Pain-Topics.org News/Research UPDATES: Stress, the Brain, & Chronic Pain

Pain-Topics.org News/Research UPDATES: Stress, the Brain, & Chronic Pain

Friday, March 1, 2013

Stress, the Brain, & Chronic Pain

Patients with chronic or recurring pain know how stressful their conditions can become. New research suggests that there may be physiologic causes of such stress, mediated by changes in brain structure that make these persons even more vulnerable to stress and pain. While there were many limitations of the study, there are implications for the importance of stress-reduction techniques in helping to better manage chronic pain.
The study — conducted primarily by Canadian researchers at the Université de Montréal — included 16 patients with chronic back pain and a control group of 18 healthy subjects [Vachon-Presseau et al. 2013]. The research objective was to analyze relationships between 4 factors: 1) cortisol levels, which were determined via saliva samples collected during 7 days; 2) assessments of clinical pain reported by patients prior to brain scans (self-perception of pain); 3) the size of brain hippocampal regions measured using magnetic resonance imaging (MRI); and 4) brain activations assessed with functional MRI (fMRI) following application of a thermal pain stimulus.
The results, reported in the journal Brain, showed that patients with chronic pain generally have higher cortisol levels than healthy individuals. Cortisol, produced by the adrenal glands, is sometimes called the “stress hormone” since it is activated in reaction to stress. It is a steroid hormone that increases blood sugar, suppresses the immune system, and may decrease bone formation; so, an excess is deleterious to health.
Data analysis further revealed that higher cortisol also was associated with smaller hippocampal volume and stronger pain-evoked activity in the anterior parahippocampal gyrus, a region involved in anticipatory anxiety and associative learning. Brain activity in response to the painful stimulus, assessed during fMRI scanning, reflected the intensity of the patient's current clinical pain condition.
The authors conclude that their findings support a “stress model of chronic pain,” suggesting that the sustained endocrine stress response observed in individuals with a smaller hippocampus induces changes in the function of the hippocampal complex, which in turn may contribute to increased and persistent pain states. Such findings imply that stress management interventions might be important adjunctive treatments in the management of chronic pain.
COMMENTARY: This study helps to elucidate understandings of the neurobiological mechanisms affecting chronic pain and the vital role of stress. It supports recent theories suggesting that chronic pain could be partly maintained by maladaptive neurophysiological responses to recurrent stressors.
We have previously discussed how structural and functional alterations in the brains of persons with various chronic pain syndromes may influence the persistence and intensity (and possibly the initial development) of their conditions. See the series of pain-and-the-brain UPDATES [here]. This present study examines a possible role of the hippocampus, a component of the limbic system that plays important roles in the consolidation of information from short-term memory to long-term memory as well as spatial navigation. Its importance for influencing the experience of pain has been less well documented.
Vachon-Presseau and colleagues [2013] suggest that their findings “open up avenues for people who suffer from pain to find treatments that may decrease its impact and perhaps even prevent chronicity. To complement their medical treatment, pain sufferers can also work on their stress management and fear of pain by getting help from a psychologist and trying relaxation or meditation techniques.”
Indeed, nonpharmacologic modalities — eg, relaxation, meditation, biofeedback, cognitive behavioral therapy, yoga, massage, and other stress and anxiety reduction techniques — may be important adjuncts for effective pain management. For example, a prior UPDATE [here] described how research on meditation found that this approach favorably influenced how participants’ brains responded to pain. In particular, as a result of meditation, multiple limbic structures reacted to pain in ways that moderated subjects’ processing of painful stimuli and their perceptions of pain. Therefore, it makes sense that structural hippocampus changes found in this present experiment by Vachon-Presseau could have an impact.
At the same time, however, there were some important limitations of this present research. For one thing, sample sizes were quite small — 16 in the experimental group, 18 controls — which raises questions about statistical veracity (eg, random effects), whether participants were truly representative of the larger population in question, and external validity. Small sample sizes are typical of imaging studies in the pain field, but larger-scale trials needed to substantiate reliability of outcomes are rarely performed.
Furthermore, causation and temporality cannot be confirmed by this study. That is, which came first, stress and an undesirable cortisol response or chronic pain? Did alterations of brain structure precede chronic pain or develop in response to it? Lastly, it would be important to know if the cortisol and brain-structure aberrations can be reversed by therapeutic approaches — whether pharmacologic or nonpharmacologic — and the ultimate effect of this remediation on chronic pain.
REFERENCE: Vachon-Presseau E, Roy M, Martel M-O, et al. The stress model of chronic pain: evidence from basal cortisol and hippocampal structure and function in humans. Brain. 2013;136(3):815-827 [abstract].

Latest Article NEJM Dr Oaklander, researcher Chosen By NORD and funded by the tarlovcystfoundation by fundraising and Membership

Photoswill show in the link itself,

The new england journal of medicine
case records of the massachusetts general hospital
Eric S. Rosenberg, m.d., Editor Alice M. Cort, m.d., Associate Editor Emily K. McDonald, Assistant Editor
Case 7-2013: A 77-Year-Old Woman with Long-Standing Unilateral Thoracic Pain and Incontinence
Anne Louise Oaklander, M.D., Ph.D., Donlin M. Long, M.D., Ph.D., Mykol Larvie, M.D., Ph.D., and Christian J. Davidson, M.D.
PRESENTATION OF CASE
A 77-year-old woman was seen in the outpatient neurology clinic at this hospital because of chronic thoracic pain.
The patient had had constant pain under her left breast for as long as she could remember, probably since adolescence, which worsened over time and spread hori- zontally to involve a band on the entire left side of the chest. Multiple evaluations performed at other hospitals throughout her adult life had been unrevealing. Eight years before this evaluation, magnetic resonance imaging (MRI) of the spine was reported as normal, with an incidental finding of small cysts associated with neural foramina on the right side. One thoracic injection and three lumbar epidural injec- tions of local anesthetic agents and glucocorticoids administered 2 years before this evaluation had only transiently improved the pain. Four months after the injections, MRI reportedly revealed a lesion (10 mm in the greatest dimension) in the left T7–T8 neural foramen, causing mild expansion and remodeling of the foramen; these findings were consistent with a perineurial cyst. One month later, a second MRI of the thoracic spine, performed with contrast material, reportedly revealed bilateral T7–T8 neural foraminal cysts measuring up to 8 mm in the greatest di- mension and smaller foraminal lesions, suspected to be bilateral, at T10–T11, T11–T12, T12–L1, and low cervical levels. A radiograph of the chest revealed slight aortic dilatation.
Twenty months before evaluation at this hospital, computed tomography (CT) after the intrathecal administration of contrast material (CT myelography) performed at a second hospital reportedly revealed a perineurial cyst (10 mm by 8 mm) at the T7–T8 level on the left side that became filled with contrast material. Eight months before this evaluation, repeat MRI reportedly revealed a perineurial cystic lesion at the T7–T8 level on the left side, with 10 or more additional cysts between T2 and T10. Four months before this evaluation, the patient sought consultation at the Johns Hopkins Hospital.
The patient rated the thoracic pain as ranging in severity from 4 to 9 on a scale
n engl j med 368;9 nejm.org february 28, 2013

Founded by Richard C. Cabot Jo-Anne O. Shepard, m.d., Associate Editor

Nancy Lee Harris, m.d., Editor Sally H. Ebeling, Assistant Editor

The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on March 1, 2013. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.

From the Departments of Neurology (A.L.O.), Radiology (M.L.), and Pathology (A.L.O., C.J.D.), Massachusetts General Hospital, and the Departments of Neu- rology (A.L.O.), Radiology (M.L.), and Pa- thology (C.J.D.), Harvard Medical School — both in Boston; and the Department of Neurosurgery, Johns Hopkins Univer- sity, Baltimore (D.M.L.).
N Engl J Med 2013;368:853-61.
DOI: 10.1056/NEJMcpc1114034
Copyright © 2013 Massachusetts Medical Society.

853

of 0 to 10, with 10 indicating the worst pain. She reported that it improved when she was lying down and was not altered by coughing, strain- ing, or palpation. It had not been relieved by physical therapy, chiropractic management, or medications, including gabapentin and transder- mal lidocaine. She also reported occasional left groin and sacral pain (the pain did not radiate and was associated with normal pudendal sensa- tion), intermittent numbness in the left leg, bilat- eral foot pain, and episodic fecal and urinary in- continence that required protective undergarments. She had severe bilateral hearing loss for which she wore hearing aids, mild hypoproliferative anemia, hypertension, hypothyroidism, hyper- cholesterolemia, gastritis, attention deficit–hyper- activity disorder, mild chronic obstructive pul- monary disease, dysthymia, and severe fatigue. During childhood, she had had a prolonged ill- ness with streptococcal pharyngitis. She had had three fractures of her arms after falling, repair of a rectal fistula, and additional minor surgical procedures. Medications included metoprolol, levothyroxine, lisinopril, omeprazole, dextroam- phetamine–amphetamine, desvenlafaxine, clonaz- epam, atorvastatin, aspirin, and multiple vitamins (including vitamin D and numerous daily doses of vitamin B complex), as well as acetaminophen for pain. She had no known allergies.
The patient was a retired nurse practitioner who lived with her husband in Massachusetts. She had smoked cigarettes for decades and did not drink alcohol or use illicit drugs. Her mother died at 75 years of age of multiple myeloma, her father at 81 years of age of congestive heart fail- ure and dementia, and a brother of pulmonary fibrosis. Among her four children, a son had been born prematurely at 31 weeks of gestation, a daughter had colon cancer and iliac-artery dis- section and required surgical mitral-valve repair at 49 years of age, another son had survived testicular cancer in youth, and another son had Asperger’s syndrome and mental retardation. A grandson was 206 cm tall and had a thyroid cyst. There was no family history of neurologic or pain conditions.
On examination, the patient was alert and oriented. The height was 178 cm and the weight 67 kg. She had arachnodactyly and normal joint flexibility. She ambulated slowly without assis- tance and was unable to walk with a tandem gait. She had mild allodynia over the left thorax

that was provoked by touch, and sensation on pin-prick testing was normal. Achilles’ tendon reflexes were not present; other reflexes were normal. The remainder of a detailed neurologic examination was normal.
A diagnostic and therapeutic procedure was performed, after which the patient was referred to the neurology clinic at this hospital.
DIFFERENTIAL DIAGNOSIS
Dr. Anne Louise Oaklander: I am aware of the diag- nosis in this case. The cause for this patient’s long-standing, left-predominant chest pain had been investigated over decades, as documented in hundreds of pages of medical records. The patient was a nurse practitioner who was motivated to pursue diagnosis and effective treatment. She had undergone extensive cardiac and pulmonary test- ing, but the lack of other cardiac or pulmonary signs or symptoms made these unlikely sources of her chest pain. There was no evidence of pain- ful rib lesions or costochondritis. The distribu- tion of pain, radiating in a band from the back to the anterior chest wall, made thoracic radiculop- athy (neuropathic pain emanating from a thoracic nerve root or intercostal nerve) the leading diag- nostic possibility.
Dr. Mykol Larvie: An MRI scan of the spine per- formed at this hospital (Fig. 1) reveals multiple cystic foci, hypointense on T1-weighted images and hyperintense on T2-weighted images, in the neural foramina of the cervical, thoracic, lum- bar, and sacral spine; they are typical of Tarlov cysts in appearance and measure up to 13 mm in the greatest dimension. The largest lesions mildly expand the neural foramina, which are otherwise intact and well demarcated. There is no abnormal enhancement associated with the lesions, and there is no invasion of adjacent bone or soft tissues. There are mild degenerative chang- es in the spine without stenosis of the spinal canal or neural foramina. Aside from the Tarlov cysts, there are no radiographic abnormalities that suggest a compressive spinal-cord or radic- ular lesion.
Differential diagnosis of thoracic radiculopathy
Dr. Oaklander: Postherpetic neuralgia from a pre- vious episode of shingles must be considered, because herpes zoster is very common and the

The new england journal of medicine

854

n engl j med 368;9 nejm.org february 28, 2013
The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on March 1, 2013. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.

case records of the massachusetts general hospital

ABC D E

Figure 1. T2-Weighted MRI Images and CT Myelogram of the Thoracic and Lumbar Spine. A sagittal MRI image of the thoracic spine (Panel A) shows exaggerated thoracic kyphosis, as well as a Tarlov cyst in the left T7–T8 neural foramen along the left T7 nerve roots (arrow). An axial view shows the same Tarlov cyst (Panel B, arrow) and a smaller cyst along the right T7 nerve root (arrowhead). A subtle hypointense focus in this right Tarlov ICM AUTHOR oaklander RETAKE 1st cyst most likely represents the actual nerve root; the left T7 nerve root is not cle2nadrly visualized in this image. In com- REG F FIGURE 1 a-e parison to the MRI images, a CT myelogram of the thoracic spine in the sagittal (Panel D) and axial (Panel E) views CASE TITLE 3rd Revised shows contrast enhancementEoMfatilhe Tarlov cyst along the left T7 nerve roots (Panels D and E, arrows), indicating Line 4-C open communication betweenEntohne cyst and the subdural compartment.STIhZeE smaller Tarlov cyst along the right T7 ARTIST: mleahy H/T H/T nerve root (Panel E, arrowheaFdIL)Ldoes not enhance andCoims nbot clearly in o3p3epn9 communication with the subdural com- partment. A sagittal MRI image of the lumbar spine (Panel C) reveals multiple Tarlov cysts involving lower thoracic AUTHOR, PLEASE NOTE: and upper sacral nerve roots (arrFoigwusr)e. has been redrawn and type has been reset. Please check carefully.

JOB: 36809
thorax is the most common location. This pa- tient had no history of shingles, but postherpetic neuralgia occasionally develops after an episode of shingles with few or no cutaneous lesions (zoster sine herpete). Less common medical causes of thoracic radicular pain, including diabetes, sar- coidosis, tuberculosis, and syphilis, had been ruled out. This patient also reported intermittent bilateral leg and foot pain and numbness and urinary and fecal incontinence, suggesting the possibility of myelopathy in addition to radicu- lopathy; therefore, a dural arteriovenous fistula or malformation was considered. Most spinal du- ral fistulas are in the low thoracic segments, and

ISSUE: 2-28-13
some are located in root sleeves.1 However, there was no gait disorder, and there were none of the characteristic imaging features of spinal dural vascular lesions (e.g., spinal cord edema, enhance- ment, and perimedullary dilated vessels).
The most common radiologically visible cause of radiculopathy is degenerative spondylosis. Nerve roots can be compressed or damaged by nearby elements, including osteophytes, herni- ated intervertebral disks, or hypertrophy of the uncovertebral or zygapophyseal joints. Degener- ation is less common at the thoracic level than at the cervical or lumbar level, and CT and MRI of the spine in this patient had revealed neither

n engl j med 368;9 nejm.org february 28, 2013 855
The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on March 1, 2013. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.

degenerative changes nor meningiomas, which are less common compressive lesions. The original MRI (8 years earlier) that first revealed abnor- malities had been interpreted as normal. The differential diagnosis of the lesion in the left T7–T8 foramen included perineurial cysts and nerve-sheath tumors, specifically schwannomas and neurofibromas. MRI performed with the administration of contrast material revealed no enhancement and revealed multiple lesions in- volving the thoracic and lower cervical nerve roots that were interpreted as perineurial cysts.
Perineurial cysts (Tarlov cysts)
The lesions seen on imaging of this patient’s spine are characteristic of perineurial cysts. Such cysts were first characterized in 1938 by the neurosur- geon Isadore M. Tarlov during an autopsy of the lower spine.2 He identified them as fluid collec- tions in the nerve-root membranes that have re- stricted communication with the cerebrospinal fluid in the thecal sac. He recognized that the cysts form only on dorsal roots and that they can invaginate into the dorsal-root ganglia or nerve- root fascicles to damage neuronal cell bodies or axons. Later he reported that these cysts could cause sciatica, backache, or the cauda equina syn- drome, similar to intervertebral disk herniations, and he published numerous papers and a mono- graph describing effective surgical treatments for these cysts that came to bear his name.3
Tarlov cysts form only on the dorsal nerve roots because only the cell bodies of sensory neurons migrate out from the spinal cord during embryonic development, leaving behind a sleeve of dura and subarachnoid space. Tarlov cysts are most common at sacral levels, presumably reflect- ing the greater hydrostatic pressure there. High intraluminal pressure can gradually expand them, and very large cysts can compress and damage motor (anterior) roots, the spinal cord, or sur- rounding bone or can cause leakage of cerebro- spinal fluid.
Symptoms Associated with Tarlov Cysts
Could this patient’s thoracic pain be due to the visualized Tarlov cysts? In large studies, the prev- alence of these cysts on lumbosacral MRI images ranges from 1.5 to 2.1%.4-7 Radicular neuropath- ic pain caused by damage to the nerve root that bears the cyst is the most common symptom,8,9

and the most common location of the pain is in the lumbosacral region.3,5,8,10 In a study involv- ing 122 patients with symptomatic Tarlov cysts (84% women; mean age, 54 years), in whom other potential causes had been ruled out by neurologic examination and diagnostic testing, all patients had local or radiating pain (sciatica) or both, and 10% had bladder or bowel incontinence.11 These symptoms, as well as the marked female predom- inance, are consistent with this patient’s presen- tation.5,11 This patient’s pain was primarily tho- racic, at the location of the cysts seen on early imaging studies; however, she also reported more typical symptoms of sacral pain, leg and foot pain and numbness, and incontinence.
Prevalence of Symptoms Due to Tarlov Cysts
How can we be sure the cysts are the cause of this patient’s symptoms? Relationships between poorly localized symptoms (e.g., back pain) and common radiologic abnormalities (e.g., Tarlov cysts) can be difficult to establish. In addition, more than one degenerative change in a patient is often identified on spine imaging. In several studies, Tarlov cysts were thought to be the cause of symptoms in 20 to 30% of patients, because of the localization of the symptoms and the ab- sence of other abnormalities on imaging.3,5,6 An electrophysiological study involving a small num- ber of patients with sacral Tarlov cysts showed abnormal sural-nerve action potentials in 45% of the patients, always on the same side as the cysts, suggesting that electrophysiological testing might help identify cysts that are causing nerve dam- age.12 Imaging studies in this patient revealed no explanation for her radicular symptoms other than the cysts, so it is reasonable to conclude that they are the cause.
I have found that most physicians are either unaware of the existence of Tarlov cysts or be- lieve that they do not cause symptoms. Radiolo- gists do not always report visualized Tarlov cysts, or they may report an imaging study as normal despite the presence of Tarlov cysts, as in the case of this patient’s initial MRI scan.
DR. ANNE LOUISE OAKLANDER’S DIAGNOSIS
Thoracic radicular pain caused by Tarlov (peri- neurial) cysts.

The new england journal of medicine

856

n engl j med 368;9 nejm.org february 28, 2013
The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on March 1, 2013. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.

case records of the massachusetts general hospital

DISCUSSION OF MANAGEMENT
Neurosurgical Management of Tarlov Cysts
Dr. Donlin M. Long: This patient requested consul- tation with me because she had a decades-long history of slowly progressive midthoracic pain. When I examined her, the complete neurologic examination was normal and there were no mus- culoskeletal findings. MRI revealed two 8-mm-by- 10-mm perineurial cysts at the T7–T8 1evel that were associated with enlarged bony foramina and had filled with contrast material on a previous CT myelography. The cysts were the most likely cause of her pain because they were located at the same thoracic level as the pain and because there were no other potential causes.
Symptomatic perineurial cysts such as those described by Tarlov2 have been reported for more than 70 years. Symptomatic dural ectasia com- plicating connective-tissue disorders of many kinds has also been described.9,13 However, most physicians are unaware of these reports and be- lieve that Tarlov cysts do not cause symptoms or require treatment.
Treatment options for symptomatic Tarlov cysts
Published treatment options for apparently symp- tomatic Tarlov cysts include perineural injection of a glucocorticoid, cyst aspiration, aspiration and injection of the cyst with blood or fibrin sealant, and surgical obliteration.6,14-18 There is no evi- dence that aspiration or perineural injection of a glucocorticoid or local anesthetic agents pro- duces anything but temporary relief, but such procedures can be used to confirm that the cysts are the cause of symptoms.19 In one study, radio- logically guided percutaneous aspiration of a cyst, followed by an injection of fibrin to seal the cyst, resulted in satisfactory pain relief in 65% of patients, including 19% who had marked or complete resolution; there were no serious complications.11 Among the patients who had pain relief, 23% had symptoms that recurred, and they were treated with repeat cyst sealing or surgery.
Neurosurgical treatment usually involves a small sacral laminectomy, unless the cyst has al- ready eroded through the sacrum. Cyst walls are opened wide, avoiding neural tissues, and dis- sected from the sacrum. Connections with the

thecal sac are occluded if possible, and the cyst cavity is packed with autologous fat and sutured shut. The largest case series show complete or substantial improvement in more than 80% of patients who underwent surgery, with a 7% rate of mostly temporary complications.15,17,18,20,21
Treatment of this patient
Because it seemed highly likely that the Tarlov cysts in the thoracic region were causing this pa- tient’s radiculopathy, I performed transcutaneous aspiration of the cysts at T7 and T8. A ball-valve type of distortion of cerebrospinal fluid dynam- ics has been proposed to explain some of the pathophysiological features of the pain associat- ed with Tarlov cysts.22 Fluid accumulates under pressure, so more fluid can enter the cyst than can escape. Pain may be generated when disten- tion of the cyst activates neural, dural, or bone nociceptors. On CT myelography in this patient, contrast material had filled the cyst promptly, showing free entry of fluid; there was no reflux of contrast material into the thecal sac, which substantiated the presence of a ball-valve effect. Both of this patient’s cysts were then partially filled with commercially available fibrin sealant.
The patient had immediate but partial relief of pain. Subsequently, a fluoroscopically guided transforaminal epidural blockade of the bilateral T6 and T9 roots was administered to determine whether adjacent roots were contributing to the pain, but no additional pain relief was achieved.
Additional evaluation and possible Genetic Connective-Tissue Disorders
Dr. Oaklander: When I saw this patient 4 months after she had been treated by Dr. Long with an injection of fibrin, the causes of her sacral and leg pain and numbness and the urinary and fecal incontinence had not been investigated. I sus- pected that they might also be related to Tarlov cysts; therefore, I requested MRI of the lumbosa- cral spine. This revealed sacral cysts, which I pre- sumed were the cause of the sacral radicular symptoms. Similarly, the patient’s reports of in- termittent hand paresthesias prompted imaging of the cervical spine, which revealed additional cysts. Spine imaging in this patient revealed 21 Tarlov cysts, suggesting a unifying cause of her symptoms.
The presence of so many Tarlov cysts strongly

n engl j med 368;9 nejm.org february 28, 2013 857
The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on March 1, 2013. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.

suggested an underlying connective-tissue disor- der. The Ehlers–Danlos syndrome, Marfan’s syn- drome (Online Mendelian Inheritance in Man [OMIM] number, 154700), and the Loeys–Dietz syndrome have all been associated with Tarlov cysts.9,13 This patient had no family history of such disorders, but possible evidence included the patient’s height and arachnodactyly, her father’s two aortic aneurysms, her son’s preterm birth, and her daughter’s iliac-artery dissection and nonrheumatic, myxomatous, prolapsing mitral valve that had required repair.11
Sacral abnormalities are associated with Mar- fan’s syndrome and related disorders.6 The clas- sic finding is dural ectasia (i.e., widening of the sacral canal due to pressure from a dilated or cystic lower thecal sac), often accompanied by neu- ral foramina that have been dilated by cysts, as in this patient.23 Genetic testing of the patient’s daughter for Marfan’s syndrome revealed no pathogenic mutation in the gene encoding fibril- lin-1 (FBN1), leaving the patient’s presumed connective-tissue disorder unidentified.
Medical treatment and follow-up
I followed the patient for just over a year after her first consultation with me. For medical manage- ment of the residual symptoms, I suggested that she lessen the risk of falls by tapering the dose of clonazepam and that she preclude potential neuro- toxic effects of vitamin B6 by tapering megadoses of toxic vitamins.24 I recommended the adminis- tration of low-dose methadone for neuropathic pain and acetazolamide to lower intrathecal pressure. I advised her that definitive manage- ment of the chest pain would require procedural intervention, and she began to consider neuro- surgical evaluation for laminectomy.
When I telephoned her home a month after our last visit (15 months after her initial evalua- tion at this hospital) to check on her progress, her family informed me that she had fallen down her basement stairs and had been admitted to a local hospital because of severe bruising. There, pneu- monia developed and hematologic abnormalities were noted, and she had recently died.
The patient had consented to autopsy at this hospital during an earlier office visit. Therefore, the body was transferred to this hospital for an investigation of the cause of death and the cause of her longtime chronic pain.

PATHOLOGICAL DISCUSSION
Dr. Christian J. Davidson: To determine the exact na- ture of this patient’s spinal cysts, we removed the spinal column, from the midcervical level to the sacrum and pelvis (Fig. 2A), with help from Dr. C. Chambliss Harrod, a resident in orthopedic surgery. The spine was frozen at −80°C and cut axially into slices 2 mm to 4 mm thick. At many levels, we observed small, extradural, thin-walled cysts where the spinal nerve exited the spinal ca- nal (Fig. 2B through 2E). Sections of representa- tive cysts were stained with hematoxylin and eo- sin, which revealed that the cysts had thin walls (Fig. 2F and 2G) and were sitting within the nerve roots just proximal to the dorsal-root ganglia, with spinal-nerve fibers passing through them. Nerve twigs, confirmed by immunohistochemical anal- ysis for neurofilament protein, were embedded in the walls of the cysts (Fig. 3A and 3C). The cells lining the cysts were small and had wispy cyto- plasm, suggestive of perineurial cells, and were positive for GLUT1 (Fig. 3B) and claudin-1 (not shown), two well-characterized markers of peri- neurial cells.
The cyst at the T7–T8 intervertebral space, which had been drained and injected with fibrin glue, had a more opaque appearance on gross examination than did the other cysts (Fig. 2E). Microscopical examination revealed increased septations but no inflammation, fibrosis, or foreign-body–type giant-cell reaction (Fig. 2H).
The location, gross appearance, and histo- logic characteristics of these cysts are the same as originally described by Tarlov in 1938 and in multiple subsequent reports.2,23,25,26 They are lo- cated just outside the spinal canal, where the spinal nerve penetrates the dura, they are intra- neural, and they abut the dorsal-root ganglia. In the absence of other anatomical or microscopical abnormalities, the multiple cysts in the sacral spine probably are responsible for the patient’s incontinence.
A chronic myeloproliferative neoplasm, which had been diagnosed shortly before the patient’s death, involved multiple organs, including the bone marrow, liver, spleen, cardiac interatrial sep- tum, and lungs, and was the most likely cause of death. Other findings included hypertensive car- diac disease, nephrosclerosis, and multiple ade- nomas of the renal cortex.

The new england journal of medicine

858

n engl j med 368;9 nejm.org february 28, 2013
The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on March 1, 2013. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.

case records of the massachusetts general hospital

ABC DE FGH

Figure 2. Characteristics of Tarlov Cysts. A gross photograph (Panel A) shows the extent of the spine (approximately 60 cm) that was removed at autopsy. Representative gross photographs of axial sections of the spine are shown at the T11–T12 intervertebral space (Panels B and D, arrows) and at S1–S2 (Panel C, arrows). Both show thin-walled cysts containing clear fluid. A gross photograph of the T7–T8 intervertebral space shows that the lining of the cyst that had been drained and injected with fibrin glue (Panel E, arrow) has a more opaque appearance than do the cysts at T10–T11. Photomicrographs at T10–T11 show thin-walled cysts (Panel F, low magnification; hematoxylin and eosin) sitting within the nerve roots (Panel G, high magnification; hematoxylin and eosin) just proximal to the dorsal-root ganglia, with spinal nerve fibers passing through them. A photomicrograph of the cyst at T7–T8 that had been injected with fibrin glue (Panel H, hematoxylin and eosin) shows increased septations in the cyst as compared with the other cysts but no evidence of inflammation.

Dr. Nancy Lee Harris (Pathology): Dr. Edward Tarlov, the son of Dr. Isadore Tarlov, is with us today. Dr. Tarlov, would you like to comment?
Dr. Edward C. Tarlov (Neurosurgery, Lahey Clinic): The incidence of perineurial Tarlov cysts has been estimated conservatively at 1 to 2% among pa- tients referred for lumbosacral MRI for low back pain or other concerns. In my experience, virtu- ally all of these are asymptomatic. Back pain is ubiquitous in the general population. When ex-

amined, most adults have degenerative changes in the spinal disks and facets. To attribute symp- toms to Tarlov cysts is, I believe, usually falla- cious, and surgical treatments, including aspira- tion, are likely to produce no lasting benefit and may even cause harm.
Dr. Long: I agree with Dr. Tarlov that the ma- jority of people with Tarlov cysts are not symp- tomatic. However, when symptoms develop and no other causes are found, especially in patients

n engl j med 368;9 nejm.org february 28, 2013 859
The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on March 1, 2013. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.

The new england journal of medicine

AB C

Figure 3. Microscopical Characteristics of Cyst Walls. A high-power photomicrograph of the cyst wall at T10 (Panel A) shows a thin layer of lining cells with the appearance of perineurial cells (arrows). Arrowheads show nerve twigs in the cyst wall. The cells that line the cyst stain for GLUT1 (Panel B), a marker of perineurial cells, confirming that the cells of the cyst lining are perineurial cells. There is a perispinal soft-tissue infiltrate of leukemic cells that are also GLUT1-positive (Panel B, top). Neurofilament immu- nostaining (Panel C) highlights the nerve twigs in the cyst wall.

with progressive neurologic deficits, these cysts can be treated surgically with good results.
I would also like to thank the patient and her family for permitting the autopsy, a very thought- ful and understanding act on her part, which gave us the opportunity to learn more about a poorly understood condition.
Dr. Oaklander: It is noteworthy that this first published pathological study of a fibrin-injected cyst showed no evidence of severe or chronic in- flammation, adhesions, or scarring, the presence of which might affect future surgical management.

ANATOMICAL DIAGNOSIS
Multiple Tarlov (perineurial) cysts.
This case was presented at the Neurology/Neuroscience Grand Rounds.
Dr. Long reports receiving consulting fees from Aptiv Solu- tions and Seikagaku. No other potential conflict of interest rele- vant to this article was reported.
Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.
We thank Dr. E. Tessa Hedley-Whyte for supervising the autopsy and for providing advice on the neuropathological dis- cussion.

References
1. Krings T, Geibprasert S. Spinal dural arteriovenous fistulas. AJNR Am J Neuro- radiol 2009;30:639-48.
2. Tarlov IM. Perineurial cysts of the spi- nal nerve roots. Arch Neurol Psychiatry 1938;40:1067-74.
3. Idem. Sacral nerve-root cysts: another cause of the sciatic or cauda equina syn- drome. Springfield, IL: Charles C. Thom- as, 1953.
4. Joo J, Kim J, Lee J. The prevalence of anatomical variations that can cause inad- vertent dural puncture when performing caudal block in Koreans: a study using magnetic resonance imaging. Anaesthesia 2010;65:23-6.

5. Langdown AJ, Grundy JR, Birch NC. The clinical relevance of Tarlov cysts. J Spinal Disord Tech 2005;18:29-33.
6. Paulsen RD, Call GA, Murtagh FR. Prevalence and percutaneous drainage of cysts of the sacral nerve root sheath (Tarlov cysts). AJNR Am J Neuroradiol 1994;15: 293-9.
7. Park HJ, Jeon YH, Rho MH, et al. Inci- dental findings of the lumbar spine at MRI during herniated intervertebral disk disease evaluation. AJR Am J Roentgenol 2011;196:1151-5.
8. Hiers RH, Long D, North RB, Oakland- er AL. Hiding in plain sight: a case of Tar- lov perineural cysts. J Pain 2010;11:833-7.

9. Isono M, Hori S, Konishi Y, et al. Ehlers-Danlos syndrome associated with multiple spinal meningeal cysts — case report. Neurol Med Chir (Tokyo) 1999;39: 380-3.
10. Van de Kelft E, Van Vyve M. Sacral meningeal cysts and perineal pain. Lancet 1993;341:500-1.
11. Murphy KJ, Nussbaum DA, Schnupp S, Long D. Tarlov cysts: an overlooked clinical problem. Semin Musculoskelet Radiol 2011;15:163-7.
12. Cattaneo L, Pavesi G, Mancia D. Sural nerve abnormalities in sacral perineural (Tarlov) cysts. J Neurol 2001;248:623-4. 13. Pyeritz RE, Fishman EK, Bernhardt

860

n engl j med 368;9 nejm.org february 28, 2013
The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on March 1, 2013. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.

case records of the massachusetts general hospital

BA, Siegelman SS. Dural ectasia is a com- mon feature of the Marfan syndrome. Am J Hum Genet 1988;43:726-32.
14. Acosta FL Jr, Quinones-Hinojosa A, Schmidt MH, Weinstein PR. Diagnosis and management of sacral Tarlov cysts: case report and review of the literature. Neurosurg Focus 2003;15:E15.
15. Mummaneni PV, Pitts LH, McCor- mack BM, Corroo JM, Weinstein PR. Mi- crosurgical treatment of symptomatic sacral Tarlov cysts. Neurosurgery 2000; 47:74-9.
16. Murphy K, Wyse G, Schnupp S, et al. Two-needle technique for the treatment of symptomatic Tarlov cysts. J Vasc Interv Radiol 2008;19:771-3. [Erratum, J Vasc In- terv Radiol 2008;19:1525.]
17. Caspar W, Papavero L, Nabhan A, Loew C, Ahlhelm F. Microsurgical exci- sion of symptomatic sacral perineurial

cysts: a study of 15 cases. Surg Neurol 2003;59:101-6.
18. GuoD,ShuK,ChenR,KeC,ZhuY, Lei T. Microsurgical treatment of symp- tomatic sacral perineurial cysts. Neuro- surgery 2007;60:1059-66.
19. Lee JY, Impekoven P, Stenzel W, Löhr M, Ernestus RI, Klug N. CT-guided percu- taneous aspiration of Tarlov cyst as a use- ful diagnostic procedure prior to opera- tive intervention. Acta Neurochir (Wien) 2004;146:667-70.
20. Tanaka M, Nakahara S, Ito Y, et al. Surgical results of sacral perineural (Tarlov) cysts. Acta Med Okayama 2006; 60:65-70.
21. Voyadzis JM, Bhargava P, Henderson FC. Tarlov cysts: a study of 10 cases with review of the literature. J Neurosurg 2001; 95:Suppl:25-32.
22. Neulen A, Kantelhardt SR, Pilgram-

Pastor SM, Metz I, Rohde V, Giese A. Mi- crosurgical fenestration of perineural cysts to the thecal sac at the level of the distal dural sleeve. Acta Neurochir (Wien) 2011; 153:1427-34.
23. Tarlov IM. Cysts, perineurial, of the sacral roots: another cause, removable, of sciatic pain. J Am Med Assoc 1948;138: 740-4.
24. Parry GJ, Bredesen DE. Sensory neu- ropathy with low-dose pyridoxine. Neu- rology 1985;35:1466-8.
25. Tarlov IM. Spinal perineurial and meningeal cysts. J Neurol Neurosurg Psy- chiatry 1970;33:833-43.
26. Rexed B. Arachnoidal proliferations with cyst formation in human spinal nerve roots at their entry into the intervertebral foramina: preliminary report. J Neuro- surg 1947;4:414-21.
Copyright © 2013 Massachusetts Medical Society.

Lantern Slides Updated: Complete PowerPoint Slide Sets from the Clinicopathological Conferences
Any reader of the Journal who uses the Case Records of the Massachusetts General Hospital as a teaching exercise or reference material is now eligible to receive a complete set of PowerPoint slides, including digital images, with identifying legends, shown at the live Clinicopathological Conference (CPC) that is the basis of the Case Record. This slide set contains all of the images from the CPC, not only those published in the Journal. Radiographic, neurologic, and cardiac studies, gross specimens, and photomicrographs, as well as unpublished text slides, tables, and diagrams, are included. Every year 40 sets are produced, averaging 50-60 slides per set. Each set is supplied on a compact disc and is mailed to coincide with the publication of the Case Record.
The cost of an annual subscription is $600, or individual sets may be purchased for $50 each. Application forms for the current subscription year, which began in January, may be obtained from the Lantern Slides Service, Department of Pathology, Massachusetts General Hospital, Boston, MA 02114 (telephone 617-726-2974) or e-mail Pathphotoslides@partners.org.

n engl j med 368;9 nejm.org february 28, 2013 861
The New England Journal of Medicine
Downloaded from nejm.org at UNIVERSITY OF WASHINGTON on March 1, 2013. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.