Neurologic manifestations of sarcoidosis.

Handbook of Clinical Neurology, Vol. 119 (3rd series) Neurologic Aspects of Systemic Disease Part I Jose Biller and Jose M. Ferro, Editors © 2014 Elsevier B.V. All rights reserved

Chapter 21

Neurologic manifestations of sarcoidosis ALLAN KRUMHOLZ* AND BARNEY J. STERN Department of Neurology, University of Maryland School of Medicine, Baltimore, MD, USA

HISTORY

CLINICAL FINDINGS

Sarcoidosis was first described in 1877 by Sir Jonathan Hutchinson (James and Williams, 1985; Lieberman et al., 1985). At the turn of the century, Caesar Boeck termed the disease “multiple benign sarkoid,” because of its histologic similarity to sarcoma, and it is from this that the term sarcoidosis is derived (Lieberman et al., 1985). Boeck also demonstrated that the unifying feature of sarcoidosis was epithelioid cell granulomas which could involve different organs, a view that serves as the basis of current thinking about sarcoidosis. In 1936, Jorgen Schaumann advanced the concept that rather than affecting only solitary organ systems, such as the skin or lymph nodes, sarcoidosis was instead a systemic disease characterized by the presence of granulomas in multiple organs. Today, it is well recognized and established that not only can sarcoidosis involve many different organ systems, but is capable of attacking any body tissue (Lieberman et al., 1985). Here we focus on neurologic manifestations of sarcoidosis. The first report of neurologic involvement by sarcoidosis was in 1917 by Heerfordt in his original account of what he described as “uveo-parotid fever.” In that report, he expressly associated cranial nerve palsies with sarcoidosis (Heerfordt, 1909). Numerous individual case reports and small series subsequently enlarged the known clinical spectrum of neurosarcoidosis. By 1948, Colover was able to collect information on 118 published cases of neurosarcoidosis (Colover, 1948). Since then, numerous additional case reports and series attest to the wide range of the neurologic manifestations of sarcoidosis. Sarcoidosis may affect any part of the nervous system but typically presents in distinctive clinical patterns.

Sarcoidosis Because diseases for which a cause is not known, such as sarcoidosis, are difficult to define with assurance, their diagnosis is often based on establishing characteristic descriptive profiles. For sarcoidosis even this approach has been difficult. Its varied manifestations and similarity to other granulomatous disorders of both an infectious and a noninfectious nature, combined with the lack of a specific confirmatory diagnostic test, have made reaching a consensus on a standard definition for sarcoidosis especially difficult. The first internationally accepted definition of sarcoidosis was proposed in 1975. It states that: “Sarcoidosis is a multisystem granulomatous disorder of unknown etiology, most commonly affecting young adults and presenting most frequently with bilateral hilar adenopathy, pulmonary infiltration, skin or eye lesions. The diagnosis is established most securely when clinical and radiographic findings are supported by histologic evidence of widespread noncaseating epithelioid-cell granulomas in more than one organ” (James et al., 1976). Although not all aspects of this proposed full definition are still widely accepted, the view that sarcoidosis is a systemic granulomatous disorder of unknown etiology diagnosed on the basis of clinical and histologic evidence is correct. The original definition, which is actually considerably longer than the excerpted quote above, has been rightly criticized for proposing that sarcoidosis is a consequence of impaired immunologic function. Instead, it is now accepted that the cause of sarcoidosis is not impaired immunity, but rather heightened immunologic activity. Still, apart from that criticism, this early definition is valuable because it offers a simple, operational definition of a very complex disease.

*Correspondence to: Allan Krumholz, M.D., S12c09, Department of Neurology, University of Maryland Medical Center, 22 South Greene Street, Baltimore, Maryland 21201, USA. E-mail: [email protected]

306

A. KRUMHOLZ AND B.J. STERN

Table 21.1 Percentage frequency of organ involvement in sarcoidosis Manifestation

Frequency

Intrathoracic Hilar node Lung parenchyma Upper respiratory tract Dermatologic Skin Erythema nodosum Ocular Lacrimal Parotid Splenomegaly Pheripheral lymphandenopathy Bone Cardiac Hepatomegaly Hypercalcemia Neurologic Hematologic, endocrinologic, gastrointestinal and genitourinary

87 72 46 6 18 15 15 3 6 10 28 3 3 10 13 5 Rare

Sarcoidosis usually presents between the ages of 20 and 40 years. However, it also occurs in children (Weinberg et al., 1983) and older populations. It appears to have similar clinical manifestations in all age groups. Intrathoracic structures are most commonly affected (87% of patients) (Table 21.1). Although it is generally understood that a diagnosis of sarcoidosis is most secure when it is based on histologic confirmation, on average 30% of the patients described in the literature lack histologic confirmation, and the diagnosis is often based solely on clinical and radiologic findings (Teirstein and Lesser, 1983). Sarcoidosis may be asymptomatic, or it can present with constitutional symptoms and pulmonary or extrapulmonary manifestations. Anatomic presence of the disease in an organ often occurs without overt clinical evidence of dysfunction. It is estimated that 20–40% of patients are asymptomatic at presentation, their disease being discovered by routine chest radiography (Teirstein and Lesser, 1983; Iannuzzi et al., 2007). When present, specific symptoms are often related to the extent of organ involvement. In the case of pulmonary disease, this may relate to the mechanical interference of sarcoidosis with normal lung function. In particular, respiratory symptoms such as cough, dyspnea, chest pain, and nasal complaints are noted as the presenting symptom of sarcoidosis in 30–50% of individuals. Constitutional symptoms are described in 20–30% of patients and include fever, fatigue, anorexia

or weight loss. Systemic symptoms are said to be more common among black populations (Katz, 1983; Israel and Kataria, 1985). In less than 10% of sarcoidosis patients the onset of symptoms is of neither a systemic nor a pulmonary nature (Israel and Kataria, 1985). Neurologic manifestations of sarcoidosis are in this category. Other forms of extrapulmonary sarcoidosis include skin lesions, lymphadenopathy, parotid gland masses, liver or spleen enlargement, and ocular and cardiac involvement. Erythema nodosum is the most common dermatologic form of sarcoidosis and lupus pernio is the most characteristic skin lesion (Katz, 1983; Israel and Kataria, 1985). Lupus pernio presents as a persistent, bluish indurated lesion with a predilection for the nose, cheeks, ears, and lips (Sharma, 1984). Most patients with systemic sarcoidosis have a good prognosis. For approximately two-thirds, the disease resolves spontaneously without major difficulties. This benign course is most common in asymptomatic patients with only hilar adenopathy on chest X-ray; they have a 70– 80% likelihood of spontaneous remission (Fanburg, 1983). However, for a third of patients, symptoms persist or the disease progressively worsens. Pulmonary dysfunction is the major problem for most patients with a persistent or progressive clinical course, and 15–20% of such sarcoidosis patients have some degree of permanent loss of lung function, as manifested clinically by fibrotic changes, such as honeycombing and retraction, on chest X-ray (Fanburg, 1983; Deremee, 1985a; Lieberman, 1985; Iannuzzi et al., 2007).

Neurosarcoidosis Neurologic symptoms are the presenting feature of sarcoidosis in one-half of individuals with neurosarcoidosis (Stern et al., 1985). Some three-quarters of patients destined to develop neurologic disease do so within 2 years of becoming afflicted with sarcoidosis. The approximate frequency of the various neurologic complications is presented in Table 21.2. Most patients with neurosarcoidosis show evidence of disease in other organ systems such as the lung (Wiederholt and Siekert, 1965; Delaney, 1977; Pentland et al., 1985; Stern et al., 1985; Oksanen, 1986; Luke et al., 1987). However, systemic disease is not always evident, particularly early in a patient’s clinical course, and in some instances it can be difficult to find or may not exist. There are numerous reports of biopsy-proven neurosarcoidosis limited to the central nervous system (CNS), and these may be considered isolated central nervous system sarcoidosis. Several large series indicate that perhaps 10% of patients with neurosarcoidosis have isolated central nervous system sarcoidosis, at least at

NEUROLOGIC MANIFESTATIONS OF SARCOIDOSIS Table 21.2 Neurosarcoidosis Clinical manifestation

Approximate frequency (%)

Cranial neuropathy Facial palsy Aseptic meningitis Hydrocephalus Parenchymal disease Endocrinopathy Mass lesion(s) Encephalopathy/ vasculopathy Seizures Neuropathy Myopathy

50–75 25–50 10–20 10 10–15 5–10 5–10 5–10 5–10 10

initial presentation over the course of several years follow-up, without other evidence of system sarcoidosis (Oksanen, 1986; Zajicek et al., 1999; Ferriby et al., 2001; Scott et al., 2007). In some instances, systemic sarcoidosis manifests or is discovered at a later time, but in others the sarcoidosis remains confined to the nervous system. Long-term outcomes for individuals with such isolated central nervous system sarcoidosis vary, and seem to relate to the specific extent and nature of the neurologic involvement or dysfunction (Oksanen, 1986; Zajicek et al., 1999; Ferriby et al., 2001; Scott et al., 2007). Although the range of clinical manifestations of sarcoidosis is exceptionally varied, most patients present in characteristic ways that can be systematically organized (Table 21.1). Similarly, the neurologic manifestations of sarcoidosis can be classified, as presented in Table 21.2. In dealing with neurosarcoidosis patients, it is useful to approach patients using such a classification. However, it should be remembered that one-third to one-half of neurosarcoidosis patients develop more than one neurologic manifestation of their disease.

CRANIAL NEUROPATHY The most frequent neurologic complication of sarcoidosis is cranial neuropathy (Stern et al., 1985). This occurs in approximately three-quarters of patients with neurosarcoidosis. Any cranial nerve can be affected, and over one-half of patients have multiple cranial nerve lesions (Stern et al., 1985). By far the most commonly affected cranial nerve is the facial nerve. Cranial nerve I. Olfactory nerve dysfunction can occur secondary to meningeal sarcoidosis involving the subfrontal region. However, anosmia or hyposmia

307

may also result from local nasal granulomatous invasion by sarcoidosis, and, therefore, in a patient with olfactory complaints, an otolaryngological evaluation is necessary before attributing impaired olfaction to CNS disease. Cranial nerve II. Optic nerve involvement in sarcoidosis is much less frequent than other ocular manifestations of sarcoidosis such as uveitis. Optic nerve involvement can present with an acute or subacute, painful, or chronic, painless visual loss (Graham et al., 1986; Stern and Corbett, 2007; Pawate et al., 2009). This visual loss may be due to bulbar or retrobulbar invasion of the optic nerve by granulomas, compression of the optic nerve by a granulomatous mass, or optic atrophy. Optic disc edema may be secondary to papilledema due to sarcoidosis-induced increased intracranial pressure or as a direct local invasive effect of sarcoidosis. A chiasmal syndrome has also been reported (Gelwan et al., 1988). Cranial nerves III, IV, and VI. Disorders of ocular motility may follow involvement of the III, IV, or VI cranial nerves in the granulomatous process (Stern et al., 1985). Typically, these nerves are damaged in their extra-axial course in the subarachnoid space as they traverse the meninges. However, they may also be involved via local orbital disease, and rarely, brainstem intra-axial central nervous system pathways can be affected by sarcoidosis (Kirkham and Kline, 1976). Uncommonly, disordered ocular motility may be due to sarcoidosis involving the extraocular muscles themselves (Obenauf et al., 1978). Occasionally, pupillary dysfunction is caused by neurosarcoidosis (Henkind and Gottlieb, 1973; Kirkham and Kline, 1976; Cohen and Reinhardt, 1982; Poole, 1984). For instance, Horner’s syndrome due to disruption of the cervical sympathetic nerves has been reported (Cohen and Reinhardt, 1982). Cranial nerve V. Trigeminal nerve disease may present as facial numbness or, rarely, trigeminal neuralgia (Stern et al., 1985). Headache may also represent trigeminal nerve dysfunction intracranially. Involvement of the muscles of mastication is unusual. Cranial nerve VII. Of all the cranial nerve syndromes, peripheral facial cranial nerve (VII) palsy is the most common, and it is also the single most frequent neurologic manifestation of sarcoidosis. It develops in 25– 50% of all patients with neurosarcoidosis. Although usually unilateral, bilateral facial palsy also can occur presenting with either simultaneous or sequential paralysis. Over half of all patients with facial palsy also have other forms of nervous system involvement. In patients with a solitary facial palsy, spinal fluid examination typically has been reported as normal, but in individuals in whom there are other associated forms of neurosarcoidosis, the spinal fluid examination is reported to be

308


abnormal in 80% of patients. The specific reason for a facial nerve palsy in sarcoidosis may vary. Rarely is the facial palsy caused by parotid inflammation. More likely, the nerve is compromised as it traverses the meninges and subarachnoid space, and as suggested by Oksanen (Oksanen, 1986; Oksanen and Salmi, 1986), facial paresis probably is also due to intra-axial sarcoidosis-induced inflammation involving the facial nerve. In general the prognosis for the facial palsy is good with over 80% of patients having a good outcome in terms of recovery of facial function (Stern et al., 1985; Luke et al., 1987). Cranial nerve VIII. Eighth cranial nerve involvement is the second most common form of cranial nerve dysfunction in sarcoidosis. Sarcoidosis may involve the auditory or vestibular portions of the nerve. When either loss of hearing or vestibular dysfunction occur, they may be sudden or insidious and often fluctuate over time (Hybels and Rice, 1976; Jahrsdoerfer et al., 1981; Colvin, 2006). If hearing loss occurs, it is typically a sensorineural loss. As with facial nerve palsy, bilateral VIII nerve disease may occur. In fact, either bilateral VII or VIII nerve dysfunction is suggestive of neurosarcoidosis (Wiederholt and Siekert, 1965; Stern et al., 1985; Oksanen and Salmi, 1986). Cranial nerves IX and X. Glossopharyngeal and vagus nerve involvement by sarcoidosis causes dysphagia and dysphonia. Hoarseness is more commonly due to laryngeal nerve dysfunction from intrathoracic disease than CNS inflammation involving the vagus nerve (El-Kassimi et al., 1990). Cranial nerves XI and XII. Eleventh and twelfth cranial nerve disease can occur but seems to be quite rare in sarcoidosis.

Meningeal disease Meningeal disease is seen in approximately 10–20% of patients with neurosarcoidosis and can present as aseptic meningitis or, less commonly, as a meningeal or dural mass lesion. Aseptic meningitis is characterized by headache, meningismus, and a sterile CSF with a predominantly mononuclear pleocytosis and may be a recurrent problem in some patients with neurosarcoidosis (Plotkin and Patel, 1986). Hypoglycorrhachia, or low cerebrospinal fluid glucose concentration, is occasionally found, and there is often an elevation of cerebrospinal spinal fluid protein. When meningeal sarcoidosis mass lesions occur, they can mimic intracranial tumors such as meningiomas (Figs 21.1 and 21.2) (Israel, 1983; Osenbach et al., 1986; Sethi et al., 1986).

Fig. 21.1. Coronal T1-weighted MRI with gadolinium enhancement demonstrating a convexity sarcoidosis mass lesion that was initially mistaken for a meningioma.

Hydrocephalus Hydrocephalus is noted in about 10% of neurosarcoidosis patients, may be associated with sarcoidosis meningitis, and can pose great danger. Hydrocephalus is a potentially lethal complication of sarcoidosis (Scott, 2000). Patients with acute hydrocephalus may die suddenly from increased intracranial pressure, and even patients with chronic hydrocephalus have the potential to acutely decompensate. Patients with acute hydrocephalus characteristically present with headache, altered mentation or consciousness, and impaired gait. On examination papilledema or other signs of raised intracranial pressure can be found. Once clinically suspected, the diagnosis of hydrocephalus is best substantiated with imaging studies such as cranial computed tomography (CT) or magnetic resonance imaging (MRI) (Fig. 21.3). Acute decompensating hydrocephalus is a medical emergency that necessitates prompt diagnosis and treatment, such as ventricular drainage. In an effort to stabilize such individuals prior to surgery or in conjunction with it, acute treatment with high-dose corticosteroids seems appropriate. The hydrocephalus in patients with neurosarcoidosis can be either of the communicating or the

NEUROLOGIC MANIFESTATIONS OF SARCOIDOSIS

309

Fig. 21.3. CT scan in the axial plane in a patient with neurosarcoidosis who presented with symptomatic hydrocephalus requiring placement of a ventricular shunt.

Fig. 21.2. (A) Axial T1-weighted MRI with gadolinium enhancement showing a frontal extracerebral mass. (B) Axial T1-weighted MRI with gadolinium enhancement showing the same patient as in (A) with a substantial reduction in the size of the mass following therapy with high-dose corticosteroids.

noncommunicating type. Chronic basilar meningitis with obliteration of subarachnoid cerebrospinal fluid flow is a major cause for communicating hydrocephalus. In addition, infiltration of the ventricular system by granulomas, granulomatous compression of the aqueduct, or outlet obstruction of the fourth ventricle by granulomas may all cause noncommunicating types of hydrocephalus. Both inflammation and fibrosis due to sarcoidosis can impair CSF flow causing ventricular enlargement (Lukin et al., 1975; Schlitt et al., 1986). Enhancement of the meninges on imaging studies suggests an active

inflammatory process which might be expected to respond to corticosteroid treatment. In contrast, lack of such enhancement may suggest a predominantly fibrotic reaction that would be less likely to respond to anti-inflammatory medications. One large series of neurosarcoidosis reported that three of 48 patients with incomplete improvement following corticosteroid and other immunotherpapy required ventriculoperitoneal shunting for hydrocephalus (Scott et al., 2007).

Central nervous system parenchymal disease Parenchymal brain disease is reported in about 50% of patients with neurosarcoidosis and can present in several forms. Hypothalamic dysfunction is the most common manifestation of CNS parenchymatous disease. When hypothalamic dysfunction occurs it usually involves the neuroendocrinologic system or “vegetative functions” such as temperature regulation, appetite, thirst, sleep, and libido. However, neuroendocrinologic disease in sarcoidosis can also occur secondary to pituitary disease. Any of the neuroendocrinologic systems can be affected by sarcoidosis (Stern et al., 1985).

310


ENDOCRINE DISORDERS Characteristically endocrinologic dysfunction in sarcoidosis is due to either a hypothalamic or pituitary granulomatous mass or a more diffuse “local” encephalopathy (Stuart et al., 1978; Ismail et al., 1985; Capellan et al., 1990; Chapelon et al., 1990; Missler et al., 1990). Probably the most common site for central nervous system parenchymal involvement is the diencephalon (Bihan et al., 2007). This can sometimes be seen with imaging studies. Given the predilection of sarcoidosis for the basal meninges, the relative frequency of such endocrinologic disturbances is not surprising. Potential endocrinologic manifestations of neurosarcoidosis include thyroid disorders, disorders of cortisol metabolism, and sexual dysfunction. An elevated serum prolactin level, found in 3–32% of patients with sarcoidosis, may be an indication of hypothalamic dysfunction. In fact, because neuroendocrinologic involvement is relatively common in individuals with CNS neurosarcoidosis, patients with more than just an isolated facial palsy probably merit a thorough endocrinologic evaluation with specific attention to hypothalamic hypothyroidism, hypocortisolism, and hypogonadism. Hypothalamic disorders that affect vegetative functions vary considerably. A disorder of thirst is the most common hypothalamic disorder related to neurosarcoidosis and is attributed to a change in the hypothalamic “osmostat” (Stuart et al., 1978). More rarely, the syndrome of inappropriate secretion of antidiuretic hormone can occur (Kirkland et al., 1983). Neurosarcoidosis-induced disruptions of hypothalamic function have also been described to cause disorders of appetite, libido, temperature control, and weight regulation (Heffernan et al., 1971; Lipton et al., 1977; Stuart et al., 1978; Bihan et al., 2007).

MASS LESIONS Intraparenchymal mass lesions due to sarcoidosis may present in two ways: an isolated mass (Fig. 21.4) or masses occurring in any cerebral area or multiple cerebral nodules. Such multiple nodules may actually represent an inflammatory process in the Virchow– Robin spaces. Subdural plaque-like masses may also occur and are discussed in the above section on meningeal manifestations of neurosarcoidosis. Calcifications may also be seen. Although historically, intraparenchymal mass lesions were considered rare, CT and MRI have shown parenchymatous disease to be more frequent than previously thought.

ENCEPHALOPATHY/VASCULOPATHY The diffuse encephalopathy and vasculopathy associated with neurosarcoidosis are not well understood

Fig. 21.4. Axial T1-weighted MRI with gadolinium enhancement demonstrating a frontal intracerebral mass that was proven by biopsy to be neurosarcoidosis.

manifestations of sarcoidosis. Moreover, it is often difficult both clinically and pathologically to separate clearly between these entities. In fact, these two manifestations of neurosarcoidosis frequently coexist. For these reasons, we find it best to consider them as a single overlapping entity, sarcoidosis-associated encephalopathy/vasculopathy, but recognize that in individual patients one form or the other may predominate (Stern et al., 1985). The diffuse encephalopathy/vasculopathy found in neurosarcoidosis can involve the cerebral hemispheres or basilar regions. Patients may present with a delirium, personality change, or isolated memory disturbance due to focal or diffuse parenchymal inflammation (Fig. 21.5) (Ho et al., 1979; Cordingley et al., 1981; Thompson and Checkley, 1981). Clinical findings correlate with the extent of enhancement on imaging studies and increased signal intensity on T2-weighted or FLAIR MRI. Sarcoidosis of the CNS has also been associated with a multifocal relapsing encephalopathy. Encephalopathic patients may have perivascular inflammation or granulomas infiltrating both arteries and veins and extending into brain parenchyma. Several investigators have observed granulomatous small-vessel arteritis in patients with neurosarcoidosis (Meyer et al., 1953; Herring and Urich, 1969). Large arteries have also been involved in the granulomatous process (Fig. 21.6).


311

Fig. 21.5. Axial T2-weighted MRI shows a large area of abnormality in the temporal lobe that proved at biopsy to be sarcoidosis manifesting with a focal encephalopathy/ vasculopathy.

Fig. 21.6. Coronal T1-weighted MRI with gadolinium enhancement showing gyral enhancement in the distribution of a posterior cerebral artery branch stroke.

Although disease is rarely evident on angiography, angiographic changes suggestive of vasculitis, as well as an ill-defined occlusive process, have been documented (Brown et al., 1989; Corse and Stern, 1990). About 20% of all patients with neurosarcoidosis develop cognitive or behavioral problems (Joseph and Scolding, 2007). These manifestations vary widely and include hallucinations (Hayashi et al., 1995), psychosis (Bona et al., 1988; O’Brien et al., 1994), depression, delirium (Bourgeois et al., 2005), and dementia (Friedman and Gould, 2002). In some cases the behavioral or cognitive changes may be a consequence of parenchymal brain involvement by sarcoidosis and in others may be a consequence of treatment with agents such as corticosteroids. In addition, psychiatric issues may arise from or be contributed to by living with neurosarcoidosis as a potentially chronic, relapsing remitting, or progressive disease. Not all such patients demonstrate overt changes in imaging studies or ancillary diagnostic testing to substantiate the exact pathologic nature of their problems. It is advisable to carefully evaluate any sarcoidosis patient who develops cognitive or behavioral symptoms with neurosarcoidosis because treatment with corticosteroids can improve outcomes (Hoitsma et al., 2004). Depression deserves particular attention as it is

reported in 60–66% or all sarcoidosis patients (Chang et al., 2001; Cox et al., 2004). Transient ischemic attacks and ischemic stroke due to neurosarcoidosis have been reported (Caplan et al., 1983; Sethi et al., 1986; Corse and Stern, 1990; Olugemo and Stern, 2008). Ischemic stroke is usually a consequence of inflammation involving large or small arteries, but other causes include compressive perivascular mass lesions and emboli from sarcoidosis-associated cardiomyopathy or cardiac arrhythmias. Caplan et al. emphasized the arterial and venous involvement of the meninges and parenchyma in the angiitic form of sarcoidosis and related this to observable perivascular lesions in the optic fundus (Caplan et al., 1983). Also, dural sinus obstruction causing intracranial hypertension has been related to inflammation from sarcoidosis (Byrne and Lawton, 1983; Chapelon et al., 1990). The vasculopathy associated with neurosarcoidosis can also cause intracranial hemorrhage, including intracerebral and subarachnoid bleeding (Olugemo and Stern, 2008).

SEIZURES Seizures are another important manifestation of CNS parenchymal disease due to neurosarcoidosis. They have

312


been reported in up to 20% of patients with neurosarcoidosis and may be focal or generalized. Seizures have been correlated with a poor prognosis in neurosarcoidosis (Delaney, 1980). However, it has only recently been shown that the cause of this poor prognosis is not directly due to seizures. Instead, the poor prognosis of neurosarcoidosis patients with seizures is due to that fact that seizures are an indicator for the presence of severe central nervous system parenchymal disease or hydrocephalus. This more severe pathology actually accounts for the poor outcomes because of its higher risk for progressive or recurrent disease or death (Delaney, 1980; Krumholz et al., 1991). Importantly, seizures in patients with neurosarcoidosis are reported to be relatively easy to control if the underlying central nervous system inflammatory process can be effectively treated (Krumholz et al., 1991).

MYELOPATHY Spinal cord involvement is another form of CNS parenchymal sarcoidosis. Spinal cord sarcoidosis may manifest as intramedullary, intradural extramedullary, or extradural granulomatous masses (Junger et al., 1993; Olek et al., 1995). Intramedullary spinal cord disease can also present with a myelitis that is analogous to the encephalopathy/vasculopathy that is described earlier as a form of intracranial sarcoidosis. Intraspinal mass lesions due to sarcoidosis present with a nonspecific imaging appearance (Fig. 21.7). Also, spinal arachnoiditis may occur. In addition, sarcoidosis may present as a

radiculopathy, polyradiculopathy, or cauda equina syndrome. Finally spinal cord sarcoidosis may appear, typically in the late stages of spinal cord disease as focal spinal cord atrophy (Junger et al., 1993; Olek et al., 1995).

Peripheral neuropathy Although sarcoidosis mainly impacts cranial nerves, peripheral neuropathy is not uncommon and is noted in about 20% of patients with neurosarcoidosis (Galassi et al., 1984; Hoitsma et al., 2004). A variety of peripheral neuropathies are reported including chronic sensorimotor, pure motor or sensory neuropathies, small fiber neuropathy, mononeuritis multiplex, and acute Guillain–Barre´ like syndromes (acute demyelinating polyneuropathy) (Zuniga et al., 1991; Hoitsma et al., 2002). The most common form of noncranial sarcoidosis, peripheral neuropathy, appears as a chronic sensorimotor neuropathy of the axonal type (Zuniga et al., 1991), although the authors have been impressed with the frequency of small fiber neuropathy encountered in a referral setting. Sarcoidosis neuropathy typically begins months to years following an initial diagnosis of sarcoidosis, but, in some instances, symptoms of neuropathy precede the discovery of systemic sarcoidosis. The neuropathy is usually mild and classically manifests with distal paresthesias, decreased vibration and proprioception sensation, and reduced ankle jerk reflexes (Zuniga et al., 1991). Pain and impaired temperature sensibility may also be a feature of the neuropathy, and

Fig. 21.7. MRI of the spine, (A) axial, (B) sagittal, with gadolinium enhancement, demonstrates an intraspinal mass due to sarcoidosis.

NEUROLOGIC MANIFESTATIONS OF SARCOIDOSIS documentation of such a small fiber neuropathy is facilitated with skin biopsies with quantitative nerve fiber analysis (Hoitsma et al., 2002).

Myopathy Sarcoidosis can also directly involve muscles and present with myopathy of various types. Manifestations of sarcoidosis myopathy include acute, subacute, or chronic weakness, and fatigue, muscle pain, and palpable muscle nodules. Symptomatic muscle disease is described in 1.4–2.3% of patients with neurosarcoidosis (Jamal et al., 1988). Sarcoidosis can manifest as a polymyositis with predominantly proximal weakness, muscular atrophy, or occasional pseudohypertrophy. Severe muscle disease can also result in fibrosis and cause contractures. Muscle involvement by noncaseating granulomas can be demonstrated with muscle biopsy. Incidental noncaseating granulomas have been found in blind muscle biopsy specimens in sarcoidosis patients without clinical evidence of muscle disease. Differentiating between sarcoidosis myopathy, polymyositis or dermatomyositis, and granulomatous myopathy may also be difficult, and it is important to evaluate other organ systems carefully for evidence of sarcoidosis (Hewlett and Brownell, 1975; Stjernberg et al., 1981; Ando et al., 1985).

CLINICAL COURSE The clinical course and prognosis for neurosarcoidosis varies but is somewhat predictable. For example, some two-thirds of patients have a monophasic neurologic illness; the remainder have a chronically progressive or remitting relapsing course. Those with a monophasic illness typically have an isolated cranial neuropathy, most often involving the facial nerve, or an episode of aseptic meningitis. Those with a chronic course usually have CNS parenchymal disease, hydrocephalus, multiple cranial neuropathies (especially involving cranial nerves II and VIII), peripheral neuropathy, and myopathy (Luke et al., 1987; Ferriby et al., 2001). Patients with CNS parenchymal disease or hydrocephalus are at highest risk of death, either from the inflammatory process itself or complications of therapy. Mortality of neurosarcoidosis is historically reported as 5–10% percent (Stern et al., 1985; Luke et al., 1987; Krumholz et al., 1991), but the authors believe that the mortality rate has decreased with modern management strategies. Since most patients with neurosarcoidosis are treated with immunosuppressive agents, it is impossible to determine the untreated natural history of the disorder (Stern et al., 1985). Moreover, since treatment with corticosteroids clearly benefits some patients, it would be

313

inappropriate to withhold treatment for patients with neurosarcoidosis. Although therapy can certainly improve patients in the short term, it is not certain exactly how it changes the ultimate outcome of the disease. Even in severely ill or impaired patients the inflammatory process may spontaneously subside over time. Other patients with remitting relapsing or progressive disease can become severely incapacitated even with aggressive treatment. Patients benefit from a comprehensive approach to care based on an understanding of the full clinical spectrum of neurosarcoidosis, an appreciation of the whole range of treatment options, and the anticipation of complications, such as those relating to corticosteroid treatment.

INVESTIGATIONS Differential diagnosis Diagnostically, there are principally two clinical scenarios with which patients with neurosarcoidosis present, and they occur with roughly equal frequency: (1) the patient without established sarcoidosis presents with a clinical picture suggestive of neurosarcoidosis; in this situation the major goal is to establish the presence of systemic sarcoidosis; (2) the patient with already established systemic sarcoidosis develops neurologic symptoms; here, one should focus on confirming that the neurologic problem is due to neurosarcoidosis rather some other cause. In each case the diagnostic approach may be somewhat different. A diagnosis of neurosarcoidosis is most secure when one has confirming pathologic tissue evidence from the site of neurologic involvement. However, neural tissue is not readily accessible for pathologic examination. Therefore it can be helpful to grade the likelihood of a patient having neurosarcoidosis based on available clinical and pathologic evidence in a given patient. Several groups have proposed diagnostic criteria for a potential diagnosis of sarcoidosis in an organ system based on particular tests and clinical presentations (Judson et al., 1999; Zajicek et al., 1999). The following definitions adapted from Zajicek and colleagues are proposed for grading the certainty of a diagnosis of neurosarcoidosis (Zajicek et al., 1999): ●

●

Possible neurosarcoidosis. The clinical syndrome and diagnostic evaluation suggest neurosarcoidosis. However, alternate conditions, such as an infection or malignancy, are not excluded or there is no pathologic confirmation of systemic sarcoidosis. Probable neurosarcoidosis. The clinical syndrome and diagnostic evaluation suggest neurosarcoidosis. Alternate diagnoses are excluded. There is pathologic confirmation of systemic sarcoidosis.

314

A. KRUMHOLZ AND B.J. STERN Definite neurosarcoidosis. The clinical presentation eyes or mouth are noted, lacrimal, parotid, or salivary is suggestive of neurosarcoidosis. Other diagnoses gland involvement should be suspected. are excluded. In addition, there is supportive nerIn addition, a thorough ocular examination is indivous system pathology, or criteria for “probable cated to search for uveitis, retinal periphlebitis, or neurosarcoidosis” are met, with the patient having superficial lesions for conjunctival biopsy. Ocular finda beneficial response to immunotherapy over a ings in sarcoidosis include lacrimal gland inflamma1 year or longer observation period. tion, conjunctival nodules, iritis, uveitis, retinal lesions (vascular sheathing, granulomas, vascular occlusions, and chorioretinitis), and optic disc patholLaboratory investigations ogy (edema, nodules, granulomas, and atrophy) (Sharma, 1984; James and Williams, 1985). In general, ESTABLISHING SYSTEMIC SARCOIDOSIS evidence supports that a conjunctival biopsy is a reasonWhen a patient without documented systemic sarcoidoable option for a blind biopsy when trying to confirm sis develops a clinical syndrome suggestive of neurosarthe presence of sarcoidosis. It is relatively innocuous coidosis, confirming evidence for sarcoidosis should be in nature and has a moderately high yield even in sought in other organ systems. Such systemic disease can patients without clinical eye findings specific for sarbest be documented when a thorough, systematic evalucoidosis (Leavitt and Campbell, 1998). ation based on the known natural history of sarcoidosis is Various laboratory measures have been described as undertaken (Table 21.1). Consequently, histologic supabnormal in sarcoidosis. Although none is highly specific port for a diagnosis of sarcoidosis should be pursued folfor sarcoidosis, they can be of some value. These abnorlowing leads obtained from the patient’s clinical malities include increased serum angiotensin-converting evaluation and those statistics (Iannuzzi et al., 2007). enzyme (SACE), increased serum g globulins, anemia, Since corticosteroid therapy can mask signs of systemic leukopenia thrombocytopenia, hypercalcemia, and disease, treatment should be postponed, if possible, hypercalciuria, and hepatic and renal dysfunction while a search for systemic disease is initiated. (Katz, 1983; Israel and Kataria, 1985). Pulmonary involvement is so common in sarcoidosis One of the laboratory tests most associated with sarthat this should be the first organ system to consider coidosis is the SACE. SACE is thought to be produced by when attempting to establish the presence of systemic alveolar macrophages and epithelioid giant cells and, in sarcoidosis. Indeed, nearly 90% of patients with sarcoideffect, reflects the “granulomatous load” of a patient. osis are reported to show radiographic evidence of pulIts sensitivity is not that high, however, with just 50– monary involvement (Rodan and Putman, 1983). Still, 60% of active sarcoidosis patients showing abnormalialthough an abnormal chest roentgenogram or CT scan ties. In addition, SACE is not very specific because it is often seen in sarcoidosis and can be supportive eviis abnormal in other conditions such as liver disease, diadence for that diagnosis, these imaging findings are betes mellitus, hyperthyroidism, systemic infection, not necessarily specific or pathognomonic for sarcoidomalignancy and Gaucher’s disease (Lieberman, 1985). sis. Additional evidence to support pulmonary involveAlthough much controversy exists as to the proper place ment can be obtained from pulmonary function of the SACE assay in the diagnosis of sarcoidosis, most testing including diffusion capacity. Moreover, when investigators accept that it is a useful marker for syschest imaging or pulmonary function studies suggest temic disease activity (Lieberman, 1985). pulmonary involvement, a diagnosis of sarcoidosis is confirmed by obtaining histologic evidence of sarcoidoCONFIRMING NEUROSARCOIDOSIS sis with a transbronchial biopsy (Iannuzzi et al., 2007). Now we turn to the second category of neurosarcoidosis Apart from the lungs, other organ systems also deserve patients, those with well documented systemic sarcoidoconsideration. Because of ease of access, early considersis who develop neurologic disease suspected to be due ation should be given to skin or lymph node biopsy of susto neurosarcoidosis. In general they require less of a syspicious sites. Clinical information should also guide the temic assessment, but they still merit careful appraisal evaluation of other organ systems for evidence of sysbecause neurosarcoidosis can be confused with many temic sarcoidosis. For instance, nasal mucosal, conjunctiother neurologic diseases, and it is usually not reasonval, lacrimal gland, liver, and muscle biopsies can be able to confirm neurosarcoidosis with biopsy tissue from considered on the basis of individualized clinical assessthe nervous system. ments. In particular, if a patient has impaired smell or However, such patients deserve consideration of taste, nasal or olfactory nerve disease might be present. disease entities that may mimic neurosarcoidosis, parIn that situation, an endoscopic nasal and sinus examinaticularly infection and neoplasia (Table 21.3) (Choi et al., tion can show abnormal mucosa amenable to biopsy. If dry ●

NEUROLOGIC MANIFESTATIONS OF SARCOIDOSIS Table 21.3 Differential diagnosis of neurosarcoidosis Inflammatory/autoimmune/ idiopathic Multiple sclerosis Neuromyelitis optica Lymphocytic hypophysitis Idiopathic pachymeningitis Rasai–Dorfman disease Wegener’s granulomatosis Sj€ ogren’s-related vasculitis CNS lupus Primary central nervous system angiitis Vogt–Koyanagi–Harada disease IgG4-related sclerosing pachymeningitis

Infectious Fungal meningitis Histoplasmosis Tuberculous meningitis Tuberculoma Toxoplasmosis Neurobrucellosis Central nervous system Whipple’s disease Neurosyphilis Neuroborrelliosis HIV infection CMV ventriculitis CMV polyradiculopathy PML Cryptoccocal meningitis

Neoplastic CNS lymphoma Leptomeningeal metastasis Germ cell tumors Craniopharyngioma Primary CNS neoplasia Glioma Meningiomas CNS, central nervous system; CMV, cytomegalovirus; PML, progressive multifocal leukoencephalopathy; HIV, human immunodeficiency virus.

2010; Desestret et al., 2010; Kosakai et al., 2010). Once such disorders are excluded, the patient can reasonably be treated for neurosarcoidosis. Nevertheless, if a patient does not respond to treatment as expected, the diagnosis of neurosarcoidosis should be revisited and a more extensive evaluation considered, including biopsy, to exclude other etiologies for the symptoms (Table 21.3). Spinal fluid analysis is another useful method for diagnosis of neurosarcoidosis. Over 50% of patients with CNS sarcoidosis will have some cerebrospinal fluid (CSF) abnormality (Stern et al., 1985; Oksanen, 1986). Reported abnormalities include an elevated CSF pressure, a high protein level, hypoglycorrhachia, and a predominantly mononuclear pleocytosis of up to several hundred cells. In addition, some patients have oligoclonal bands in the CSF or an elevated IgG index. However, none of these abnormalities are specific for neurosarcoidosis. When performing a CSF examination, fluid should be sent for routine, fungal, and mycobacterium cultures,

315

Mycobacterium tuberculosis polymerase chain reaction (PCR), cytology, and flow cytometry, as appropriate to the clinical setting. CSF angiotensin-converting enzyme (ACE) activity tends to be raised in some 50% of untreated patients with CNS sarcoidosis (Oksanen et al., 1985), although abnormalities are also seen in the presence of infection and malignancy. The CSF ACE level may be abnormal even with steroid therapy, but less consistently than in untreated patients. The degree of elevation of CSF ACE may parallel the clinical course (Chan Seem et al., 1985; Oksanen et al., 1985). CSF ACE is thought to be produced by CNS granulomas, especially those near the meninges. A normal CSF ACE assay does not exclude the diagnosis of neurosarcoidosis. Moreover, CSF ACE’s diagnostic value is further limited because assay methodology and normative values have as yet not been well standardized. In summary, the CSF ACE assay is not a reliable measure of CNS sarcoidosis (Khoury et al., 2009). Evoked potentials, visual, auditory and somatosensory, can be useful in evaluating some patients with neurosarcoidosis. Visual evoked potentials (VER) can reveal abnormalities of the anterior visual pathways. VERs are often abnormal in patients with symptomatic optic nerve disease and may be abnormal in some patients with central nervous system sarcoidosis but no clinical evidence of optic nerve dysfunction (Streletz et al., 1981; Stern et al., 1985; Oksanen and Salmi, 1986). Streletz et al. report VER abnormalities in a high proportion of patients with neurosarcoidosis. Even some patients with sarcoidosis but without ocular or neurologic disease were found to have abnormal visual evoked potentials (Streletz et al., 1981). Their study suggests that subclinical neurosarcoidosis may be more common than previously realized. However, other observations of visual evoked potentials do not confirm a high incidence of abnormalities in asymptomatic patients (Stern et al., 1985). Some investigators suggest, instead, that visual evoked potential abnormalities in particular, and evoked potential disturbances in general, are useful in the evaluation of patients with symptomatic neurosarcoidosis because they demonstrate relevant abnormalities that may be used to monitor the clinical course of the disease (Stern et al., 1985). Similarly, brainstem auditory evoked responses (BAERs) are often abnormal in neurosarcoidosis patients with brainstem or eighth nerve symptoms and can be abnormal in neurologically ill patients without overt disease in these areas (Oksanen and Salmi, 1986). However, they are rarely, if ever, abnormal in patients without clinically evident CNS sarcoidosis. The degree of abnormality of an evoked potential can fluctuate depending on disease activity, and an abnormal pattern can even normalize (Oksanen and Salmi, 1986).

316 A. KRUMHOLZ AND B.J. STERN Somatosensory evoked responses have not been comresponding to treatment, and the diagnosis of neurosarprehensively studied in the assessment of patients with coidosis is in doubt, a biopsy of involved tissue should be sarcoidosis. Preliminary evidence suggests that their considered, if feasible, to exclude alternate diagnoses or clinical utility is similar to VER and BAER in confirming confirm the diagnostic impression of sarcoidosis. the involvement of a specific sensory system that may be clinically affected (Stern et al., 1992). Neuroimaging Nerve conduction studies in patients with sarcoidosis Although not specifically diagnostic, imaging tests can neuropathy usually reveal changes compatible with an support a presumptive diagnosis of neurosarcoidosis. axonal neuropathy, though slowing can be more pronounced and suggestive of demyelinating disease The preferred imaging technique is MRI with contrast (Challenor et al., 1984; Galassi et al., 1984). Electromyogenhancement (Sherman and Stern, 1990; Stern et al., raphy can demonstrate a denervation pattern in patients 2010; Terushkin et al., 2010). T1-weighted images prowith a neuropathy or radiculopathy and myopathic vide less useful information than T2-weighted and changes in patients with a symptomatic myopathy fluid-attenuated inversion recovery (FLAIR) studies. (Ando et al., 1985). MRI may reveal a characteristic With T2-weighted and FLAIR imaging, areas of increased signal intensity can be appreciated, especially “star-shaped” pattern for muscle nodules (Otake et al., in a periventricular distribution. Contrast administration 1990). Muscle or nerve biopsy is informative if the diagnosis of neuromuscular disease is in doubt and can be helps by demonstrating leptomeningeal enhancement as targeted to an area of enhancement. However, when well as parenchymal abnormalities (Figs 21.1 and 21.2). there is no symptomatic myopathy, blind muscle biopsy Enhancement presumably reflects a breakdown of the is reported to have a low yield and does not appear warblood–brain barrier and implies active inflammation. ranted (Aksamit and Norona, 1999). In patients with neuSpine MRI can visualize intramedullary disease, which ropathies and symptoms such as painful feet and appears as an enhancing fusiform enlargement, focal or diffuse enhancement, or atrophy (Fig. 21.7) (Junger impaired pain sensitivity and temperature appreciation, et al., 1993; Olek et al., 1995). Enhancing nodules or skin biopsy with quantitative sensory nerve terminal assessment is useful (Hoitsma et al., 2002). thickened or matted nerve roots can also be appreciated Patients with CNS sarcoidosis should be carefully in MRI images of the cauda equina. MRI scans of large questioned about symptoms relating to neuroendocrinomuscles are reported to highlight muscle granulomas logic or hypothalamic dysfunction, since problems in and provide a target for biopsy for pathologic confirmathese arenas are the most common brain parenchymal tion of sarcoidosis (Karadeli and Ulu, 2009). A common disorders found in sarcoidosis. Inquiry should focus MRI abnormality in neurosarcoidosis is the presence of multiple white matter changes, which has been confused on alterations in menses, libido, and potency as well as with multiple sclerosis (Fig. 21.8) (Scott et al., 2010). the presence of galactorrhea. Excessive thirst, a known manifestation of neurosarcoidosis, can be caused by Angiography usually has little to contribute in the evalhypothalamic damage that affects the normal osmouation of CNS sarcoidosis (Brooks et al., 1982; Brown static mechanism, or diabetes insipidus, hypercalcemia, et al., 1989). Granulomatous masses are typically avascuhypercalciuria, and corticosteroid-induced diabetes lar (Kendall and Tatler, 1978). Rarely, angiographic mellitus. Alterations in body temperature, sleep, and abnormalities are found such as smooth narrowing of appetite can also develop as a consequence of hypothathe caudal internal carotid artery (Kendall and Tatler, 1978), tapering occlusion of the anterior cerebral artery lamic sarcoidosis. Patients with CNS symptoms other (Corse and Stern, 1990), or multiple small vessel segmenthan transient cranial nerve palsies or aseptic meningitis should undergo a neuroendocrinologic evaluation tal narrowings suggesting an arteritis (Lawrence et al., including thyroid function tests (hypothalamic hypothy1974). Patients with a diffuse encephalopathy or strokeroidism needs to be considered), prolactin, testosterone like episodes are the most likely to demonstrate clinically or estradiol, follicle-stimulating hormone (FSH) and relevant angiographic findings (Oksanen, 1986). luteinizing hormone (LH), and cortisol assays. Gallium 67 scanning is a valuable imaging method for The diagnosis of sarcoidosis is most secure when initial detection of systemic sarcoidosis (Chapelon et al., 1990), but it has limited utility for longitudinal clinical based on pathology and when more than one organ sysfollow-up because of its expense and the potential raditem can be documented to be involved. However, since tissue from the nervous system is often difficult to ation exposure (Turner-Warwick et al., 1986). Detection secure for pathologic analysis and other tests are not of inflammation in the lacrimal, minor salivary, and diagnostic of neurosarcoidosis, the diagnosis of neuroparotid glands on a gallium scan is especially suggestsarcoidosis, despite the best efforts, must sometimes ive of sarcoidosis (Savolaine and Schlembach, 1990). remain tentative. That being said, if a patient is not Fluorodeoxyglucose positron emission tomography


317

Fig. 21.8. MRI findings of multifocal white matter type changes in neurosarcoidosis. (Reproduced from Scott et al., 2010, with permission.)

(FDG-PET) scans are a sensitive but nonspecific modality to search for systemic sarcoidosis. Case reports note increased or decreased FDG activity with PET scans of the brain and nervous system in patients with neurosarcoidosis, particularly the encephalopathy/vasculopathy form. These types of scans also have been used to follow response to treatment (Braun et al., 2008).

EPIDEMIOLOGY, PATHOPHYSIOLOGY, AND GENETICS The prevalence of sarcoidosis is estimated to be in the order of 60 per 100 000 population, while the annual incidence is approximately 11 per 100 000 population. In certain populations sarcoidosis may be even more common. However, the exact prevalence and incidence of sarcoidosis is difficult to validate because there is no single diagnostic test for sarcoidosis. Population differences for sarcoidosis have been described. For instance, there is a reported increased incidence of sarcoidosis in black compared to white people in the US; the disease also seems to be more severe in black people (James, 1983; Teirstein and Lesser, 1983). Certain areas of the world, such as Sweden, also seem to have a higher incidence of sarcoidosis, while it appears to be quite rare in other areas such as China or Southeast Asia (Tanoue et al., 1994). These studies also raise the possibility of a genetic predisposition to the development of sarcoidosis. Indeed, sarcoidosis does seem to occur with greater likelihood in some families, but as yet there is no well-defined genetic pattern determined and no consistent mode of inheritance has been discovered (James,

1983; Teirstein and Lesser, 1983; Rybicki et al., 2001b, 2003; Sverrild et al., 2008). One large study found that patients with sarcoidosis had a fivefold increased risk of having first-degree relatives with sarcoidosis (Rybicki et al., 2001b). In addition, although there are some indications that sarcoidosis may be slightly more common in women, overall there seems to be no clear sex predilection (James, 1983; Teirstein and Lesser, 1983). One recent genome association study identified variants in the annexin A11 gene as risk factors for sarcoidosis. This annexin 11 gene is involved in apoptosis and cellular proliferation (Hoffman et al., 2008). The current consensus is that susceptibility to sarcoidosis is polygenic and complex (Smith et al., 2008). Sarcoidosis can occur at any age. Although the likelihood of sarcoidosis is greatest in the third and fourth decades, the youngest reported patient is a 28-monthold child (Teirstein and Lesser, 1983; Weinberg et al., 1983). The exact frequency in children is difficult to assess because most countries understandably limit the use of routine screening chest radiography in children. Clinically, when sarcoidosis affects children, the distribution of organ involvement appears similar to that of reported adult cases, and the patterns of neurosarcoidosis seem analogous as well (Weinberg et al., 1983). Overall, our view of the neurologic manifestations of sarcoidosis is probably somewhat skewed toward the more serious manifestations of the disease. This is because neurosarcoidosis is a rare disorder and virtually all series and case reports are biased by patient selection issues. Individual case reports characteristically detail only specific, rare or extremely unusual clinical

318 A. KRUMHOLZ AND B.J. STERN presentations. Even large series are derived from either were reported by Gerdes et al. (1992) and Ghossein autopsy data, which are obviously biased toward the et al. (1994). The PCR has been applied to CSF samples most severe types of problems, or referral centers, which in patients with neurosarcoidosis; but only one of six again tend to see the more unusual or complex patients. samples from two patients was positive for M. There are no definite provoking or exacerbating factuberculosis (Liedtke et al., 1993). The debate on the relators reliably associated with sarcoidosis. The granulomationship of sarcoidosis and mycobacterial infection contous lesions in sarcoidosis are similar to those caused by tinues (Joyce-Brady, 1992). infectious agents such as mycobacteria and fungi, inorPCR analysis may be useful in the occasional patient ganic materials such as zirconium and beryllium, and in whom a distinction between sarcoidosis and tubercuthose seen in association with hypersensitivity reactions lous infection is difficult. The rapidity of PCR compared to various organic agents (Tanoue et al., 1994). It has been to conventional culture techniques can be an important speculated that perhaps infectious agents, or organic consideration. However, the most useful setting in which agents, such as pine pollen, or inorganic substances to apply PCR may be in the evaluation of “sarcoidosis” may be etiologic factors for sarcoidosis (Desai and patients with progressive disease in spite of “optimal” Simon, 1985; James, 1994; Tanoue et al., 1994). In particimmunosuppressive disease. The ability to exclude ular, interest has centered on the possibility that sarcoidmycobacterial infection in these patients can be reassurosis is the consequence of an unusual host response to a ing before proceeding with more intense immunosupcommon antigen or infectious agent. Of the various pospressive therapy. sible etiologies, mycobacterial infections have received Despite the fact that a specific etiologic agent has not the most attention. Propionibacterium acnes is another been established to cause sarcoidosis, it appears that saragent that has been considered (Hoitsma et al., 2004). coidosis probably relates to an initiating agent persisting The potential relationship of sarcoidosis to tuberculoat the site of disease activity for a period of time and trigsis is an important one that has evolved over the years. gering a hypersensitivity or otherwise amplified type of Initially, sarcoidosis and tuberculosis were distinguished immune reaction in a potential, perhaps genetically susas separate diseases by various diagnostic tests. Skin ceptible, host (Fig. 21.9). testing for tuberculosis exposure, staining tissue for Although the precise etiology of sarcoidosis remains mycobacteria organisms, and culture of mycobacteria unknown, major strides have been made in understandfrom tissue and body fluids were the mainstays of difing its pathogenesis. For example, there is growing eviferential diagnosis between these diseases. Based on dence that sarcoidosis is caused by heightened immune the absence of evidence of mycobacterial infection in processes at sites of disease activity (Rocklin, 1983; most patients with sarcoidosis, the clear dichotomy Kataria, 1985; Daniele et al., 1986; Semenzato and between the two diseases had been thought to be well Agostini, 1994; Moller and Chen, 2002; Iannuzzi et al., established during the 1970s and 1980s. Still a perplexing 2007). This contrasts sharply with earlier concepts that challenge has developed with the possibility that mycorelated sarcoidosis to impaired immunity and to generbacterial infection, manifesting in an atypical fashion, alized anergy (Lieberman et al., 1985). may cause sarcoidosis. Our current understanding of the immunopathology With the advent of DNA techniques to detect of sarcoidosis derives largely from studies of pulmonary mycobacterial infection, the issue of the relationship sarcoidosis. The initial lesion in pulmonary sarcoidosis of sarcoidosis and mycobacterial infection has been reoappears to be an alveolitis, an inflammation of the alvepened. Mitchell et al. reported evidence for the presence olar structures of the lung. This inflammation provides of Mycobacterium tuberculosis in sarcoidosis tissue the appropriate environment for granuloma formation. samples using a DNA/RNA hybridization technique. Evidence indicates that these granulomas contain actiHybridization was 4.8 times more likely in sarcoidosis vated mononuclear cells that primarily have a secretory patients than in controls (Mitchell et al., 1985). Saboor rather than a phagocytic role. There is also activation of and colleagues, using PCR, found M. tuberculosis other inflammatory cells, particularly CD4 lymphocytes, DNA in 10 of 20 patients with sarcoidosis and an addithat congregate at sites of disease activity and secrete tional 20% of patients had nontuberculosis mycobactevarious cytokines, including interleukin 2, interleukin rial DNA detected (Saboor et al., 1992). On the other 1, interferon g, and tumor necrosis factor (Rocklin, hand, also using PCR, Bocart et al. detected evidence 1983; Kataria, 1985; Semenzato and Agostini, 1994). of M. tuberculosis in only 2 of 24 patients (Bocart Most probably, processes similar to what one finds in et al., 1992) and Popper et al. found M. tuberculosis in the lung underlie the pathogenesis of other forms of saronly 2 of 15 sarcoidosis patients (Popper et al., 1994). coidosis, including neurosarcoidosis (Fig. 21.9). The presence of other mycobacteria species could not Despite the lack of research addressing the specific be documented (Bocart et al., 1992). Similar findings immunopathogenesis of neurosarcoidosis, there is reason


319

Fig. 21.9. Immunopathogenic mechanisms postulated as responsible for development of the granulomatous inflammation found in neurosarcoidosis.

to believe it is similar to that of systemic sarcoidosis. Therefore the information available from studies of bronchoalveolar lavage fluid and other sites of inflammation represent a reasonable model of processes found in the nervous system. Although there are occasional discrepant results among investigators, a consensus about the immunologic mechanisms behind sarcoidosis is emerging. Sarcoidosis can be thought of as an inflammatory response to an as yet unidentified foreign antigen (Hunninghake and Crystal, 1981; Agostini et al., 2000). As described elsewhere in this chapter, there has long been a suspicion that sarcoidosis is an inflammatory disorder; fever, malaise, weight loss, cutaneous anergy, and polyclonal hyperglobulinemia are all consistent with this hypothesis. The central pathologic hallmark of sarcoidosis, the granuloma, consists of macrophages, macrophage-derived epithelioid cells, and multinucleated giant cells which secrete cytokines (Myatt et al., 1994). About this central core exist CD4 and CD8 lymphocytes, B lymphocytes, plasma cells, and fibroblasts, in proportions that vary with time. The lymphocytes are thought to be stimulated by antigen presentation by activated macrophages present at sites of inflammation (Semenzato and Agostini, 1994). The macrophages exhibit class II major histocompatibility antigen and the a chain of the IL-2 receptor and release tumor necrosis factor (TNF)-a, interleukin-6 (IL-6), interleukin-1 (IL-1) b, granulocytemacrophage colony-stimulating factor, fibronectin,

1,25 vitamin D, prostaglandin, and angiotensinconverting enzyme (Steffen et al., 1993; Bost et al., 1994; Fireman and Topilsky, 1994). Prostaglandin E2 decreases granuloma size whereas TNF and IL-1 increase granuloma size (Pueringer et al., 1993). Polyclonal B cells aggregate about granulomas (Fazel et al., 1992). Occasional plasma cells are also found, producing IgM and IgA more so than IgG. k and l light chains are present. Presumably local cytokine production leads to the proliferation and differentiation of the B cells, which then contribute to the inflammatory process. Genetic analysis of the CD4 T cell receptor, and its potential relationship to class II major histocompatibility molecules, has been pursued to search for clues as to the antigen or antigens inciting the inflammation (Weissler, 1994). If sarcoidosis were caused by a specific antigen, whether exogenous or endogenous, there might be an over-representation of specific CD4 markers expressed in patients with sarcoidosis. Indeed, an abundance of V (variable) b3, b5, b8, b14, b15, a2.3, g9, and d1 lymphocyte phenotypes is described (Forman et al., 1993; Fireman and Topilsky, 1994; Klein et al., 1995). An association between Va2.3 and HLA-DR3(17), DQ2 haplotype has been documented (Grunewald et al., 1994). There is also a bias in the expression of C (constant) b1 elements (Tamura et al., 1991) and an increased percentage of g/d cells (Raulf et al., 1994) with the Vd1 marker (Forrester et al., 1993). Unfortunately, the inciting antigen or antigens remain unknown.

320


Of critical importance is whether chronic inflammation requires continuous antigen presentation or whether an intrinsic genetic predisposition allows persistent inflammation if the inciting antigen is no longer locally available. Nonetheless, evidence suggests that CD4 cells are persistently activated and proliferate in a limited clonal response to a specific repertoire of as yet unknown foreign or native antigens (Hol et al., 1993; Semenzato et al., 1993). As the reactive monocytes and macrophages form granulomas, ultimately irreversible, obliterative fibrosis can develop. Furthermore, small foci of ischemic necrosis can be found, probably as a consequence of vascular compromise due to perivascular inflammation. Importantly, these granulomas are not specific for sarcoidosis and indistinguishable or nearly identical lesions occur in a variety of other conditions that must be excluded before a diagnosis of sarcoidosis can be made with certainty (Thomas and Hunninghake, 1987).

PATHOLOGY Noncaseating granulomas and the accompanying diffuse mononuclear cell infiltrates that are characteristic of sarcoidosis can be found in any part of the neuraxis including peripheral nerve or muscle (Fig. 21.10). The most common site of inflammation is the meninges, especially in the basal region of the brain (Fig. 21.11) (Delaney, 1977; Waxman and Sher, 1979). Sarcoidosis granulomas can be distributed widely or concentrated in one or more areas to form a mass. Although sarcoidosis is not usually considered to be a primary vasculitis (Moore and Cupps, 1983), arteriolar and venous infiltration does occur and can lead to infarction (Douglas and Maloney, 1973; Delaney, 1977; Urich, 1977; Caplan et al., 1983; Mirfakhraee et al., 1986). The granulomatous inflammation found pathologically can correlate directly with clinical signs and symptoms or may be subclinical and unexpressed (Waxman and Sher, 1979; Manz, 1983). Central nervous system (CNS) involvement can be conceptualized as an inflammatory process primarily affecting the leptomeninges. Inflammation may spread along the Virchow–Robin perivascular spaces to invade the brain or spinal cord, or remain more localized to involve the cranial nerves (Waxman and Sher, 1979; Trombley et al., 1981; Mirfakhraee et al., 1986). Presumably, at an early stage, much of the active inflammation is reversible. Ultimately, as the disease progresses, irreversible fibrosis can develop which leads to permanent neurologic damage and persisting neurologic deficits. Inflammation can also extend to the CSF pathways, leading to hydrocephalus. Brain or spinal cord disease can take the appearance of discrete granulomatous mass lesions or a diffuse encephalopathy/vasculopathy. The

Fig. 21.10. (A) Photomicrograph at 200 magnification of the brain showing an intraparenchymal noncaseating or non-necrotizing sarcoid granuloma. (B) Photomicrograph at 400 magnification of a sarcoid granuloma in brain demonstrating a multinucleated giant cell.

Fig. 21.11. Coronal midfrontal section of the brain of a patient with neurosarcoidosis showing thickening and inflammatory changes of the basal and optic region. (Courtesy of Dr. Steven C. Bauserman.)

NEUROLOGIC MANIFESTATIONS OF SARCOIDOSIS hypothalamic region is the most common site for parenchymal disease (Fig. 21.11). Granulomas are apparent in the epineurium and perineurium of peripheral nerve in symptomatic patients. The endoneurium may contain a mononuclear cell infiltrate. Perivascular and vascular inflammation can be seen in the epineural and perineurial vessels. All nerve fiber sizes can be affected. Documentation of a small fiber neuropathy is facilitated with skin biopsies with quantitative nerve fiber analysis (Hoitsma et al., 2002). There seems to be a predominantly axonal neuropathy with only a minor component of segmental demyelination. The exact mechanism of peripheral nerve damage in sarcoidosis seems to vary, but it includes injury resulting from vascular compromise, direct compression from granulomas, and immunologic factors that may affect the peripheral nerve’s axons (Galassi et al., 1984). Muscle pathology is common in sarcoidosis. Muscle biopsy of symptomatic patients reveals typical noncaseating granulomas, and there may be more diffuse inflammation with muscle fiber degeneration and regeneration and fibrosis (Hewlett and Brownell, 1975; Ando, 1989). Moreover, asymptomatic noncaseating granulomas have been found in up to one-half of all sarcoidosis patients having a muscle biopsy (Stjernberg et al., 1981).

MANAGEMENT Treatment There are no scientifically rigorous studies comparing various treatments for neurosarcoidosis. However, most experts agree that corticosteroid therapy is the mainstay

321

of treatment and is indicated for any patient without a specific contraindication to it (Patel et al., 2007). Decisions on the optimal therapeutic dose and duration of therapy should be guided by the patient’s clinical course, expected natural history or prognosis, and adverse treatment effects. A treatment paradigm is given in Figure 21.12. Treatment with corticosteroids is widely accepted and recommended for all forms of neurosarcoidosis. In support of this, many individual case reports and series provide evidence that in the short term it can produce impressive responses and alleviate symptoms. However, it is not absolutely certain that treatment changes the natural history and long-term course of neurosarcoidosis. A major theoretical goal for long-term treatment with corticosteroids is to diminish the irreversible fibrosis that can develop and to minimize tissue ischemia that might result from perivascular inflammation. However, once corticosteroid or other immunotherapy is begun, it need not continue indefinitely, particularly not at extremely high doses. Theoretically, with time, the inflammatory process can recede, allowing immunosuppressive therapy to be withdrawn. Specific proposed corticosteroid treatment regimens for the various manifestations of neurosarcoidosis vary (Stern, 1992; Constantino et al., 2000), and some recommendations are detailed below.

PERIPHERAL FACIAL NERVE PALSY The most common neurologic manifestation of neurosarcoidosis is a peripheral facial nerve palsy. The facial weakness may improve without any specific treatment. A controlled trial of treatment has not been done, and the

Fig. 21.12. Therapeutic algorithm. (Reproduced from Terushkin et al., 2010, with permission.)

322


long-term prognosis for these patients seems favorable. It seems reasonable to give a short course of 2–12 weeks of prednisone treatment for a peripheral facial nerve palsy due to sarcoidosis. The recommended first week’s prednisone dose is in the range of 0.5–1.0 mg/kg daily (or 40– 60 mg/day), followed by a gradual reduction leading to discontinuation of prednisone over the ensuing weeks. General supportive care, as used for any patient with a peripheral facial nerve palsy, should be provided, with special attention directed at protecting the eye. Occasionally a patient who was improving will relapse as the dose of prednisone is decreased. In that situation, the dose can be increased for a week and a slower taper attempted.

OTHER CRANIAL NERVE PALSIES AND ASEPTIC MENINGITIS

Patients with other cranial neuropathies and aseptic meningitis can be managed with a corticosteroid protocol similar to that described for peripheral facial nerve palsy. However, often more than 2 weeks of therapy is needed. In particular, patients with optic neuropathy or cranial nerve VIII dysfunction typically may need more prolonged, aggressive therapy. However, even prolonged, aggressive therapy may not prevent irreversible optic and eighth nerve damage. Aseptic meningitis may also respond to a short course of prednisone, 0.5–1.0 mg/kg/day. However, the goal of therapy should not be complete clearing of an asymptomatic cerebrospinal fluid pleocytosis. Attempts to completely normalize the spinal fluid can needlessly expose patients to the adverse side-effects of the high doses of corticosteroids. Indeed, there is no evidence that clearing the cerebrospinal fluid of an abnormal cellular response necessarily corresponds to clinical wellbeing in patients with aseptic meningitis who are otherwise asymptomatic.

HYDROCEPHALUS Asymptomatic ventricular enlargement usually does not require immediate therapy. Mild, symptomatic hydrocephalus can respond to corticosteroid treatment, and prolonged therapy is often appropriate. However, there is some evidence that neurosarcoidosis patients who fail to respond to immunotherapy need to be carefully followed for symptomatic hydrocephalus requiring a shunt (Scott et al., 2007). Life-threatening hydrocephalus or corticosteroid-resistant hydrocephalus requires ventricular shunting or a ventricular drain. Unfortunately, patients can evolve from having mild hydrocephalus to severe, life-threatening disease quite rapidly. Scott et al. have highlighted the potential for life-threatening clinical deterioration in patients undergoing a diagnostic lumbar puncture for hydrocephalus (Scott, 2000; Scott

et al., 2007). Therefore, patients and care providers should be well educated as to the symptoms of acute progressive hydrocephalus and know how to obtain prompt emergency care (Scott, 2000; Scott et al., 2007). At times, high-dose intravenous corticosteroid therapy (methylprednisolone 20 mg/kg/day for 3 days) can stabilize a patient with life-threatening hydrocephalus, although usually prompt surgical intervention with a ventricular drain or ventriculoperitoneal shunt is necessary (Fig. 21.3). Shunt placement is not without risk in this patient population, which is why “prophylactic” shunting in asymptomatic patients with hydrocephalus is not readily advocated. Shunt obstruction from the inflamed CSF and ependyma is common, and placement of a foreign object in the CNS of an immunosuppressed host predisposes to infection.

MASS LESION Patients without known systemic sarcoidosis who develop a brain or spinal cord mass are frequently biopsied for determination of the diagnosis. If pathologic examination suggests noncaseating granulomas, appropriate cultures should be obtained. Also, since granulomas can surround some malignancies, adequate biopsy samples are encouraged (Peoples et al., 1991). However, as discussed previously, surgical excision of the mass should usually be avoided since surgery is rarely curative and patients have been known to develop new major fixed neurologic deficits or deteriorate from surgical excision of a granulomatous mass (Cahill and Salcman, 1981). After biopsy, an empiric trial of highdose corticosteroid treatment is appropriate for most such patients. Furthermore, patients should be evaluated for systemic sarcoidosis, but it is possible that corticosteroid therapy can mask the presence of systemic disease. If a patient with known sarcoidosis develops a CNS mass, an empiric trial of corticosteroid therapy is appropriate, especially if infection and malignancy are reasonably excluded by CSF examination and other means. If the patient does not respond to corticosteroid therapy, a biopsy should be pursued. In either of these scenarios, if a mass progressively enlarges in spite of corticosteroid therapy, surgical exploration should be strongly considered to evaluate the possibility of an infection or malignancy (Peoples et al., 1991). Some of the important entities that can be mistaken for sarcoidosis are listed in Table 21.3.

CNS PARENCHYMAL DISEASE Corticosteroid therapy can improve the status of patients with a diffuse encephalopathy/vasculopathy or a CNS mass lesion. Only rarely does immunosuppressive

NEUROLOGIC MANIFESTATIONS OF SARCOIDOSIS treatment improve neuroendocrine dysfunction or vegetative symptoms. Corticosteroid treatment for CNS parenchymal disease and other severe neurologic manifestations of sarcoidosis usually starts with prednisone 0.5–1.0 mg/kg daily. The higher doses are used in patients with particularly severe disease. Such patients should be observed on high corticosteroid doses for 2–4 weeks to ascertain the clinical response. Also, these types of patients often require more prolonged therapy, and so even when appropriate to lower it, prednisone should be tapered very slowly. Patients should be followed by imaging or similar relevant studies at intervals of not less than 6 months. The prednisone dose can be tapered by 5 mg decrements every 2 weeks as the clinical course is monitored. Neurosarcoidosis may exacerbate at a low prednisone dose approximating 10 mg or less daily. Some patients exhibit an individual therapeutic lower limit or dosage below which worsening can almost be predicted, based on their prior response to medication tapers. Once a dose of prednisone of 10 mg daily can be achieved, the patient should be evaluated for evidence of clinical as well as subclinical worsening of disease. Clinical disease can be monitored by symptoms, but subclinical disease can also be monitored. For patients with CNS disease, an enhanced MRI scan is useful. Intense enhancement, in the meninges for example, suggests that neurosarcoidosis is active and further decreases in corticosteroid dose may lead to a clinical exacerbation. Other manifestations of neurosarcoidosis can be evaluated for subclinical deterioration on an individualized basis; for instance, evaluating nerve conduction studies or a serum creatine kinase level. On the other hand, persistent mild CSF abnormalities are usually not an indication for continuing or escalating high-dose corticosteroid therapy, since patients can remain quite functional in spite of an abnormal CSF. Efforts to “normalize” the CSF often require powerful immunosuppression, with its attendant adverse effects. If the inflammation appears quiescent, then even a low daily prednisone dose of about 10 mg can be tapered further by 1 mg every 2–4 weeks. If a patient has a clinical relapse, the dose of prednisone should be doubled (unless the dose is very modest, in which case a prednisone dose of 10–20 mg daily can be prescribed). The patient should than be observed for approximately 4 weeks before another taper is contemplated. Patients may require multiple cycles of higher and lower corticosteroid dosage during attempts to taper medications. For sustained remission, some patients require 12 months or more of therapy. Still, this effort to taper corticosteroid therapy is warranted since the disease can become quiescent and without attempts at withdrawing medication, patients may be needlessly exposed to the harmful side-effects of long-term, high-dosage corticosteroids.

323

Daily dosing of prednisone is usually superior to alternate day therapy. However, if a patient has been clinically stable for a long time on a modest daily prednisone dose, an attempt can be made to gradually change the patient to alternate day therapy.

SEIZURES In sarcoidosis patients with seizures, the seizures are generally not a major limiting problem and can usually be well controlled with antiepileptic medication if the underlying central nervous system inflammatory reaction can be effectively treated. However, seizures have been shown to be an indication of, or marker for, the presence of parenchymatous involvement of the brain, which is in itself a very serious manifestation of neurosarcoidosis (Luke et al., 1987; Krumholz et al., 1991). Consequently, it is important to recognize this clinical relevance of seizures in patients with neurosarcoidosis. Seizures are a useful, often early warning sign that a patient may have one of the more serious forms of neurosarcoidosis such as an intracranial mass lesion (Krumholz et al., 1991).

PERIPHERAL NEUROPATHY AND MYOPATHY Patients with a peripheral neuropathy or myopathy can respond to a several week course of corticosteroids, usually beginning with prednisone 1 mg/kg/day (or approximately 60 mg/day). Here, too, prolonged treatment may be indicated. Corticosteroids should be tapered slowly, as discussed above.

OTHER ISSUES Critically ill patients may be considered for high-dose intravenous corticosteroid therapy (Soucek et al., 1993). A short course of methylprednisolone 20 mg/kg daily intravenously for 3 days, followed by prednisone 1.0–1.5 mg/kg/day for 2–4 weeks, is occasionally warranted. Aggressive corticosteroid therapy can also clarify whether a patient on a modest corticosteroid dose can incrementally improve. The clinical response should be closely observed and predefined measures of improvement carefully monitored. One or two target measures such as a specific clinical sign, symptom, functional assessment or neurodiagnostic test can be used to judge the response to treatment. For instance, the results of psychometric tests or a timed walk can be used for clinical assessment in some patients, while in patients with an intracranial mass, MRI imaging may be a helpful neurodiagnostic measure. This type of treatment regimen can be used to judge a response over a relatively short time period.

324

A. KRUMHOLZ AND B.J. STERN radiation therapy. These agents or radiation therapy have Surgical considerations been not been studied in a scientifically controlled manner If the patient has a CNS sarcoid mass lesion unresponagainst placebos or comparable treatments. Such trials are sive to high-dose intravenous corticosteroids, surgical difficult because neurosarcoidosis is an extremely rare resection may be appropriate in life-threatening situadisorder with very varied clinical presentations. Consetions. There are several other clinical scenarios in which quently, it has been difficult to conduct controlled treatneurosurgical intervention may be necessary for patients ment studies for either conventional or alternative with neurosarcoidosis. First, without biopsy confirmatherapies. tion of affected neural tissue, a diagnosis of neurosarPractically, consideration should be given to introduccoidosis must remain somewhat in question. Biopsy of ing alternative therapy whenever a patient shows signs of peripheral nerve or muscle for such a purpose is relaserious corticosteroid side-effects or requires frequent tively easy to obtain. In contrast, CNS tissue for patholarge increases in corticosteroid dosage to control symplogic examination is much more difficult to secure in toms. Alternative treatment with an immunosuppressive such patients. Despite this difficulty, if a patient with agent or irradiation is a logical adjunctive therapy for presumptive CNS sarcoidosis is not responding to treatrefractory neurosarcoidosis given what we understand ment as anticipated, consideration should be given to of the immunopathogenic mechanisms of the disease. biopsy of a demonstrated CNS lesion so as to exclude Alternative therapy may allow a gradual decrease in corother disorders that may be confused with neurosarcoiticosteroid dosage to prevent or minimize corticosteroid dosis and to confirm the diagnosis of sarcoidosis (Fidler complications, often without deterioration in clinical staet al., 1993; Cheng et al., 1994). Neurosarcoidosis is a contus. Rarely, however, can corticosteroids completely be dition that does not lend itself to definitive neurosurgical eliminated. Still, some patients may continue to deteriotreatment, and aggressive surgical interventions have rate despite combination therapy and the reported mortalrarely produced sustained benefit (Cahill and Salcman, ity in neurosarcoidosis remains a substantial 5–10% (Stern 1981). Still, two major situations in which surgical interet al., 1985; Luke et al., 1987; Krumholz et al., 1991). ventions are warranted or deserve consideration include An algorithm for the use of alternative treatment for patients with intracranial mass lesions unresponsive to refractory neurosarcoidosis is presented in corticosteroids, or with hydrocephalus. Figure 21.12. The choice of alternative treatment should Patients with symptomatic hydrocephalus not responbe determined by the potential adverse effects of the sive to corticosteroid therapy need to be considered for therapy and the extent of systemic disease. It is wise ventricular shunting as previously discussed. The surgito choose an agent whose adverse effects spare an cal management of such patients can be quite challengorgan or organ system that may already be comproing because of the potential for shunt obstruction and the mised (Table 21.4). For example, cyclosporine should risk of infection. In addition, cerebral ventricles or parts usually be avoided in patients with significant hypertenof ventricles can sometimes become “trapped” or isosion or renal disease, while azathioprine may not be the lated from the rest of the ventricular system necessitatbest choice for patients with liver or hematologic ing the placement of multiple shunts. problems. Mycophenolate mofetil is an inhibitor of monophosphate dehydrogenase, which is an important enzyme Alternative treatments in purine synthesis, attenuates B and T cell proliferation. Treatment alternatives to corticosteroids must sometimes It has primarily been used to treat mucocutaneous and be considered for patients with neurosarcoidosis gastrointestinal manifestations of sarcoidosis, but has (Fig. 21.12 and Table 21.4). Experience in this area is limnow also shown good tolerability and efficacy in the ited and optimal strategies are not well established. Inditreatment of neurosarcoidosis (Chaussenot et al., cations for the use of alternate treatments include 2007). It is increasingly used as the alternative drug of contraindications to corticosteroids as initial therapy, serichoice by the authors because of its ease of use, low toxous adverse chronic corticosteroid effects, and progresicity profile, and apparent efficacy. sive disease activity in spite of aggressive corticosteroid Methotrexate is a folic acid analog that may be useful therapy. Medication alternatives to corticosteroids that for treating patients with sarcoidosis or neurosarcoidohave been used to treat sarcoidosis include cyclosporine, sis. It is typically initiated in conjunction with corticosteazathioprine, methotrexate, mycophenolate mofetil, roid therapy, with corticosteroids gradually reduced or cyclophosphamide, chloroquine and hydroxychloroquine, even eventually totally withdrawn in some instances. thalidomide, infliximab, rituximab (Bomprezzi et al., When used in conjunction with corticosteroids, metho2010), and chlorambucil. One nondrug treatment that trexate is reported to reduce the necessary corticosteroid has also been reported to have some limited success is dose by half (Baughman et al., 2000). In another study, a

Table 21.4 Therapy for neurosarcoidosis (adapted from Terushkin et al., 2010, with permission) Therapy

Dose

Prednisone

Initially 40–80 mg/day or 0.5–1.0 mg/kg/; slow taper based on response. Chronic dose often in the range of 0.1 mg/kg/day

Methylprednisolone

1 g daily or 20 mg/kg IV daily for 3–5 days 7.5 mg weekly, with 2.5 mg increases every 2 weeks until 10–25 mg/week is reached and continued for 4–6 weeks; continue for 12 months if needed 500 mg twice daily for 1 week. If tolerated, 1000 mg twice daily. If persistent disease activity after 2–3 months, 1500 mg twice daily (B.J.S., personal observation, January, 2009) 2 mg/kg for up to 200 mg/day; adjust dosage if WBC decreases below 3000 or if platelets drop below 100 000. Consider target dose to achieve WBC of 3000–4000 or total lymphocyte count of 1000 250 mg twice daily (chloroquine) or 200 mg twice daily (hydroxycholorquine)

Methotrexate

Mycophenolate mofetil

Azathioprine

Chloroquine, hydroxychlorquine

Infliximab

3–5 mg/kg IV at 0, 2, and 6 weeks for induction; then dose at every 4–8 þ week

Cyclophosphamide

Initial dose 500 mg IV with escalation to 750 mg and maintenance of 1000 mg; oral doses of 2–5 mg/kg/day may also be used 1.5 Gy/day: total dose 20 Gy

Radiotherapy

Laboratory tests

Notable adverse reactions Hypertension, adrenal suppression, hyperglycemia, psychosis, insomnia, gastritis, weight gain, osteoporosis As above

CBC with diff, LFTs

Hepatotoxicity, interstitial pneumonitis, bone marrow suppression, anemia

CBC with diff, LFTs

Leukopenia, anemia, esophagitis, gastritis, gastrointestinal hemorrhage

CBC with diff, LFTs, pretreatment assay

Bone marrow suppression, hypersensitivity syndrome (fever, myalgia, arthralgia, cutaneous eruptions), possible risk for lymphoma or leukemia Irreversible retinal damage, seizures, deafness, tinnitus, lichen planus-like eruptions, alopecia, hypertension, ECG changes Anaphylaxis, risk for infection, rash, hepatotoxicity, pancytopenia, delayed allergic reactions, potential to exacerbate congestive heart failure, CNS demyelination/ inflammation Hemorrhagic cystitis, amenorrhea, azotemia, cardiac toxicity, infertility, neutropenia

WBC with diff. LFTs, urinalysis

Transient enlargement of CNS lesion and encephalopathy, necrosis, atrophy, calcification, tumor inductions, CSF fistula, demyelination

CBC, complete blood count; with diff., with differential; LFTs, liver function tests; IV, intravenous; Gy, gray; CNS, central nervous system; WBC, white blood cell count.

326 A. KRUMHOLZ AND B.J. STERN 61% response rate for methotrexate as a substitute for hydroxychloroquine with either improving or stabilizing corticosteroids is reported (Lower et al., 1997). Methosymptoms or signs (Sharma, 1998). In order to prevent trexate is a medication that is easy to take and administer the adverse side-effect of severe hemolysis, glucose-6 and that is usually well tolerated, but when using it, regphosphate dehydrogenase deficiency should be excluded ular evaluation of complete blood counts and liver funcbefore the administration of these agents. Retinal damage tion tests are required because of the risk of blood can be a consequence of treatment with chloroquine or dyscrasias and hepatotoxicity. In addition, folic acid suphydroxychloroquine, so periodic ophthalmologic examiplementation at a dosage of 1 mg daily is recommended nations are recommended to detect problems early. Such to prevent potential adverse effects such as macrocytic ocular problems are more likely with chloroquine than anemia, potential liver function abnormalities, gastroinwith hydroxychloroquine. It is reasonable to avoid these testinal difficulties, and cardiovascular disease related drugs in patients with already impaired vision. to a drug-associated elevation in plasma homocysteine. Infliximab is a recombinant monoclonal antibody Cyclophosphamide is an alkylating agent that can be against TNF-a, and has been used for neurosarcoidosis used in severe cases of neurosarcoidosis. For example, in therapy. TNF-a is an important macrophage-secreted one study of seven corticosteroid-refractory neurosarcytokine involved in granuloma formation, so its supcoidosis patients treated with intravenous cyclophosphapression may be of benefit in treating various manifesmide, well-documented favorable clinical responses tations of sarcoidosis. Numerous case reports and were observed in 57%, and the dose of required concomsmall series attest to the value of infliximab in patients itant corticosteroids was able to be reduced by 50% with corticosteroid-resistant sarcoidosis (Doty et al., (Doty et al., 2003). When utilizing cyclophosphamide, 2005; Baughman et al., 2006) and neurosarcoidosis leukocyte counts need to be monitored, and infections (Pettersen et al., 2002; Carter et al., 2004; Sollberger pose a major risk. Although the intravenous form of et al., 2004; Sodhi et al., 2008; Moravan and Segal, cyclophosphamide is safer, the oral form is a reasonable 2009). In particular, infliximab can be an effective agent alternative when individual patient circumstances in patients with severe disease, in which instance it can be limit or preclude using intravenous cyclophosphamide. used in conjunction with prednisone and mycophenolate When using oral cyclophosphamide, particular attention mofetil. Sometimes infliximab can be used to “bridge” a should be directed to patient hydration to prevent hemseriously ill patient until the mycophenolate mofetil, or orrhagic cystitis. other drug, exerts a beneficial effect (e.g., 2–3 months). Another medication option is azathioprine, a purine The authors have also successfully used adalimumab in analog that inhibits DNA synthesis and suppresses the selected patients, particularly those with a small fiber immune system. Azathioprine is a reasonable second polyneuropathy. agent, in part because of its moderate cost. It can be Cyclosporine has been demonstrated to be beneficial to started at a dose of 50 mg/day for 1 week and if this is some patients with disease refractory to corticosteroid tolerated, the dose can be progressively increased to therapy (Stern et al., 1992). Cyclosporine inhibits the 2–3 mg/kg/day or to achieve a white blood cell count amplification of the CD4 cell immune response, which of 3500 or absolute lymphocyte count or 1000. Azathiois suspected to be of importance in the pathogenesis of prine may excessively suppress the bone marrow, and sarcoidosis (Kahan, 1989). Although it is ineffective as blood counts should be frequently monitored as the isolated treatment for pulmonary sarcoidosis, it is an dose is escalated. Blood and liver profiles should be effective adjunct to corticosteroid therapy for refractory monitored, and there is a long-term risk for leukemia neurosarcoidosis (Stern et al., 1992). Other immunosupor lymphoma. An increase in the red blood cell mean pressive agents offer less specific antagonism of the corpuscular volume to above 100 is also indicative of immune system. However, cyclosporine is difficult to an adequate azathioprine effect. Some 10% of patients manage because of the need to adjust the dose to “theradevelop an acute idiosyncratic reaction with fever and peutic” blood levels, so it is rarely used as a first-line abnormal liver function necessitating discontinuation agent. of the drug. An assay to exclude a thiopurine methylIn comparison to the agents described above, there is transferase deficiency should be performed before not as much experience with the other immunosuppresstarting azathioprine because such a deficiency predissive medication alternatives to corticosteroid treatment poses an individual to toxicity. for neurosarcoidosis. However, in general, side-effects Chloroquine and hydroxychloroquine, which are to these alternative medications are limited, predictable agents commonly used to treat malaria, also have effiand respond to withdrawal of the offending agent. It is cacy against neurosarcoidosis (Sharma, 1998). In one even possible to restart the medication in some cases report of 12 neurosarcoidosis patients refractory to corwithout recurrent side-effects. That being said, all of ticosteroid therapy, 83% responded to chloroquine or the alternative agents can predispose the patient to

NEUROLOGIC MANIFESTATIONS OF SARCOIDOSIS infection and there is a low, but finite, increase in the risk of cancer, particularly lymphoma. For instance, mycophenolate mofetil is now associated with a mandated “black box” warning from the US Food and Drug Administration (FDA) for progressive multifocal leukoencephalopathy. One of the major advantages of such alternative immunosuppressant medications is that they may enable a gradual reduction of corticosteroid to about 15–30% of the stabilizing corticosteroid dose, thereby offering a “corticosteroid sparing” effect (Stern et al., 1992). Even with the use of alternative medication, there is the possibility of recurrent symptoms when corticosteroids are decreased below 10 mg/day. Attempts to withdraw corticosteroid totally may result in worsening symptoms, suggesting that alternative immunosuppressant medication is best used as an adjunct to corticosteroid. On the other hand, some patients do quite well on corticosteroid alone after alternative medication has been withdrawn, and some can be maintained on the alternative agent alone. We present here one approach that we favor, which is adapted from Terushkin et al., for how to use these various options for treating patients with neurosarcoidosis (Fig. 21.12 and Table 21.4) (Terushkin et al., 2010). In symptomatic neurosarcoidosis, corticosteroids are the preferred initial therapy. Should patients fail such initial therapy and exhibit severe neurologic symptoms, mycophenolate mofetil or intravenous infliximab are the drugs of choice. Because mycophenolate mofetil reaches its optimal therapeutic efficacy only after 2–3 months, if a rapid effect is required, infliximab is preferred. When one is confronted with severe life-threatening neurosarcoidosis symptoms, we also recommend consideration of cyclophosphamide. For patients with mild or moderate symptomatic neurosarcoidosis who fail corticosteroids, we suggest considering mycophenolate mofetil, azathioprine, or methotrexate. However, it should be noted that these latter drugs may take 3 or more months to become fully effective. In patients with asymptomatic neurosarcoidosis, low-dose prednisone might be considered. Although it is not possible to predict with absolute certainty which patients with neurosarcoidosis will have disease refractory to conventional corticosteroid treatment, there are patients with a particularly high-risk clinical profile. For instance, patients with central nervous system parenchymatous disease, such as mass lesions, or extensive encephalopathy/vasculopathy are at especially high risk. Such patients might benefit from the prompt use of adjunctive alternative treatment should they become refractory to corticosteroids or develop intolerable side-effects. Response to a particular alternative therapy can only be determined

327

by trial since a good clinical response is highly individualized and patients may show a good response to one agent even after an initial failure with another. There is a trend amongst physicians who manage neurosarcoidosis patients frequently to initiate early treatment with alternative agents in patients who are anticipated to require chronic, moderate to high dose corticosteroid therapy. This approach can minimize or avoid the accumulative toxicities associated with corticosteroid treatment. On occasion, corticosteroid therapy can be eliminated entirely, allowing the patient to be managed with an alternate agent alone. Most patients resistant to multiple therapies have a poor prognosis. If, on the other hand, an alternative medication is discontinued because of adverse drug effects, it is advisable to try other immunosuppressive medications before irradiation, as such patients may respond to other drugs. If a patient is stable for several months on prednisone or an alternate treatment, a slow taper of the alternate treatment can be pursued as the clinical course is monitored.

Radiation therapy Another viable option for some patients with refractory central nervous system neurosarcoidosis in whom medications have failed is radiation therapy. The mechanism of how radiation therapy works is not clear, but it is directed to the central nervous system and may destroy cells, such as macrophages and lymphocytes, that have direct roles in granuloma formation. The small number of studied and reported patients treated with CNS radiation precludes any definite conclusions about efficacy and optimal protocols for radiation therapy, but some mild to moderate effectiveness is reported (Bejar et al., 1985; Ahmad et al., 1992; Kang and Suh, 1999; Menninger et al., 2003). Some case reports suggest a beneficial response if total nodal and craniospinal irradiation is done (Ahmad et al., 1992). It appears that although radiation therapy can sometimes be of benefit, continued immunosuppressive therapy is usually necessary.

General supportive care Patients with neurosarcoidosis, and particularly those receiving treatment with immunosuppressive agents, require close attention to their general medical wellbeing. Potential adverse effects of treatment should be carefully sought. For example, an exercise and dietary program are often very beneficial to help avoid the weight gain associated with high-dose and long-term corticosteroid treatment. Rehabilitation services should be utilized as appropriate. Depression is not uncommon and treatment can be helpful.

328 A. KRUMHOLZ AND B.J. STERN Therapy of endocrinologic disturbances is important. CONCLUSIONS In particular, hypothyroidism and hypogonadism are Sarcoidosis is a multisystem granulomatous disorder of major problems in neurosarcoidosis and should be treaunknown etiology. Typical presentations include bilatted. In addition, since patients are often on protracted, eral hilar adenopathy, pulmonary infiltration, skin, and low-dose corticosteroid regimens, a supplemental dose eye lesions. Neurologic manifestations occur in more of corticosteroid is appropriate during periods of interthan 5% of sarcoidosis patients and may be the presentcurrent illness or stress. Hyperglycemia is a potential ing feature. side-effect of long-term, high-dose corticosteroid treatNeurosarcoidosis can manifest in a myriad of ways ment but fortunately it is not usually associated with including: cranial neuropathy, aseptic meningitis, mass ketoacidosis. Exercise and dietary programs are useful lesions, encephalopathy, vasculopathy, seizures, to manage the hyperglycemia, but occasionally oral hypothalamic-pituitary disorders, hydrocephalus, myehypoglycemic agents or insulin therapy may need to be lopathy, peripheral neuropathy and myopathy considered. (Wiederholt and Siekert, 1965; Delaney, 1977; Pentland Another concern is that patients with sarcoidosis are et al., 1985; Stern et al., 1985; Younger et al., 1988; at risk for osteoporosis. In particular, corticosteroid Joseph and Scolding, 2009; Pawate et al., 2009). Because therapy can cause osteoporosis. Moreover, treatment its etiology is unknown, its neurologic manifestations of osteoporosis is a challenge because sarcoidosis can are diverse, and its diagnosis cannot be readily concause hypercalcemia and hypercalciuria. Management firmed by laboratory tests, neurosarcoidosis poses major of osteoporosis in patients with neurosarcoidosis may clinical problems. Substantiation of a presumed diagnorequire reducing corticosteroid dosage when possible, sis of neurosarcoidosis requires biopsy tissue from the supplementation with calcium and vitamin D, hormonal affected region of the nervous system. However, this treatment and use of other appropriate agents. Because is not always feasible or judicious, so neurosarcoidosis sarcoidosis is associated with hypercalcemia and hyperis often a tentative diagnosis. In fact, the diagnosis of calciuria, caution should be used when prescribing vitaneurosarcoidosis is usually based on the identification min D and calcium supplementation to such patients. of characteristic neurologic findings in an individual Since the management of osteoporosis is a complex with proven systemic sarcoidosis as established by clinand rapidly changing field, it is suggested that when ical, imaging, or histologic findings. osteoporosis is identified an endocrinologist may need Even after proper diagnosis, treatment of neurosarto become involved (Schneyer, 1995). Screening should coidosis poses problems. Although corticosteroids are be done for osteoporosis, and, when appropriate, moniregarded as the foundation of treatment, they are not toring with bone density measurements should be conalways successful and have serious side-effects. Moresidered (American College of Rheumatology Ad Hoc over, some patients with neurosarcoidosis are refractory Committee on Glucocorticoid-Induced Osteoporosis, to conventional therapy, and approximately 5–10% die 2001). Serial measurements of bone density may be par(Stern et al., 1985; Luke et al., 1987; Krumholz et al., ticularly useful in judging an individual patient’s 1991; Pawate et al., 2009). Optimal management of progress. patients with neurosarcoidosis benefits from an underPatients commonly complain of fatigue, which can standing of the broad clinical spectrum of neurosarcoihave a multifactorial etiology apart from the impact dosis, appreciation of the ways to best confirm a of the systemic inflammation that characterizes sarcoiddiagnosis, and awareness of the full range of treatment osis. Patients should be evaluated for conditions such as options, including the use of alternative therapies such as sleep apnea syndrome, primary hypersomnia due to immunotherapy. hypothalamic disease, hypothyroidism, hypogonadism, We have learned a great deal about neurosarcoidosis corticosteroid myopathy, occult infection, cardiac or and its management. Nevertheless, neurosarcoidosis respiratory compromise, and depression. Aggressive remains a significant problem, and despite weight loss strategies can be helpful, if appropriate. tremendous advances in diagnosis and better insights It is obvious that patients with refractory neurosarcoiinto pathophysiology and treatment, it still poses dosis are at risk from both the adverse effects of the sarmajor challenges. In an effort to better address the coidosis inflammatory process and the dangers of complicated epidemiologic, diagnostic, and treatment treatment side-effects and complications. Consequently, issues involved in neurosarcoidosis, a consortium of if a patient with neurosarcoidosis is doing poorly and is committed experts to pool knowledge and refractory to treatment, not only should the original experience with this complex and challenging disorder diagnosis of sarcoidosis be questioned, but the patient has been proposed and could be extremely helpful should be evaluated for intercurrent complications, as (Stern et al., 2010). outlined in Table 21.4.


REFERENCES Agostini C, Adami F, Semenzato G (2000). New pathogenic insights into sarcoid granuloma. Curr Opin Rheumatol 12: 71–76. Ahmad K, Kim YH, Spitzer AR et al. (1992). Total nodal radiation in progressive sarcoidosis. Am J Clin Oncol 15: 311–313. Aksamit AJ, Norona F (1999). Neurosarcoidosis without systemic sarcoidosis. Ann Neurol 46: 471. American College of Rheumatology Ad Hoc Committee on Glucocorticoid-Induced Osteoporosis (2001). Recommendations for the prevention and treatment of glucocorticoid-induced osteoporosis. Arthritis Rheum 44: 1496–1503. Ando M (1989). Role of macrophages in the pathogenesis of sarcoidosis. Sarcoidosis 6: 76–78. Ando DG, Lynch JP, Fantone JC (1985). Sarcoid myopathy with elevated creatine phosphokinase. Am Rev Respir Dis 131: 298–300. Baughman RP, Winget DB, Lower EE (2000). Methotrexate is steroid sparing in acute sarcoidosis: results of a double blind, randomized trial. Sarcoidosis Vasc Diffuse Lung Dis 17: 60–66. Baughman R, Drent M, Kavaru M et al. (2006). Infliximab therapy in patients with chronic sarcoidosis and pulmonary involvement. Am J Respir Crit Care Med 174: 795–802. Bejar JM, Kerby GR, Ziegler DK et al. (1985). Treatment of central nervous system sarcoidosis with radiotherapy. Ann Neurol 18: 258–260. Bihan H, Christozova V, Dumas JL et al. (2007). Sarcoidosis: clinical, hormonal, and magnetic resonance imaging (MRI) manifestations of hypothalamic-pituitary disease in 9 patients and review of the literature. Medicine 86: 259–268. Bocart D, Lecossier D, De Lassence A et al. (1992). A search for mycobacterial DNA in granulomatous tissues from patients with sarcoidosis using the polymerase chain reaction [see comments]. Am Rev Respir Dis 145: 1142–1148. Bomprezzi R, Pati S, Chansakul C et al. (2010). A case of sarcoidosis successfully treated with rituximab. Neurology 75: 568–570. Bona JR, Fackler SM, Fendley MJ et al. (1988). Neurosardoidosis as a cause of refractory psychosis: a complicated case report. Am J Psychiatry 155: 1106–1108. Bost TW, Riches DW, Schumacher B et al. (1994). Alveolar macrophages from patients with beryllium disease and sarcoidosis express increased levels of mRNA for tumor necrosis factor-a and interleukin-6 but not interleukin-1. Am J Respir Cell Mol Biol 10: 506–513. Bourgeois JA, Maddock RJ, Rogers L et al. (2005). Neurosarcoidosis and delirium. Psychosomatics 46: 148–150. Braun JJ, Kessler R, Constaninesco A et al. (2008). 18-FDG PET/CT in sarcoidosis: a review and report of 20 cases. Eur J Nucl Med Mol Imaging 35: 1537–1543. Brooks BS, Gammal TE, Hungerford GD (1982). Radiologic evaluation of neurosarcoidosis: role of computed tomography. AJNR Am J Neuroradiol 3: 513–521. Brown MM, Thompson AJ, Wedzicha JA et al. (1989). Sarcoidosis presenting with stroke. Stroke 20: 400–405.

329

Byrne JV, Lawton CA (1983). Meningeal sarcoidosis causing intracranial hypertension secondary to dural sinus thrombosis. Br J Radiol 56: 755–757. Cahill DW, Salcman M (1981). Neurosarcoidosis: a review of the rarer manifestations. Surg Neurol 15: 204–211. Capellan JI, Cuellar Olmedo L, Martinez Martin J et al. (1990). Intrasellar mass with hypopituitarism as a manifestation of sarcoidosis. J Neurosurg 73: 283–286. Caplan L, Corbett J, Goodwin J et al. (1983). Neuroophthalmologic signs in the angiitic form of neurosarcoidosis. Neurology 33: 1130–1135. Carter JD, Valeriano J, Vasey FB et al. (2004). Refractory neurosarcoidosis: a dramatic response to infliximab. Am J Med 117: 277–279. Challenor YB, Felton CP, Brust JCM (1984). Peripheral nerve involvement in sarcoidosis: an electrodiagnostic study. J Neurol Neurosurg Psychiatry 47: 1219–1222. Chan Seem CP, Norfolk G, Spokes EG et al. (1985). CSF angiotensin-converting enzyme in neurosarcoidosis. Lancet 1: 456–457. Chang B, Steimel J, Moller DR et al. (2001). Depresssion in sarcoidosis. Am J Respir Crit Care Med 163: 320–324. Chapelon C, Ziza JM, Piette JC et al. (1990). Neurosarcoidosis: signs, course and treatment in 35 confirmed cases. Medicine 69: 261–276. Chaussenot AV, Bourg V, Chanalet S et al. (2007). Neurosarcoidosis treated with myocphenolate mofetil: two cases. Rev Neurol 163: 471–475. Cheng TM, O’Neill BP, Scheithauer BW et al. (1994). Chronic meningitis: the role of meningeal or cortical biopsy. Neurosurgery 34: 590–595. Choi SH, Lee SH, Khang SK et al. (2010). IgG4-related sclerosing pachymeningitis causing spinal cord compression. Neurology 75: 1388–1390. Cohen DM, Reinhardt RA (1982). Systemic sarcoidosis presenting with Horner’s syndrome and mandibular paresthesia. Oral Surg Oral Med Oral Pathol 53: 577–581. Colover J (1948). Sarcoidosis with involvement of the nervous system. Brain 71: 451–475. Colvin IB (2006). Audiovestibular manifestations of sarcoidosis: a review of the literature. Laryngoscope 116: 75–82. Constantino T, Digre K, Zimmerman P (2000). Neuroophthalmologic complications of sarcoidosis. Semin Neurol 20: 123–137. Cordingley GC, Navarro C, Brust JCM (1981). Sarcoidosis presenting as senile dementia. Neurology 31: 1148–1151. Corse AM, Stern BJ (1990). Neurosarcoidosis and stroke [letter]. Stroke 21: 152–153. Cox CE, Donohue JF, Brown CD et al. (2004). Health related quality of life of persons with sarcoidosis. Chest 125: 997–1004. Daniele RP, Rossman MD, Kern JA et al. (1986). Pathogenesis of sarcoidosis. State of the art. Chest 89 (Suppl): 174S–177S. Delaney P (1977). Neurologic manifestations in sarcoidosis: review of the literature, with a report of 23 cases. Ann Intern Med 87: 336–345.

330


Delaney P (1980). Seizures in sarcoidosis: a poor prognosis. Ann Neurol 7: 494. Deremee R (1985a). Chest roentgenology of sarcoidosis. In: J Lieberman (Ed.), Sarcoidosis. Grune and Stratton, Orlando, pp. 117–135. Desai SG, Simon MR (1985). Epidemiology of sarcoidosis. In: J Lieberman (Ed.), Sarcoidosis. Grune and Stratton, Orlando, pp. 25–37. Desestret V, Didelot A, Meyronet D et al. (2010). Neurosarcoidosis with diencephalitis and anti-Ma2 antibodies. Neurology 74: 772–774. Doty JD, Mazur JE, Judson MA (2003). Treatment of corticosteroid-resistant neurosarcoidosis with a shortcourse cyclophosphamide regimen. Chest 124: 2023–2026. Doty JD, Mazur JE, Judson MA et al. (2005). Treatment of sarcoidosis with infliximab. Chest 127: 1064–1071. Dougals AC, Maloney AFJ (1973). Sarcoidosis of the central nervous system. J Neurol Neurosurg Psychiatry 36: 1024–1033. El-Kassimi FA, Ashour M, Vijayaraghavan R (1990). Sarcoidosis presenting as recurrent left laryngeal nerve palsy. Thorax 45: 565–566. Fanburg BL (1983). Treatment of sarcoidosis. In: BL Fanburg (Ed.), Sarcoidosis and Other Granulomatous Diseases of the Lung. Marcel Dekker, New York, pp. 381–388. Fazel SB, Howie SEM, Krajewski AS et al. (1992). B lymphocyte accumulations in human pulmonary sarcoidosis. Thorax 47: 964–967. Ferriby D, De Seze J, Stojkovic T et al. (2001). Long-term follow up on neurosarcoidosis. Neurology 57: 927–929. Fidler HM, Rook GA, Johnson NM et al. (1993). Mycobacterium tuberculosis DNA in tissue affected by sarcoidosis. BMJ 306: 546–549. Fireman EM, Topilsky MR (1994). Sarcoidosis: an organized pattern of reaction from immunology to therapy. Immunol Today 15: 199–201. Forman JD, Silver RF, Klein JT et al. (1993). T cell receptor variable beta-gene expression in the normal lung and in active pulmonary sarcoidosis. Chest 103 (Suppl 2): 78S. Forrester JM, Newman LS, Wang Y et al. (1993). Clonal expansion of lung V delta 1þ T cells in pulmonary sarcoidosis. J Clin Invest 91: 292–300. Friedman SH, Gould DJ (2002). Neurosarcoidosis presenting as psychosis and dementia: a case report. Int J Psychiatry Med 32: 401–403. Galassi GM, Gibertoni M, Mancini A (1984). Sarocidosis of the peripheral nerve: clinical, electrophysiological and histological study of two cases. Eur Neurol 23: 459–465. Gelwan MJ, Kellen RI, Burde RM (1988). Sarcoidosis of the anterior visual pathway: successes and failures. J Neurol Neurosurg Psychiatry 5: 1473–1480. Gerdes J, Richter E, Rusch-Gerdes S et al. (1992). Mycobacterial nucleic acids in sarcoid lesion. Lancet 339: 1536–1537. Ghossein RA, Ross DG, Salomon RN et al. (1994). A search for mycobacterial DNA in sarcoidosis using the polymerase chain reaction. Am J Clin Pathol 101: 733–737.

Graham EM, Ellis CJK, Sanders MD (1986). Optic neuropathy in sarcoidosis. J Neurol Neurosurg Psychiatry 49: 756–763. Grunewald J, Olerup O, Perssson U et al. (1994). T-cell receptor variable region gene usage by CD4 þ and CD8þ T cells in bronchoalveolar lavage fluid and peripheral blood of sarcoidosis patients. Proc Natl Acad Sci 91: 4965–4969. Hayashi T, Onodera J, Nagata T et al. (1995). A case of biopsyproven sarcoid meningoencephalitis presented with hallucination, nominal aphasia and dementia. Rinsho Shinkeigaku 35: 1008–1011. Heerfordt C (1909). Uber eine “Febrid uveo-parotidea subchronica” an den Glandula parotis und der Uvea des Auges lokalisiert und haufig mit Paresencerebrospinaler Nerven kompliziert. Albrecht v. Graefes Arch Ophthalmol 70: 254–273. Heffernan AM, Cullen M, Towers R (1971). Sarcoidosis of the hypothalamus. Hormones 2: 1–12. Henkind P, Gottlieb MB (1973). Bilateral internal ophthalmoplegia in a patient with sarcoidosis. Br J Ophthalmol 57: 792. Herring A, Urich H (1969). Sarcoidosis of the central nervous system. J Neurol Sci 9: 405–422. Hewlett RH, Brownell B (1975). Granulomatous myopathy: its relationship to sarcoidosis and polymyositis. J Neurol Neurosurg Psychiatry 38: 1090–1099. Ho SU, Berenberg RA, Kim KS (1979). Sarcoid encephalopathy with diffuse inflammation and focal hydrocephalus shown by sequential CT. Neurology 29: 1161–1165. Hoffman S, Franke A, Fisher A et al. (2008). Genome-wide association study identifies ANXA 11 as a new susceptibility locus for sarcoidosis. Nat Genet 40: 1103–1106. Hoitsma E, Marziniak M, Faber CG et al. (2002). Small fibre neuropathy in sarcoidosis. Lancet 359: 2085–2086. Hoitsma E, Faber CG, Drent M et al. (2004). Neurosarcoidosis: a clinical dilemma. Lancet Neurol 3: 397–407. Hol BE, Hintzen RQ, Van Lier RAW et al. (1993). Soluble and cellular markers of T cell activation in patients with pulmonary sarcoidosis. Am Rev Respir Dis 148: 643–649. Hunninghake GW, Crystal RG (1981). Pulmonary sarcoidosis: a disorder mediated by excess helper T-lymphocyte activity at sites of disease activity. N Engl J Med 305: 429–434. Hybels RL, Rice DH (1976). Neuro-otologic manifestations of sarcoidosis. Laryngoscope 86: 1873–1878. Iannuzzi MC, Rybicki BA, Teirstein AS (2007). Sarcoidosis. N Engl J Med 359: 2153–2165. Ismail F, Miller JL, Kahn SE (1985). Hypo-thalamic-pituitary sarcoidosis. S Afr Med J 97: 139–142. Israel HL (1983). Diagnostic value of the Kveim reaction. In: BL Fanburg (Ed.), Sarcoidosis and Other Granulomatous Diseases of the Lung. Marcel Dekker, New York, pp. 273–286. Israel HL, Kataria YP (1985). Clinical aspects of sarcoidosis. In: J Lieberman (Ed.), Sarcoidosis. Grune and Stratton, Orlando, pp. 7–23. Jahrsdoerfer RA, Thompson EG, Johns MM et al. (1981). Sarcoidosis and fluctuating hearing loss. Ann Otol Rhinol Laryngol 90: 161–163.

NEUROLOGIC MANIFESTATIONS OF SARCOIDOSIS Jamal MM, Cilursu AM, Hoffman EL (1988). Sarcoidosis presenting as acute myositis. Report and review of the literature. J Rheumatol 15: 1868–1871. James DG (1983). Genetics and familial sarcoidosis. In: BL Fanburg (Ed.), Sarcoidosis and Other Granulomatous Diseases of the Lung. Marcel Dekker, New York, pp. 135–146. James DG (1994). State-of-the art lecture: Etiology of sarcoidosis. In: OP Sharma (Ed.), Proceedings of the Fourteenth International Conference on Sarcoidosis and Other Granulomatous Disorders. Edizioni Bongraf, Milan, pp. 43–58. James DG, Williams WJ (1985). Sarcoidosis and other granulomatous disorders. WB Saunders, Philadelphia. James DG, Turiaf J, Hosada Y et al. (1976). Description of sarcoidosis: report of the subcommittee on classification and definition. Ann N Y Acad Sci 278: 742. Joseph FG, Scolding NJ (2007). Sarcoidosis of the nervous system. Pract Neurol 7: 234–244. Joseph FG, Scolding NJ (2009). Neurosarcoidosis: a study of 30 new cases. J Neurol Neurosurg Psychiatry 80: 297–304. Joyce-Brady M (1992). “Tastes great, less filling”. The debate about mycobacteria and sarcoidosis. Am Rev Respir Dis 145: 986–987. Judson MA, Baughman RP, Tierstein AS et al. (1999). A case controlled etiologic study of sarcoidosis. Sarcoidosis Vasc Diffuse Lung Dis 16: 75–86. Junger SS, Stern BJ, Levine SR (1993). Intramedullary spinal sarcoidosis: clinical and magnetic resonance imaging characteristics. Neurology 43: 333–337. Kahan BD (1989). Cyclosporine. N Engl J Med 321: 1781–1782. Kang S, Suh JH (1999). Radiation therapy for neurosarcoidosis: report of three cases from a single institution. Radiat Oncol Investig 7: 309–312. Karadeli E, Ulu EM (2009). Acute sarcoid myositis with unusual radiographic findings. Diagn Interv Radiol 16: 232–235. Kataria YP (1985). Immunology of sarcoidosis. In: J Lieberman (Ed.), Sarcoidosis. Grune and Stratton, Orlando, pp. 39–63. Katz S (1983). Clinical presentation and natural history of sarcoidosis. In: BL Fanburg (Ed.), Sarcoidosis and Other Granulomatous Diseases of the Lung. Marcel Dekker, New York, pp. 3–36. Kendall BE, Tatler GLV (1978). Radiological findings in neurosarcoidosis. Br J Radiol 51: 81–92. Khoury J, Wellik KE, Demaerschalk BM et al. (2009). Cerebrospinal fluid angiotensin-converting enzyme for diagnosis of central nervous system sarcoidosis. Neurologist 15: 108–111. Kirkham TH, Kline LB (1976). Monocular elevator paresis, Argyll Robertson pupils and sarcoidosis. Can J Ophthalmol 11: 330–335. Kirkland JL, Pearson DJ, Goddard C et al. (1983). Polyuria and inappropriate secretion of arginine vasopressin in hypothalamic sarcoidosis. J Clin Endocrinol Metab 56: 269–272. Klein JT, Horn TD, Forman JD et al. (1995). Selection of oligoclonal V-beta-specific T cells in the intradermal response to Kveim–Siltzbach reagent in individuals with sarcoidosis. J Immunol 154: 1450–1460.

331

Kosakai A, Ito D, Yamada S et al. (2010). A case of definite IgG4-related pachymeningitis. Neurology 75: 1390–1392. Krumholz A, Stern BJ, Elyce EG (1991). Clinical implications of seizures in neurosarcoidosis. Arch Neurol 48: 842–844. Lawrence WP, Gammal TE, Pool WH Jr (1974). Radiological manifestations of neurosarcoidosis: report of three cases and review of literature. Clin Radiol 25: 343–348. Leavitt JA, Campbell J (1998). Cost-effectiveness in the diagnosis of sarcoidosis: the conjunctival biopsy. Eye 12: 959–962. Lieberman J (1985). Angiotensin-converting enzyme (ACE) and serum lysozyme in sarcoidosis. In: J Lieberman (Ed.), Sarcoidosis. Grune and Stratton, Orlando, pp. 145–159. Lieberman J, Kataria YP, Young JR (1985). Historical perspective of sarcoidosis. In: J Lieberman (Ed.), Sarcoidosis. Grune and Stratton, Orlando, pp. 1–5. Liedtke WE, May A, Dux R (1993). Mycobacterium tuberculosis polymerase chain reaction findings in neurosarcoidosis. J Neurol Sci 120: 118–119. Lipton JM, Kirkpatrick J, Rosenberg RN (1977). Hypothermia and persisting capacity to develop fever. Arch Neurol 34: 498–504. Lower EE, Broderick JP, Brott TG et al. (1997). Diagnosis and management of neurological sarcoidosis. Arch Intern Med 157: 1864–1868. Luke RA, Stern BJ, Krumholz A et al. (1987). Neurosarcoidosis: the long-term clinical course. Neurology 37: 461–463. Lukin RR, Chambers AA, Soleimanpour M (1975). Outlet obstruction of the fourth ventricle in sarcoidosis. Neuroradiology 10: 65. Manz HJ (1983). Pathobiology of neurosarcoidosis and clinicopathologic correlation. Can J Neurol Sci 10: 50–55. Menninger MD, Amdur RJ, Marcus RB Jr (2003). Role of radiotherapy in the treatment of neurosarcoidosis. Am J Clin Oncol 26: e115–e118. Meyer J, Foley J, Campagna-Pinto D (1953). Granulomatous angiitis of the meninges in sarcoidosis. Arch Neurol 69: 587–600. Mirfakhraee M, Crofford MJ, Guinot FC Jr et al. (1986). Virchow–Robin space: a path of spread in neurosarcoidosis. Radiology 158: 715–720. Missler U, Mack M, Nowack G (1990). Pituitary sarcoidosis. Klin Wochenschr 68: 342–345. Mitchell JD, Yap PL, Milne LA et al. (1985). Immunological studies on the cerebrospinal fluid in neurological sarcoidosis. J Neuroimmunol 7: 249–253. Moller DR, Chen ES (2002). What causes sarcoidosis? Curr Opin Pulm Med 8: 429–434. Moore PM, Cupps TR (1983). Neurological complications of vasculitis. Ann Neurol 14: 155–167. Moravan M, Segal BM (2009). Treatment of CNS sarcoidosis with infliximab and myocophenolate mofetil. Neurology 72: 337–340. Myatt NG, Coghill G, Morrison K et al. (1994). Detection of tumour necrosis factor in sarcoidosis and tuberculosis granulomas using situ hybridization. J Clin Pathol 47: 423–426. O’Brien GM, Baughman R, Broderick R et al. (1994). Paranoid psychosis due to neurosarcoidosis. Sarcoidosis 11: 34–36.

332


Obenauf CD, Shaw HE, Syndor CF (1978). Sarcoidosis and its ophthalmic manifestations. Am J Ophthalmol 86: 648–655. Oksanen V (1986). Neurosarcoidosis: clinical presentations and course in 50 patients. Acta Neurol Scand 73: 283–290. Oksanen V, Salmi T (1986). Visual and auditory evoked potentials in the early diagnosis and follow-up of neurosarcoidosis. Acta Neurol Scand 74: 38–42. Oksanen V, Fyhrquist F, Gronhagen-Riska C (1985). Angiotensin converting enzyme in cerebrospinal fluid: a new assay. Neurology 35: 1220–1223. Olek M, Jaitly R, Kuta A (1995). Sarcoidosis and spinal cord atrophy. Neurologist 1: 240–243. Olugemo OA, Stern BJ (2008). Stroke and neurosarcoidosis. In: LR Caplan (Ed.), Uncommon Causes of Stroke. 2nd edn. Cambridge University Press, New York, pp. 75–80. Osenbach RK, Blumenkopf B, Ramirez H Jr et al. (1986). Meningeal neurosarcoidosis mimicking convexity enplaque meningioma. Surg Neurol 26: 387–390. Otake S, Banno T, Ohaba S (1990). Muscular sarcoidosis: findings at MR imaging. Radiology 176: 145–148. Patel AV, Stickler DE, Tyor WR (2007). Neurosarcoidosis. Curr Treat Options Neurol 9: 161–168. Pawate S, Moses H, Sriam S (2009). Presentations and outcomes of neurosarcoidosis: a study of 54 cases. QJM 102: 449–460. Pentland B, Mitchell JD, Cull RE et al. (1985). FORD: central nervous system sarcoidosis. QJM 56: 457–465. Peoples DM, Stern BJ, Jiji V et al. (1991). Germ cell tumors masquerading as central nervous system sarcoidosis. Arch Neurol 48: 554. Pettersen JA, Zochodne DW, Bell RB et al. (2002). Refractory neurosarcoidosis responding to infliximab. Neurology 59: 1660–1661. Plotkin GR, Patel BR (1986). Neurosarcoidosis presenting as chronic lymphocytic meningitis. Pa Med 89: 36–37. Poole CJM (1984). Argyll Robertson pupils due to neurosarcoidosis: evidence for site of lesion. BMJ 289: 356. Popper HH, Winter E, Hofler G (1994). DNA of Mycobacterium tuberculosis in formalin-fixed, paraffinembedded tissue in tuberculosis and sarcoidosis detected by polymerase chain reaction. Am J Clin Pathol 101: 738–741. Pueringer RJ, Schwartz DA, Dayton CS et al. (1993). The relationship between alveolar macrophage TNF, IL-1, and PGE2 release, alveolitis, and disease severity in sarcoidosis. Chest 103: 832–838. Raulf MV, Liebers V, Steppert C et al. (1994). Increased gamma/delta-positive T-cells in blood and bronchoalveolar lavage of patients with sarcoidosis and hypersensitivity pneumonitis. Eur Respir J 7: 140–147. Rocklin R (1983). Cell-mediated immunity in sarcoidosis. In: BL Fanburg (Ed.), Sarcoidosis and other granulomatous diseases of the lung. Marcel Dekker, New York, pp. 203–224. Rodan BA, Putman CE (1983). Radiological alterations in sarcoidosis. In: BL Fanburg (Ed.), Sarcoidosis and Other Granulomatous Diseases of the Lung. Marcel Dekker, New York, pp. 37–75.

Rybicki BA, Iannuzzi MC, Frederick MM et al. (2001b). Familial aggregation of sarcoidosis: a case-control etiologic study of sarcoidosis (ACCESS). Am J Respir Crit Care Med 164: 2085–2091. Rybicki BA, Maliarik MJ, Poisson LM et al. (2003). The major histocompatibility complex gene region and sarcoidosis susceptibility in African Americans. Am J Respir Crit Care Med 167: 444–447. Saboor SA, Johnson NM, McFadden J (1992). Detection of mycobacterial DNA in sarcoidosis and tuberculosis with polymerase chain reaction. Lancet 339: 1012–1015. Savolaine ER, Schlembach PJ (1990). Gallium scan diagnosis of sarcoidosis in the presence of equivocal radiographic and CT findings: value of lacrimal gland biopsy. Clin Nucl Med 15: 198–199. Schlitt M, Duval ER, Bonnin J et al. (1986). Neurosarcoidosis causing ventricular loculation, hydrocephalus, and death. Surg Neurol 26: 67–71. Schneyer CR (1995). Abnormal bone metabolism in a neurological setting. Neurologist 1: 259–272. Scott TF (2000). Cerebral herniation after lumbar puncture in sarcoid meningitis. Clin Neurol Neurosurg 102: 26–28. Scott TF, Yandora K, Valeri A et al. (2007). Aggressive therapy for neurosarcoidosis: long-term follow-up of 48 treated patients. Arch Neurol 64: 691–696. Scott TF, Yandora K, Kunschere LJ et al. (2010). Neurosarcoidosis mimcry of multiple sclerosis: clinical, laboratory, and imaging characteristics. Neurologist 16: 386–389. Semenzato G, Zambello R, Trentin L et al. (1993). Cellular immunity in sarcoidosis and hypersensitivity pneumonitis. Chest 103 (Suppl): 139S–143S. Semenzato G, Agostini C (1994). State-of-the art lecture: immunopathologies of sarcoidosis. In: OP Sharma (Ed.), Proceedings of the Fourteenth International Conference on Sarcoidosis and Other Granulomatous Disorders. Edizioni Bongraf, Milan, pp. 59–71. Sethi KD, El Gammal T, Patel BR et al. (1986). Dural sarcoidosis presenting with transient neurologic symptoms. Arch Neurol 43: 595–597. Sharma OP (1984). Sarcoidosis: Clinical Management. Butterworths, London. Sharma OP (1998). Effectiveness of chloroquine and hydroxychloroquine in treating selected patients with sarcoidosis with neurological involvement. Arch Neurol 55: 1248–1254. Sherman JL, Stern BJ (1990). Sarcoidosis of the CNS: comparison of unenhanced and enhanced MR images. AJNR Am J Neuroradiol 11: 915–923. Smith G, Brownell I, Sanchez M et al. (2008). Advances in the genetics of sarcoidosis. Clin Genet 73: 401–412. Sodhi M, Pearson K, White ES et al. (2008). Infliximab therapy rescues cyclophosphamide failure in severe central nervous system sarcoidosis. Respir Med 103: 268–273. Sollberger M, Fluri F, Baumann S et al. (2004). Successful treatment of steroid-refractory neurosarcoidosis with infliximab. J Neurol 251: 760–761.

NEUROLOGIC MANIFESTATIONS OF SARCOIDOSIS Soucek D, Prior C, Loef G (1993). Successful treatment of spinal sarcoidosis by high-dose intravenous methylprednisolone. Clin Neuropharmacol 16: 464–467. Steffen M, Pettersen J, Oldigs A et al. (1993). Increased secretion of tumor necrosis factor-alpha, interleukin-1-beta, and interleukin-6 by alveolar macrophages from patients with sarcoidosis. J Allergy Clin Immunol 91: 939–948. Stern BJ (1992). Neurosarcoidosis. Neurol Chron 2: 1–6. Stern BJ, Corbett J (2007). Neurophthalmologic manifestations of sarcoidosis. Curr Treat Options Neurol 9: 63–71. Stern BJ, Krumholz A, Johns C et al. (1985). Sarcoidosis and its neurological manifestations. Arch Neurol 42: 909–917. Stern BJ, Schonfeld SA, Sewell C et al. (1992). The treatment of neurosarcoidosis with cyclosporine. Arch Neurol 49: 1065–1072. Stern BJ, Aksamit A, Clifford D et al. (2010). Neurologic presentations of sarcoidosis. Neurol Clin 28: 185–198. Stjernberg N, Cajander S, Truedsson H (1981). Muscle involvement in sarcoidosis. Acta Med Scand 209: 213–216. Streletz LJ, Chambers LJ, Bae SH (1981). Visual evoked potentials in sarcoidosis. Neurology 31: 1545–1549. Stuart CA, Neelon FA, Lebovitz HD (1978). Hypothalamic insufficiency: the cause of hypopituitarism in sarcoidosis. Ann Intern Med 88: 589–594. Sverrild A, Nacker V, Kyvikko et al. (2008). Heredity in sarcoidosis: Aregistry-based twin study. Thorax 63: 894–896. Tamura N, Moller DR, Balbi B et al. (1991). Differential usage of the T-cell antigen receptor-chain constant region C1 element by lung T-lymphocytes of patients with pulmonary sarcoidosis. Am Rev Respir Dis 143: 635–639. Tanoue LT, Zitnik R, Elias JA (1994). Systemic sarcoidosis. In: GL Baum, E Wolinsky (Eds.), Textbook of Pulmonary Diseases. Vol. 1. Little, Brown and Co, Boston, pp. 745–774. Teirstein AS, Lesser M (1983). Worldwide distribution and epidemiology of sarcoidosis. In: BL Fanburg (Ed.), Sarcoidosis and other granulomatous diseases of the lung. Marcel Dekker, New York, pp. 101–134.

333

Terushkin V, Stern BJ, Judson MA et al. (2010). Neurosarcoidosis: presentations and management. Neurologist 16: 2–15. Thomas PD, Hunninghake GW (1987). Current concepts of the pathogenesis of sarcoidosis. Am Rev Respir Dis 135: 747–760. Thompson C, Checkley S (1981). Short term memory deficit in a patient with cerebral sarcoidosis. Br J Psychiatry 139: 160–161. Trombley IK, Mirra SS, Miles ML (1981). An electron microscopic study of central nervous system sarcoidosis. Ultrastruct Pathol 2: 257–267. Turner-Warwick M, Haslam PL, McAllister W (1986). Do measurements of bronchoalveolar lymphocytes and neutrophils, serum angiotensin-converting enzyme and gallium uptake help the clinician to treat patients with sarcoidosis? Ann N Y Acad Sci 465: 387–394. Urich H (1977). Neurosarcoidosis or granulomatous angiitis: a problem of definition. Mt Sinai J Med 44: 718–725. Waxman JS, Sher JH (1979). The spectrum of central nervous system sarcoidosis: a clinical and pathologic study. Mt Sinai J Med 46: 309–317. Weinberg S, Bennett H, Weinstock I (1983). CNS manifestations of sarcoidosis in children. Clin Pediatr 22: 447–481. Weissler JC (1994). Southwestern internal medicine conference: sarcoidosis: immunology and clinical management. Am J Med Sci 307: 233–245. Wiederholt WC, Siekert RG (1965). Neurological manifestations of sarcoidosis. Neurology 15: 1147–1154. Younger DS, Hays AP, Brust JC et al. (1988). Granulomatous angiitis of the brain. An inflammatory reaction of diverse etiology. Arch Neurol 45: 514–518. Zajicek JP, Scolding NJ, Foster O et al. (1999). Central nervous system sarcoidosis: diagnosis and management. QJM 92: 103–117. Zuniga G, Ropper AH, Frank J (1991). Sarcoid peripheral neuropathy. Neurology 41: 1558–1561.

Neurologic manifestations of achondroplasia.

Neurologic manifestations in sarcoidosis: review of the literature, with a report of 23 cases.

Sarcoidosis: multiple chest manifestations.

Neurologic manifestations of malabsorption syndromes.

Cardiovascular manifestations of neurologic disease.

Neurologic manifestations of hypothalamic disease.

Porphyria and its neurologic manifestations.

Sarcoidosis and its ophthalmic manifestations.

Sarcoidosis: abdominal manifestations at CT.

Neurologic manifestations of Henoch-Schönlein purpura.

Neurologic manifestations of progressive systemic sclerosis.

Chronic neurologic manifestations of Lyme disease.

Neurologic manifestations of giant cell arteritis.

Neurologic manifestations of gastrointestinal and liver diseases.

Neurologic manifestations of acute liver failure.

Neurologic manifestations of childhood rheumatic diseases.

Neurologic manifestations of the neglected tropical diseases.

Neurologic Manifestations of Inflammatory Bowel Disease.

Chronic neurologic manifestations of Lyme disease.

Neurologic manifestations of diphtheria and pertussis.

Neurologic manifestations of inflammatory bowel diseases.

Sarcoidosis: overview of pulmonary manifestations and imaging.

Lyme disease: neurologic and ophthalmic manifestations.

Neurologic manifestations of inherited disorders of connective tissue.