg Computed tomography (CT) angiography done in patient with abdominal pain and hypotension may reveal aortic dissection and CT spine done in case of cervical spine trauma may reveal vertebral artery injury). However, if SCI is suspected, MRI is essential for diagnosis and should include the following sequences: T2W Sagittal, T2W axial, Short tau inversion recovery (STIR) sagittal, DWI, contrast enhanced T1W sagittal, and axial sequences. Specifically we look for pencil thin hyperintensity, owl's eye appearance (ASA infarct), and posterior paramedian triangular hyperintensity (PSA infarct). Marrow signal changes/diffusion restriction/contrast enhancement can be noted in ASA/PSA infarct depending on time of imaging from onset. Follow-up of patients with ASA infarct was available for 9 patients. Of these, 4 could walk independently, 1 patient could walk with support, 2 patients were wheelchair bound, 1 patient died due to unknown cause, and 1 succumbed to respiratory failure with involvement of cord above C3 (C2-C7). Previous studies have advocated various factors in predicting prognosis of patients of SCI. Severe neurological deficits at onset and involvement of
ARTICLE IN PRESS CLINICAL AND RADIOLOGICAL FEATURES
proprioception were associated with poor outcome, whereas the presence of owl's eye sign was associated with a good outcome.2 Management largely depends on the underlying cause of infarction. Emergency surgery is needed if acute aortic event, vascular compression is the underlying cause of spinal ischaemia. In the event of global hypoperfusion, maintaining adequate blood pressure to maintain spinal perfusion is crucial. Antiplatelet therapy is needed to prevent vascular occlusions or embolism. Corticosteroids are indicated in cases of vasculitis or aortitis. Novel findings in our study include: depiction of spinal cord infarct on DWI at 3 hours after symptom onset (3 hours being the earliest in literature for DWI to demonstrate the signal changes). Next is a girl presenting with “(wo)man-in-the-barrel syndrome” due to cervical cord infarct; literature describes less than 10 such cases. Another observation is of concomitant enhancement of anterior horn and cauda equina noted in one of our patients; this adds to the limited literature on this characteristic finding in the course of spinal cord ischemia. Further, 2 patients developed cord ischemia secondary to hypotension in dorsolumbar cord, which in accordance with studies suggesting that global ischemia may affect the low thoracic and lumbosacral cord to a greater extent than the other levels of the spinal cord as opposed to historically described spinal cord “watershed zone” of ischemic vulnerability centered at the midthoracic level (D4-D6 level). However, our study also had certain limitations. Ours being only a neurosciences tertiary referral center, relative late presentation to the hospital was common. Diffusion sequence was only available for 11 cases: the imaging protocols have now been modified. Follow-up was available only for 9 patients. In conclusion, ASA and PSA infarcts are rare but important causes of acute spinal syndrome. A high index of suspicion is essential to identify these lesions. The earliest change is noted on DWI sequence while enhancement is noted in the subacute stage. Typical distribution and imaging appearance of lesions with signal changes evolving over time on various sequences can help in differentiating SCI from other etiologies. Spinal imaging with addition of DWI can thus help in diagnosing this uncommon entity early in the disease course.
References 1. Lynch K, Oster J, Apetauerova D, Hreib K. Spinal cord stroke: acute imaging and intervention. Case Rep Neurol Med 2012;2012:1-3. 2. Masson C. Spinal cord infarction: clinical and magnetic resonance imaging findings and short term outcome. J Neurol Neurosurg Psychiatry 2004;75:1431-1435. 3. de la Barrera SS, Barca-Buyo A, Montoto-Marques A, et al. Spinal cord infarction: prognosis and recovery in a series of 36 patients. Spinal Cord 2001;39:520.
11 4. Millichap JJ, Sy BT, Leacock RO. Spinal cord infarction with multiple etiologic factors. J Gen Intern Med 2007;22:151-154. 5. Kaneki M, Inoue K, Shimizu T, et al. Infarction of the unilateral posterior horn and lateral column of the spinal cord with sparing of posterior columns: demonstration by MRI. J Neurol Neurosurg Psychiatry 1994;57:629-631. 6. Matsubayashi J, Tsuchiya K, Shimizu S, et al. Posterior spinal artery syndrome showing marked swelling of the spinal cord: a clinico-pathological study. J Spinal Cord Med 2013;36:31-35. 7. Weidauer S, Nichtweiss M, Lanfermann H, et al. Spinal cord infarction: MR imaging and clinical features in 16 cases. Neuroradiology 2002;44:851-857. 8. Thurnher MM, Bammer R. Diffusion-weighted MR imaging (DWI) in spinal cord ischemia. Neuroradiology 2006;48:795-801. 9. Sibon I, Menegon P, Moonen CTW, et al. Early diagnosis of spinal cord infarct using magnetic resonance diffusion imaging. Neurology 2003;61. 1622 1622. 10. Fujikawa A, Tsuchiya K, Takeuchi S, et al. Diffusionweighted MR imaging in acute spinal cord ischemia. Eur Radiol 2004;14:2076-2078. 11. Amato ACM, Stolf NAG. Anatomia da circula¸c ~ ao medular. J Vasc Bras 2015;14:248-252. 12. Williams GM, Perler BA, Burdick JF, et al. Angiographic localization of spinal cord blood supply and its relationship to postoperative paraplegia. J Vasc Surg 1991;13:23-35. 13. Martirosyan NL, Feuerstein JS, Theodore N, et al. Blood supply and vascular reactivity of the spinal cord under normal and pathological conditions: a review. J Neurosurg Spine 2011;15:238-251. 14. Hassler O. Blood supply to human spinal cord. A microangiographic study. Arch Neurol 1966;15:302-307. 15. Prince E, Ahn S. Basic vascular neuroanatomy of the brain and spine: what the general interventional radiologist needs to know. Semin Interv Radiol 2013;30:234-239. 16. Novy J, Carruzzo A, Maeder P, et al. Spinal cord ischemia: clinical and imaging patterns, pathogenesis, and outcomes in 27 patients. Arch Neurol 2006;63:1113-1120. 17. Vargas MI, Gariani J, Sztajzel R, et al. Spinal cord ischemia: practical imaging tips, pearls, and pitfalls. Am J Neuroradiol 2015;36:825-830. 18. Antelo MJG, Facal TL, S anchez TP, et al. Man-in-the-barrel. A case of cervical spinal cord infarction and review of the literature. Open Neurol J 2013;7:7-10. 19. Berg D, Miillges W, Koltzenburg M, et al. Case report man-in-the-barrel syndrome caused by cervical spinal cord infarction. Acta Neurol Scand 1998;97:417-419. 20. Ben Sassi S, El Euch G, Regaieg A, et al. Man-in-the-barrel syndrome with combination of infarctions in the anterior spinal artery and posterior inferior cerebellar artery territories. Cerebrovasc Dis 2009;27:201-202. 21. Rouanet C, Reges D, Rocha E, et al. “Man in the barrel” syndrome with anterior spinal artery infarct due to vertebral artery dissection. J Stroke Cerebrovasc Dis 2017;26: e41-e42. 22. Bartsch W, Hopf HC. New observations on the relations between cardiac performance and spinal cord blood supply. Dtsch Z Nervenheilkd 1963;184:288-307. 23. Marcus ML, Heistad DD, Ehrhardt JC, et al. Regulation of total and regional spinal cord blood flow. Circ Res 1977;41:128-134. 24. Sliwa JA, Maclean IC. Ischemic myelopathy: a review of spinal vasculature and related clinical syndromes. Arch Phys Med Rehabil 1992;73:365-372.
ARTICLE IN PRESS 12 25. Imaizumi H, Ujike Y, Asai Y, et al. Spinal cord ischemia after cardiac arrest. J Emerg Med 1994;12:789-793. 26. Cheshire WP, Santos CC, Massey EW, et al. Spinal cord infarction: etiology and outcome. Neurology 1996;47: 321-330. 27. Gilles FH, Nag D. Vulnerability of human spinal cord in transient cardiac arrest. Neurology 1971;21:833-839. 28. Azzarelli B, Roessmann U. Diffuse “anoxic” myelopathy. Neurology 1977;27:1049-1052. 29. Duggal N, Lach B. Selective vulnerability of the lumbosacral spinal cord after cardiac arrest and hypotension. Stroke 2002;33:116-121. 30. Monteiro L, Leite I, Pinto JA, et al. Spontaneous thoracolumbar spinal cord infarction: report of six cases. Acta Neurol Scand 1992;86:563-566. 31. Alexander M, Patil A, Mani S, et al. Spectrum of clinicoradiological findings in spinal cord infarction: report of three cases and review of the literature. Ann Indian Acad Neurol 2013;16:190. 32. Cheng M-Y, Lyu R-K, Chang Y-J, et al. Spinal Cord Infarction in Chinese patients. Cerebrovasc Dis 2008;26:502-508. 33. Yuh WT, Marsh EE, Wang AK, et al. MR imaging of spinal cord and vertebral body infarction. Am J Neuroradiol 1992;13:145-154. 34. Weidauer S, Dettmann E, Krakow K, et al. Diffusionsgewichtete MRT bei spinalen Infarkten. Nervenarzt 2002;73:999-1003. 35. Loher TJ, Bassetti CL, Lovblad KO, et al. Diffusionweighted MRI in acute spinal cord ischaemia. Neuroradiology 2003;45:557-561.
N. YADAV ET AL. 36. Kuker W, Weller M, Klose U, et al. Diffusion-weighted MRI of spinal cord infarction: high resolution imaging and time course of diffusion abnormality. J Neurol 2004;251:818-824. Available from: http://link.springer. com/10.1007/s00415-004-0434-z. 37. Bammer R, Fazekas F, Augustin M, et al. Diffusionweighted MR imaging of the spinal cord. Am J Neuroradiol 2000;21:587-591. 38. Friedman DP, Flanders AE. Enhancement of gray matter in anterior spinal infarction. Am J Neuroradiol 1992;13:983-985. 39. Amano Y, Machida T, Kumazaki T. Spinal cord infarcts with contrast enhancement of the cauda equina: two cases. Neuroradiology 1998;40:669-672. 40. Mull M, Thron A. Spinal infarcts. In R€ udiger von Kummer and Tobias Back (Eds.). Magnetic Resonance Imaging in Ischemic Stroke, Springer-Verlag Berlin Heidelberg 2006, 251-267. 41. Faig J, Busse O, Salbeck R. Vertebral body infarction as a confirmatory sign of spinal cord ischemic stroke. Stroke 1998;29:239-243. 42. Suzuki T, Kawaguchi S, Takebayashi T, et al. Vertebral body ischemia in the posterior spinal artery syndrome: case report and review of the literature. Spine 2003;28: E260-E264. 43. Mascalchi M, Cosottini M, Ferrito G, et al. Posterior spinal artery infarct. Am J Neuroradiol 1998;19:361-363. 44. Williamson RT. Spinal softening limited to the parts supplied by the posterior arterial system of the cord. Lancet 1895;146:520-522.