- Email: [email protected]
Neuralgic amyotrophy. An update
Joint Bone Spine 84 (2017) 153–158
Available online at
ScienceDirect www.sciencedirect.com
Review
Neuralgic amyotrophy. An update Paul Seror Laboratoire d’électroneuromyographie, 146, avenue Ledru-Rollin, 75011 Paris, France
a r t i c l e
i n f o
Article history: Accepted 10 March 2016 Available online 2 June 2016 Keywords: Neuralgic amyotrophy Plexopathy Parsonage and Turner syndrome
a b s t r a c t A century after the first description of neuralgic amyotrophy (NA), its pathophysiology remains unknown. An inflammatory (auto)immune pathophysiology is presumed, with mechanical or infectious precipitating conditions, which triggers attacks. Clinically, NA is an acute and painful unique or multiple mononeuropathy that causes palsy, amyotrophy and sensory loss in an asymmetric and patchy distribution. It involves the upper brachial plexus rather than the other parts but also may involve the cervical plexus, lumbosacral plexus and cranial nerves. The impairment can be restricted to one fascicule of one nerve, plexus or root; limited to a few ones; or extensive, involving both upper limbs. Its evolution is usually monophasic and auto-limited and never leads to generalized polyneuropathy. Electrodiagnostically, NA is characterized by severe axonal damage. The recovery is usually good after 6 months to 3 years in 80% of cases. Persistent disability is present in 20% of idiopathic NA cases and is more frequent in hereditary NA, with frequent recurrences, more frequent bilateral impairment, and more atypical distribution (cervical plexus, lumbosacral plexus or cranial nerves) than with idiopathic NA. Hereditary NA is mainly linked to a mutation in the gene of the Septin-9 protein. When the patient is seen early after disease onset, treatment with corticosteroids for 2 weeks seems to shorten the pain duration and the delayed recovery. With diagnosis during the palsy period, treatment is based on pharmacologic and non-pharmacologic therapies according to the complaints of the patient. ´ e´ franc¸aise de rhumatologie. Published by Elsevier Masson SAS. All rights reserved. © 2016 Societ
1. Introduction Neuralgic amyotrophy (NA) [1], also called Parsonage–Turner syndrome [1,2] or brachial plexus neuritis [3], is an acute and painful neuropathy that involves mainly the upper brachial plexus. It was first reported after anti-tetanic serotherapy, by Dyke, in 1918. Wyburn-Mason [3] described 42 cases of “brachial plexus neuritis” in 1941, and Spillane [4] reported 46 cases of “localised neuritis of the shoulder girdle” in 1943. Then in 1948, Parsonage and Turner [1] reported 136 cases of “neuralgic amyotrophy”. Of note, Dyke, Spillane, Parsonage and Turner were all military medical doctors, which explains the observation of frequent precipitating conditions with the first descriptions. “Neuralgic amyotrophy”, first used by Parsonage and Turner, remains the best term to describe this pathology because it is purely descriptive and presumes none of the pathophysiology nor the exact level of the nerve lesion. It immediately provides the main description of a painful disorder related to a nerve lesion that is severe enough to cause amyotrophy. Other terminologies focus on the region impaired (shoulder girdle) [1,4], the localisation of the nerve lesion (brachial plexus neuritis, plexopathy, neuritis, etc.)
E-mail address: [email protected]
[2,3,5,6], or the pathophysiology (immune, idiopathic, hereditary) [7,8]. Since then, some rare series were published that confirmed the previous data or provided some new insights considering diverse clinical, electrodiagnostic (EDX), biological, histological, genetic, imaging and therapeutic data. In 2000, a national expert center for NA was created in Radboud medical university, in The Netherlands. Since the last reviews considering NA in the French language in 1992 [9] and 2009 [10], the knowledge of NA has progressed. This article revisits this topic with new data from the literature during the last 2 decades.
2. Epidemiology of neuralgic amyotrophy NA is more frequent in men than women and in the right than left upper limb (both ratios 2 to 1); the disease is asymmetric in 97% of cases. It involves patients from 3 to 80 years old [8]. NA is usually considered a rare disorder (annual incidence 2/100,000 people) [11,12], but recently, an annual incidence of 1/1000 people was reported in primary care, following a specific training of general practitioners on how to diagnose NA (Box 1) [13]. This study confirmed that NA was poorly known and explains why it is diagnosed, in 3 of 4 cases, 28 weeks after disease onset (mean 44 weeks).
http://dx.doi.org/10.1016/j.jbspin.2016.03.005 ´ e´ franc¸aise de rhumatologie. Published by Elsevier Masson SAS. All rights reserved. 1297-319X/© 2016 Societ
154
P. Seror / Joint Bone Spine 84 (2017) 153–158
Box 1: Suggestive symptoms and exclusion criteria for considering a diagnosis of neuralgic amyotrophy [13]
Box 2: Non-exhaustive list of the precipitating conditions in the literature
• Neuralgic amyotrophy is probable or definite with: ◦ new onset shoulder pain (uni- or bilateral); ◦ numeric rating scale score for pain of 7 on a scale of 0–10; ◦ abnormal shoulder movement (glenohumeral and/or scapulothoracic) during maximum abduction/anteflexion movement; ◦ when first seen 3 weeks after onset: paresis of long thoracic nerve, suprascapular nerve, anterior interosseus nerve. • Optional signs and symptoms: ◦ less severe initial pain with otherwise typical clinical multifocal distribution of weakness and monophasic course; ◦ more extensive multifocal paresis of upper extremity(-ies); ◦ asymmetric involvement of other upper extremity; ◦ areas of hypesthesia and/or paresthesia in the upper extremity; ◦ involvement of other peripheral nerves: lumbosacral plexus, phrenic, recurrent laryngeal nerve. • Neuralgic amyotrophy is excluded with: ◦ progression of pain and/or weakness > 3 months (except for pain associated with abnormal compensatory shoulder movements); ◦ only passive range of motion constraints in the glenohumeral joint; ◦ Horner syndrome; ◦ perfectly symmetric weakness distribution; ◦ diabetes mellitus.
• Trauma (benign to gunshot wound), exercise, psychological stress, lumbar puncture, cold exposure, burns. • All kinds of surgery: tonsillectomy, vegetation, appendectomy, hysterectomy, coronary bypass surgery, knee surgery. • Pregnancy, childbirth and post-partum period. • Anti-tetanic serotherapy. • Vaccination: typhoid, diphtheria, tetanus, smallpox, flu, human papillomavirus, etc. • Bacterial infection: upper respiratory tract infection, typhus, typhoid, malaria, diphtheria, pneumonia, malaria, rheumatic fever, dysentery, sepsis, rickettsia coroni, bartonella henselae (cat claw), borrelia burgdorferi? • Viral infection: influenza; cytomegalovirus; herpes virus; varicella-zoster virus; parvovirus B19; Epstein-Barr virus; cocksackie virus A2, B; hepatitis B and E virus (10% vs. 0.9% NA among blood donors); Echo 13/30 virus; smallpox; poliomyelitis.
3. The clinical picture 3.1. Classical clinical pattern The classical clinical pattern of NA includes 3 successive phases: painful phase, then weakness, amyotrophy and sensory complaints, then recovery [1–4,6,8,9,11,14–18]. These 3 phases are encountered in a similar way with idiopathic NA (INA) and hereditary NA (HNA). 3.1.1. The painful phase Pain is the first symptom in 90% of cases; it has an acute onset, during the night in 61% of the cases, and is severe, relentless and neuropathic. Usually, pain on a numeric rating scale is greater than 7/10, and involves the shoulder girdle. Its duration varies from 1 day to 2 months; and is longer for males than females (mean 45 vs. 23 days); only 5% of patients have pain for < 24 h and only 10% for > 2 months. After the acute painful phase, 30% to 70% of patients have persistent pain related to diverse causes. Some have neuropathic pain, others develop pain related to muscular compensation required with weakness of palsied muscles, and others have chronic pain without clear significance. 3.1.2. Phase of weakness, amyotrophy and sensory complaints Weakness may precede the pain in 5% of cases but occurs within 24 h in 34% of cases, after 1–7 days in 39% and after 1–4 weeks in 27%. The amyotrophy appears usually between 2 and 6 weeks and reflects the importance of the axonal loss. It is obvious for superficial muscles (deltoid, supra- and infraspinatus) or is completely invisible for other deep muscles (serratus anterior and pronator quadratus muscles). In descending order of frequency, the most commonly affected muscles are the infraspinatus (72%), serratus anterior (70%), supraspinatus (65%), biceps (61%), rhomboid and pronator teres (53%). The other muscles of the upper limbs are affected in less than 50% of cases, and muscles of the neck (trapezius 20% and sternocleidomastoid 7%) are still less frequently
affected [8]. The distribution of the affected muscles corresponds to a root (C5, C6 or C7 more frequently than C8T1), a part of the brachial plexus (the upper and middle plexus more frequently than the lower plexus), a nerve trunk (suprascapular and long thoracic nerves, etc.) or some fascicules of a nerve (anterior interosseous nerve, a motor branch of the median nerve). Lasegue manoeuvre of the upper limb is frequently positive, and frozen shoulder is reported in 17% of cases as a consequence of the NA [8]. Sensory complaints are present in 66% [6] to 78% [8] of cases, were described in the first reports [1,3,4], and appear with the motor weakness and are an integral part of the classical pattern. Their functional consequences are fewer than those of motor weakness and wasting, consequently, NA has been frequently considered by practitioners as a purely a motor pathology. They usually involve the upper cervical root or brachial plexus but may involve some nerve trunks such as the lateral and medial ante-brachial cutaneous nerves, and exceptionally, the distal median or ulnar nerves [16,19,20]. 3.1.3. Recovery phase Overall, 75% of INA cases show good to complete recovery between 6 months and 3 years [2,6,11,14]. In HNA, the multiple attacks of NA (up to 74% of recurrence) increases the risk of incomplete recovery and frequently leads to progressive weakness and disability of one or both upper limbs (20% to 30%) [8]. When the nerve lesion is partial, the recovery is provided by collateral reinnervation and may be completed in 6 to 12 months. In contrast, when the lesion is severe, recovery is based on nerve growth and direct reinnervation, requires 1 to 3 years, and is not always as good. On the whole, the quality and the delay in recovery depend closely on the severity of the axonal loss and the length of nerve growth. 3.2. Precipitating conditions Precipitating conditions were reported from the very first publication of NA: anti-tetanic serotherapy [15]. Since then, numerous precipitating conditions have been reported (Box 2), in large series and in case reports. Such conditions are common benign trauma, simple or strenuous exercise, pregnancy, any kind of surgery, numerous bacterial or viral infections, and also diverse vaccinations. These conditions are found in 45% to 53% of cases [1,6,8], and occur a few hours to 1 month before the NA attack (within one week 52%) [8,21].
P. Seror / Joint Bone Spine 84 (2017) 153–158
Box 3: A dysmorphic syndrome is found in some patients with hereditary NA (HNA) In considering the facial features of HNA, Dunn et al. claimed that these patients resemble a Modigliani painting [25]. The syndrome combines: • • • • • • • • •
hypotelorism; blepharophimosis; slight ptosis; epicanthal folds; microstomia; dysmorphic ears; cleft uvula; cleft palate; partial syndactyly of fingers or toes.
3.3. Atypical cases or symptoms NA may be atypical, when it is hereditary, non-painful, distal, extensive, extra-brachial, recurrent, purely sensitive or autonomic. 3.3.1. HNA HNA was first described by Taylor, in 1960 [22], and Dunn et al., in 1978 [23], and was mainly investigated by Van Alfen et al. in a national expert center dedicated to NA, that compiled 36 of the 300 HNA families known worldwide. [8,24]. HNA represented almost 20% of the cases reported by Van Alfen et al. but certainly less than 10% in a more common medical practice [13]. Patients with HNA are characterized by other cases in the family, a first attack in childhood (42%), more frequent recurrence (74%), more frequent extra-brachial attack (56%), longer duration of pain, and more frequent disability due to multiple attacks. HNA is related to a mutation of the Septin-9 protein (SEPT9), on chromosome 17q25, with some repeated motifs that altered bind and bundle microtubules [25]. This autonomic-dominant form [26–28] represents 55% to 85% of the cases in The Netherlands and the United States, respectively. HNA patients may have a dysmorphic syndrome (Box 3) [9,23,26,28]. 3.3.2. Non-painful NA Less than 10% of cases show no pain or little pain at onset of the disease [8], and NA should not be ruled out if other positive signs are present (Box 1). In fact, Parsonage and Turner reported 2 of 136 cases with no pain [1]. 3.3.3. Distal NA Patients with motor and/or sensory symptoms distal to the elbow joint are considered to have distal NA. England and Sumner in 1987 described 9 cases and considered NA as “an increasingly diverse entity” [16]; later, 4 cases were reported with distal sensory and mild motor impairment [29]. However, many cases of distal NA involving the anterior interosseous nerve (AIN) and posterior interosseous nerve (PIN) or distal median and ulnar nerves were reported with the first descriptions and confirmed by the large series of Van Alfen et al. [8]. The AIN syndrome was frequently described as “isolated neuritis” [5,30,31]. Currently, the syndrome is largely considered part of NA and was more clinically and electrodiagnostically investigated by Seror [32], who pointed out the frequency of partial AIN palsies (43%) [33]. 3.3.4. Extensive NA In some rare cases, NA may involve many roots, division cords or nerves of both brachial plexus, leading to a severe, extensive asymmetric palsy of both upper limbs. The disease affects proximal and distal muscles, is very painful, and usually recovers
155
incompletely. The diagnosis requires, more than any other, all additional investigations, NA always being an elimination diagnosis. Some of these cases are associated with abnormally elevated liver enzymes, phrenic nerve palsy [8,34] and hepatitis E [35]. 3.3.5. NA with extra-brachial nerve lesions In some cases, NA may involve nerves not in the brachial plexus. Although some were described by Parsonage and Turner (9 involved trapezius muscles: 11th cranial nerve to C4 roots), extra-brachial forms are particularly unusual, and some have been slow to emerge as NA, particularly those affecting the lumbosacral plexus [36–39], and the phrenic nerve [17,34]. The most frequently involved extra-brachial nerve is the spinal accessory nerve [40,41] and the most exceptional nerves are the recurrent nerve and the cranial nerves [8,14,42]. These lesions, which are rare in INA (17%), are quite frequent in HNA (56%) [8]. All extra-brachial NA may be an inaugural presentation, are always an elimination diagnosis, and require additional examinations, except perhaps for some unilateral spinal accessory nerve lesion [40]. 3.3.6. Recurrent NA The rate of recurrence varies from 5% [6] to 26% [8] in INA patients as compared with 74% in HNA patients. In one HNA patient, the number of attacks varies from ×2 (19%) to ×5 (23%) and up to ×20! 3.3.7. Sensory NA Pure sensory cases are exceptionally reported [19,43] but are certainly underdiagnosed. In NA, sensory complaints are underestimated, and for practitioners not familiar with NA, the presence of persistent sensory complaints rules out the NA diagnosis. Altered sensory nerve action potentials of sensory-motor nerves or purely sensory nerves such as radial, median, ulnar, lateral and medial ante-brachial cutaneous nerves are rarely documented and reported [16,19,20,43]. 3.3.8. Autonomic symptoms Vasomotor symptoms (oedema, temperature dysregulation, and increased sweating) are rarely reported [8], and are more frequent (15% to 32%) in patients with motor involvement of the posterior cord and lower part of the brachial plexus. When autonomic dysfunction is present, sensory symptoms are also present in 94% of cases. 4. Additional investigations No additional investigations can confirm the diagnosis of NA. Its diagnosis is essentially clinical with a typical pattern; with an atypical pattern, NA diagnosis is an elimination diagnosis. Some investigations are frequently used to confirm the peripheral nature of the motor and sensory deficit and others are used to rule out other diagnoses such as proximal entrapment neuropathies of the upper limb, compressive cervical root disorder, rotator cuff syndrome, meningo-radiculitis, facio-scapulo-humeral myopathy, infiltrative or post-radiation plexopathy, and vasculitis. 4.1. Electrodiagnosis (EDX) EDX is the first examination required when a NA is suspected. It demonstrates that the palsy is of peripheral origin, due to an important to severe axonal damage [16,18]. The distribution is patchy, according to a root, division, cord, branch of the plexus, or nerve, separately or in combination, and has very unequal severity from one nerve to another. The needle examination provides the most characteristic features for NA diagnosis, when it is performed at least 10 to 15 days after onset. Then, it reveals
156
P. Seror / Joint Bone Spine 84 (2017) 153–158
an obvious neurogenic pattern: single unit pattern to only simplified interference pattern with high firing rate of motor units (according to the severity of the axonal loss) and numerous denervation signs (2+ to 3+ fibrillations and positive sharp waves). The motor and sensory nerve conduction velocities remain normal, thereby eliminating demyelinating disorders such as chronic idiopathic demyelinating polyneuropathy and hereditary neuropathy with liability to pressure palsy. The motor nerve conduction study demonstrates low or dramatically decreased amplitude of compound motor action potential in palsied muscles. The sensory nerve conduction study may demonstrate normal or decreased amplitude of sensory nerve action potential (SNAP) according to the level of nerve involvement: normal for pre-ganglionic damage (root); decreased for post-ganglionic damage [44,45]. Because NA involves mainly the upper plexus and SNAP tests mainly the middle and lower plexus, the proportion of abnormal SNAP remains low, about 17% [16,19,20,29]. Exceptionally, conduction block aspects were observed [46].
Histological examination of peripheral nerves during the NA is never typically done. However, some authors have performed surgery for certain forms of NA that could be diagnosed as nerve entrapment of AIN, PIN, lateral ante-brachial cutaneous and other nerves [51–53]. The authors found, macroscopically, fascicular hourglass-like constrictions and microscopically, infiltrates of CD8positive T lymphocytes in all excised specimens, with thickening of the perineurium at the distal and proximal portions of the constriction [52]. Other authors, concerned about malignancy, found mononuclear inflammatory infiltrates (T lymphocytes and rarely CD20+ B lymphocytes), surrounding epineurial and endoneurial vessels, without signs of true necrotizing vasculitis [7]. Histological study of HNA patients showed that “the sensory nerves are definitely involved despite the mainly motor symptoms, and that the lesions in nerves and muscles are more widespread than clinically presumed” [24].
4.2. Laboratory findings
5. Pathophysiology of NA
No routine immunological test searching for specific antibodies (anti-myelin, anti-axon, or anti-ganglioside), proving cross-immune reaction, is available to assist in diagnosis of NA [8,47]. Blood investigations may be useful to define the recent viral infection (IgM antibodies), that may have triggered the NA attack, before some specific clinical signs (Box 2) [21]. Hepatitis E infection should be searched when NA is bilateral, with elevated serum transaminase levels, diaphragm palsy, or extensive NA [35]. Searching for mellitus diabetes and antibodies such as ANCAs, FAN, Latex and others according to the clinical pattern (lumbosacral plexus or extensive NA) is useful to rule out some auto-immune disorders. Lumbar puncture findings are usually normal in NA and the procedure is rarely performed but is necessary with extensive NA, with involvement of the lumbosacral plexus or the cranial nerves, or with chronic erythema migrans to search for Lyme disease and other causes of meningo-radiculitis. Genetic investigations of the SEPT9 protein will be performed when HNA is suspected. In some rare cases a deletion of the gene of the PMP22 can be searched, when the clinical and EDX pattern is evocative of hereditary neuropathy with liability to pressure palsy [8].
The pathophysiology of NA remains uncertain, although it often suggests an auto-immune origin such as in Guillain-Barré syndrome [7,35,47,54,55]. Indeed, NA often occurs after similar triggering such as a viral or bacterial infection, vaccination, surgery, or childbirth; these circumstances were noted by all authors since its first description. The other hypothesis is the possibility of a direct effect of the virus on the nerve, for viral radiculitis, mononeuritis, or plexitis. In any case, these two mechanisms, which are not mutually exclusive and can coexist, remain until now simple presumptions. Regardless, an intense, focal inflammation [7,52] of vessels of the nerve causes acute axonal damage, with very unequal severity from one nerve to another. A concept of unique or multiple focal inflammation immune-triggered, best defines the pathophysiology of NA. So the use of strenuous anti-inflammatory and immunomodulating drugs, at the acute phase of NA is actually discussed, to minimize the focal inflammation, and perhaps the fibrous, stenotic evolution of some fascicular hourglass-like constrictions and finally prevent the axonal damage.
4.3. Imaging studies Cervical spine imaging must be systematically performed when NA has a radicular distribution. Then, the pitfall is to find disc and osteophytic abnormalities in patients ≥ 45 years old and to consider that this imaging abnormality is responsible for the clinical involvement, actually related to the NA. Magnetic resonance imaging (MRI) of the brachial plexus may reveal abnormal T2 hypersignals of some trunk, division, or cord of the brachial plexus in 6% of NA cases [8,48] but their specificity is not demonstrated, as far as they may be frequently found in association with cervical root compression [49] and other causes of brachial plexopathy such as inflammatory, invasive, and postradiation plexus lesions. Recently, MRI performed in 20 cases of AIN syndrome found an abnormal T2 signal restricted to AIN fascicules inside the main trunk of the median nerve in all cases [30]. MRI may also show abnormal T2 hypersignal of palsied muscles [50], but they are usually discovered by chance, during an MRI of the shoulder for another reason, and thus may allow for rectifying the diagnosis, if the radiologist is aware of NA.
4.4. Histological examination
6. Treatment The treatment for NA is now fairly well codified and varies by the phase when the patient is seen, even if treatment strategies are still empirical and based on clinical series. Painful phase, if the patient is seen rapidly after onset of the attack (< 1 month), corticosteroids for 2 weeks is the treatment of choice as it seems to shorten the painful phase and provide more frequent recovery after 1 year (56% vs. 11%); the regimen of oral corticosteroids proposed by Van Eijk et al. was “60 mg/day (prednisolone) in the first week, tapered by 10 mg every day during the next 5 days, ending with 5 mg on day 13” [56]. The major limitation of this treatment is that patients are rarely seen so quickly by specialists who are familiar with NA; in fact NA is little known and is usually diagnosed late (cf. epidemiology) in common medical practice. In fact, only patients with NA recurrence have a “chance” to benefit from this treatment. In the most painful cases, opioids may be added, and in a few cases of extensive NA, intravenous corticosteroids and immunoglobulins have been used, with apparent efficacy on functional outcome [48]. With NA diagnosed later than 1 or 2 months after onset, pain is usually less, and symptomatic treatment combining analgesics and/or non-steroidal anti-inflammatory drugs are usually sufficient, but in the most severe cases, opioids, corticosteroids and intravenous immunoglobulins may still be discussed [57]. Later,
P. Seror / Joint Bone Spine 84 (2017) 153–158
during the palsy period, pharmacologic and non-pharmacologic treatments are used, according to the complaints of the patient, during recovery. Analgesics are not required in most cases, but with neuropathic pain, specific neuropathic pain-relief medications may be required. In some other cases, residual pain is related to muscular compensation due to the weakness of palsied muscles, then rehabilitation is the logical treatment. More exceptionally, because the weakness does not recover, a compensation orthotic or a surgical procedure may be considered. Surgery is quite never required, and Pan et al., analysing results of surgery for different nerve palsies with fascicular hourglass-like constriction, or stenosis, concluded that “the clinical presentation of patients with hourglass-like constrictions in their nerves is similar to that of patients with neuralgic amyotrophy, and the role of surgery in this condition is uncertain” [52]. Furthermore, “a neurolysed fascicular constriction may improve or disappear soon after neurolysis, or after deflation of the air tourniquet” [31], and then, has not the significance of definitive fibrous stenosis absolutely requiring surgery [52]. 7. The differential diagnosis There is no differential diagnosis when NA is responsible for a palsy of a right long thoracic nerve and a left suprascapular nerve in the same time. Otherwise, the main differential diagnoses are hyperalgesic cervical root disorder and shoulder joint disorders [8]. In cases of mononeuropathies, an entrapment neuropathy related to a fibrous arcade, synovial cyst, or lipoma may be considered. In other cases, according to the clinical and additional investigations, meningo-radiculitis, neoplastic plexopathy, or vasculitis, for example, may be considered. With no-pain “NA”, a chronic idiopathic demyelinating polyneuropathy, a multifocal motor neuropathy [58], a Lewis Sumner syndrome, and an hereditary neuropathy with liability to pressure palsy [59] or a facio-scapulo-humeral myopathy (unilateral accessory spinal palsy) may be considered. 8. Conclusions On the whole, in 2016, NA is still diagnosed on presumptions that combine evocative history, physical examination, and EDX examination (and without any identifiable cause on imaging). It corresponds neurophysiologically to an acute axonal damage that may involve alone or in combination a root, plexus or nerve lesion. It seems to best match an immune-triggered peripheral neuropathy that mostly involves the brachial plexus and more especially the long thoracic and the suprascapular nerves but may also involve the cervical and lumbosacral plexus. HNA is better known, and such cases more frequently exhibit a longer duration of pain, an attack in childhood, recurrent attacks, extra-brachial attacks, and a final disability. Hepatitis E should be searched when NA is bilateral, with modestly elevated serum transaminase levels, or phrenic nerve lesions or in extensive cases. When by chance the patient is seen early after the attack, treatment with corticosteroids must be promptly administered for 2 weeks. NA evolution is usually favourable and never progresses to a generalized neuropathy. Disclosure of interest The author declares that he has no competing interest. References [1] Parsonage MJ, Turner JW. Neuralgic amyotrophy. The shoulder girdle syndrome. Lancet 1948:973–8. [2] Turner JW, Parsonage MJ. Neuralgic amyotrophy (paralytic brachial neuritis); with special reference to prognosis. Lancet 1957;273:209–12.
157
[3] Wyburn-Mason. Brachial plexus neuritis occurring in epidemic form. Lancet 1941:662–4. [4] Spillane. Localised neuritis of the shoulder girdle. A report of 46 cases in the MEF. Lancet 1943:532–5. [5] Kiloh LG, Nevin S. Isolated neuritis of the anterior interosseous nerve. Br Med J 1952;1:850–1. [6] Tsairis P, Dyck PJ, Mulder DW. Natural history of brachial plexus neuropathy. Report on 99 patients. Arch Neurol 1972;27:109–17. [7] Suarez GA, Giannini C, Bosch EP, et al. Immune brachial plexus neuropathy: suggestive evidence for an inflammatory-immune pathogenesis. Neurology 1996;46:559–61. [8] van Alfen N, van Engelen BG. The clinical spectrum of neuralgic amyotrophy in 246 cases. Brain 2006;129:438–50. [9] Serratrice G, Baudoin D, Pouget J, et al. [Typical and atypical forms of neuralgic amyotrophy of the shoulder: 86 cases]. Rev Neurol 1992;148:47–50. [10] Seror P. Le syndrome de Parsonage et Turner. In: Brunon-Martinez A, Codine P, Herisson C, editors. Épaule et médecine de rééducation. Paris: Masson; 2009. p. 81–8. [11] Beghi E, Kurland LT, Mulder DW, et al. Brachial plexus neuropathy in the population of Rochester, Minnesota, 1970–1981. Ann Neurol 1985;18: 320–3. [12] MacDonald BK, Cockerell OC, Sander JW, et al. The incidence and lifetime prevalence of neurological disorders in a prospective community-based study in the UK. Brain 2000;123:665–76. [13] van Alfen N, van Eijk JJ, Ennik T, et al. Incidence of neuralgic amyotrophy (Parsonage Turner syndrome) in a primary care setting – a prospective cohort study. PloS One 2015;10:e0128361. [14] Cruz-Martinez A, Barrio M, Arpa J. Neuralgic amyotrophy: variable expression in 40 patients. J Peripher Nerv Syst 2002;7:198–204. [15] Dyke SC. Peripheral nerve lesions after anti-tetanic serum. Lancet 1918;1:570. [16] England JD, Sumner AJ. Neuralgic amyotrophy: an increasingly diverse entity. Muscle Nerve 1987;10:60–8. [17] Lahrmann H, Grisold W, Authier FJ, et al. Neuralgic amyotrophy with phrenic nerve involvement. Muscle Nerve 1999;22:437–42. [18] Subramony SH. AAEE case report #14: neuralgic amyotrophy (acute brachial neuropathy). Muscle Nerve 1988;11:39–44. [19] Seror P, Kuntz P, Maisonobe T, et al. Sensory nerve action potential abnormalities in neuralgic amyotrophy: a report of 18 cases. J Clin Neuromusc Dis 2002:445–9. [20] van Alfen N, Huisman WJ, Overeem S, et al. Sensory nerve conduction studies in neuralgic amyotrophy. Am J Phys Med Rehabil 2009;88:941–6. [21] Winer JB, Hughes RA, Anderson MJ, et al. A prospective study of acute idiopathic neuropathy. II. Antecedent events. J Neurol Neurosurg Psychiatry 1988;51:613–8. [22] Taylor RA. Heredofamilial mononeuritis multiplex with brachial predilection. Brain 1960;83:113–37. [23] Dunn HG, Daube JR, Gomez MR. Heredofamilial brachial plexus neuropathy (hereditary neuralgic amyotrophy with brachial predilection) in childhood. Dev Med Child Neurol 1978;20:28–46. [24] van Alfen N, Gabreels-Festen AA, Ter Laak HJ, et al. Histology of hereditary neuralgic amyotrophy. J Neurol Neurosurg Psychiatry 2005;76:445–7. [25] Bai X, Bowen JR, Knox TK, et al. Novel septin 9 repeat motifs altered in neuralgic amyotrophy bind and bundle microtubules. J Cell Biol 2013;203: 895–905. [26] Laccone F, Hannibal M, Neesen J, et al. Dysmorphic syndrome of hereditary neuralgic amyotrophy associated with a SEPT9 gene mutation – a family study. Clin Genet 2008;74:279–83. [27] Kuhlenbaumer G, Hannibal MC, Nelis E, et al. Mutations in SEPT9 cause hereditary neuralgic amyotrophy. Nat Genet 2005;37:1044–6. [28] van Alfen N, Hannibal MC, Chance PF, et al. In: Pagon RA, Adam MP, Ardinger HH, et al., editors. Hereditary neuralgic amyotrophy. Seattle (WA): University of Washington; 1993–2016. [Source: GeneReviews® (Internet). 2008 Feb 27. Updated 2012 Dec 6]. [29] Vanneste JA, Bronner IM, Laman DM, et al. Distal neuralgic amyotrophy. J Neurol 1999;246:399–402. [30] Pham M, Baumer P, Meinck HM, et al. Anterior interosseous nerve syndrome: fascicular motor lesions of median nerve trunk. Neurology 2014;82:598–606. [31] Ochi K, Horiuchi Y, Tazaki K, et al. Fascicular constrictions in patients with spontaneous palsy of the anterior interosseous nerve and the posterior interosseous nerve. J Plast Surg Hand Surg 2012;46:19–24. [32] Seror P. Electrodiagnostic examination of anterior interosseous nerve. Normal and pathological data (21 cases). Electromyogr Clin Neurophysiol 1999;39:183–9. [33] Seror P. [Anterior interosseous nerve lesions revisited in 2004]. Rev Med Liege 2004;59:58–66. [34] van de Ven AC, van Alfen N, Heijdra YF. [An unusual cause of acute dyspnoea: neuralgic amyotrophy]. Ned Tijdschr Geneeskd 2009;153:A181. [35] van Eijk JJ, Madden RG, van der Eijk AA, et al. Neuralgic amyotrophy and hepatitis E virus infection. Neurology 2014;82:498–503. [36] Bradley WG, Chad D, Verghese JP, et al. Painful lumbosacral plexopathy with elevated erythrocyte sedimentation rate: a treatable inflammatory syndrome. Ann Neurol 1984;15:457–64. [37] Evans BA, Stevens JC, Dyck PJ. Lumbosacral plexus neuropathy. Neurology 1981;31:1327–30. [38] Seror P, Maisonobe T, Viala K, et al. [Idiopathic lumbosacral plexopathy]. Presse Med 2005;34:856–8.
158
P. Seror / Joint Bone Spine 84 (2017) 153–158
[39] Dyck PJ, Windebank AJ. Diabetic and nondiabetic lumbosacral radiculoplexus neuropathies: new insights into pathophysiology and treatment. Muscle Nerve 2002;25:477–91. [40] Seror P. [Injury of the external branch of the accessory nerve]. Rev Neurol 2002;158:759–62. [41] Friedenberg SM, Zimprich T, Harper CM. The natural history of long thoracic and spinal accessory neuropathies. Muscle Nerve 2002;25:535–9. [42] Pierre PA, Laterre CE, Van den Bergh PY. Neuralgic amyotrophy with involvement of cranial nerves IX, X, XI and XII. Muscle Nerve 1990;13:704–7. [43] Seror P. Isolated sensory manifestations in neuralgic amyotrophy: report of eight cases. Muscle Nerve 2004;29:134–8. [44] Ferrante MA, Wilbourn AJ. The utility of various sensory nerve conduction responses in assessing brachial plexopathies. Muscle Nerve 1995;18:879–89. [45] Aminoff MJ, Olney RK, Parry GJ, et al. Relative utility of different electrophysiologic techniques in the evaluation of brachial plexopathies. Neurology 1988;38:546–50. [46] Watson BV, Nicolle MW, Brown JD. Conduction block in neuralgic amyotrophy. Muscle Nerve 2001;24:559–63. [47] Vriesendorp FJ, Dmytrenko GS, Dietrich T, et al. Anti-peripheral nerve myelin antibodies and terminal activation products of complement in serum of patients with acute brachial plexus neuropathy. Arch Neurol 1993;50:1301–3. [48] Johnson NE, Petraglia AL, Huang JH, et al. Rapid resolution of severe neuralgic amyotrophy after treatment with corticosteroids and intravenous immunoglobulin. Muscle Nerve 2011;44:304–5. [49] Yoshida T, Sueyoshi T, Suwazono S, et al. Three-tesla magnetic resonance neurography of the brachial plexus in cervical radiculopathy. Muscle Nerve 2015;52:392–6.
[50] Gaskin CM, Helms CA. Parsonage-Turner syndrome: MR imaging findings and clinical information of 27 patients. Radiology 2006;240:501–7. [51] Yamamoto S, Nagano A, Mikami Y, et al. Fascicular constriction in the anterior interosseous nerve and other motor branches of the median nerve. Muscle Nerve 1999;22:547–8. [52] Pan Y, Wang S, Zheng D, et al. Hourglass-like constrictions of peripheral nerve in the upper extremity: a clinical review and pathological study. Neurosurgery 2014;75:10–22. [53] Comtet JJ, Chambaud D. [“Spontaneous” paralysis of the posterior interosseous nerve by unusual lesions]. Rev Chir Orthop Reparatrice Appar Mot 1975;61:533–41. [54] van Eijk JJ, van Alfen N, Tio-Gillen AP, et al. Screening for antecedent Campylobacter jejuni infections and anti-ganglioside antibodies in idiopathic neuralgic amyotrophy. J Peripher Nerv Syst 2011;16:153–6. [55] Vighetto A, Grand C, Confavreux C, et al. [The Parsonage-Turner syndrome and similar diseases, 29 cases]. Rev Neurol 1988;144:494–8. [56] van Eijk JJ, van Alfen N, Berrevoets M, et al. Evaluation of prednisolone treatment in the acute phase of neuralgic amyotrophy: an observational study. J Neurol Neurosurg Psychiatry 2009;80:1120–4. [57] Naito KS, Fukushima K, Suzuki S, et al. Intravenous immunoglobulin (IVIg) with methylprednisolone pulse therapy for motor impairment of neuralgic amyotrophy: clinical observations in 10 cases. Intern Med 2012;51:1493–500. [58] Seror P, Leger JM, Maisonobe T. Anterior interosseous nerve and multifocal motor neuropathy. Muscle Nerve 2002;26:841–4. [59] Amato AA, Jackson CE, Kim JY, et al. Chronic relapsing brachial plexus neuropathy with persistent conduction block. Muscle Nerve 1997;20:1303–7.
Available online at
ScienceDirect www.sciencedirect.com
Review
Neuralgic amyotrophy. An update Paul Seror Laboratoire d’électroneuromyographie, 146, avenue Ledru-Rollin, 75011 Paris, France
a r t i c l e
i n f o
Article history: Accepted 10 March 2016 Available online 2 June 2016 Keywords: Neuralgic amyotrophy Plexopathy Parsonage and Turner syndrome
a b s t r a c t A century after the first description of neuralgic amyotrophy (NA), its pathophysiology remains unknown. An inflammatory (auto)immune pathophysiology is presumed, with mechanical or infectious precipitating conditions, which triggers attacks. Clinically, NA is an acute and painful unique or multiple mononeuropathy that causes palsy, amyotrophy and sensory loss in an asymmetric and patchy distribution. It involves the upper brachial plexus rather than the other parts but also may involve the cervical plexus, lumbosacral plexus and cranial nerves. The impairment can be restricted to one fascicule of one nerve, plexus or root; limited to a few ones; or extensive, involving both upper limbs. Its evolution is usually monophasic and auto-limited and never leads to generalized polyneuropathy. Electrodiagnostically, NA is characterized by severe axonal damage. The recovery is usually good after 6 months to 3 years in 80% of cases. Persistent disability is present in 20% of idiopathic NA cases and is more frequent in hereditary NA, with frequent recurrences, more frequent bilateral impairment, and more atypical distribution (cervical plexus, lumbosacral plexus or cranial nerves) than with idiopathic NA. Hereditary NA is mainly linked to a mutation in the gene of the Septin-9 protein. When the patient is seen early after disease onset, treatment with corticosteroids for 2 weeks seems to shorten the pain duration and the delayed recovery. With diagnosis during the palsy period, treatment is based on pharmacologic and non-pharmacologic therapies according to the complaints of the patient. ´ e´ franc¸aise de rhumatologie. Published by Elsevier Masson SAS. All rights reserved. © 2016 Societ
1. Introduction Neuralgic amyotrophy (NA) [1], also called Parsonage–Turner syndrome [1,2] or brachial plexus neuritis [3], is an acute and painful neuropathy that involves mainly the upper brachial plexus. It was first reported after anti-tetanic serotherapy, by Dyke, in 1918. Wyburn-Mason [3] described 42 cases of “brachial plexus neuritis” in 1941, and Spillane [4] reported 46 cases of “localised neuritis of the shoulder girdle” in 1943. Then in 1948, Parsonage and Turner [1] reported 136 cases of “neuralgic amyotrophy”. Of note, Dyke, Spillane, Parsonage and Turner were all military medical doctors, which explains the observation of frequent precipitating conditions with the first descriptions. “Neuralgic amyotrophy”, first used by Parsonage and Turner, remains the best term to describe this pathology because it is purely descriptive and presumes none of the pathophysiology nor the exact level of the nerve lesion. It immediately provides the main description of a painful disorder related to a nerve lesion that is severe enough to cause amyotrophy. Other terminologies focus on the region impaired (shoulder girdle) [1,4], the localisation of the nerve lesion (brachial plexus neuritis, plexopathy, neuritis, etc.)
E-mail address: [email protected]
[2,3,5,6], or the pathophysiology (immune, idiopathic, hereditary) [7,8]. Since then, some rare series were published that confirmed the previous data or provided some new insights considering diverse clinical, electrodiagnostic (EDX), biological, histological, genetic, imaging and therapeutic data. In 2000, a national expert center for NA was created in Radboud medical university, in The Netherlands. Since the last reviews considering NA in the French language in 1992 [9] and 2009 [10], the knowledge of NA has progressed. This article revisits this topic with new data from the literature during the last 2 decades.
2. Epidemiology of neuralgic amyotrophy NA is more frequent in men than women and in the right than left upper limb (both ratios 2 to 1); the disease is asymmetric in 97% of cases. It involves patients from 3 to 80 years old [8]. NA is usually considered a rare disorder (annual incidence 2/100,000 people) [11,12], but recently, an annual incidence of 1/1000 people was reported in primary care, following a specific training of general practitioners on how to diagnose NA (Box 1) [13]. This study confirmed that NA was poorly known and explains why it is diagnosed, in 3 of 4 cases, 28 weeks after disease onset (mean 44 weeks).
http://dx.doi.org/10.1016/j.jbspin.2016.03.005 ´ e´ franc¸aise de rhumatologie. Published by Elsevier Masson SAS. All rights reserved. 1297-319X/© 2016 Societ
154
P. Seror / Joint Bone Spine 84 (2017) 153–158
Box 1: Suggestive symptoms and exclusion criteria for considering a diagnosis of neuralgic amyotrophy [13]
Box 2: Non-exhaustive list of the precipitating conditions in the literature
• Neuralgic amyotrophy is probable or definite with: ◦ new onset shoulder pain (uni- or bilateral); ◦ numeric rating scale score for pain of 7 on a scale of 0–10; ◦ abnormal shoulder movement (glenohumeral and/or scapulothoracic) during maximum abduction/anteflexion movement; ◦ when first seen 3 weeks after onset: paresis of long thoracic nerve, suprascapular nerve, anterior interosseus nerve. • Optional signs and symptoms: ◦ less severe initial pain with otherwise typical clinical multifocal distribution of weakness and monophasic course; ◦ more extensive multifocal paresis of upper extremity(-ies); ◦ asymmetric involvement of other upper extremity; ◦ areas of hypesthesia and/or paresthesia in the upper extremity; ◦ involvement of other peripheral nerves: lumbosacral plexus, phrenic, recurrent laryngeal nerve. • Neuralgic amyotrophy is excluded with: ◦ progression of pain and/or weakness > 3 months (except for pain associated with abnormal compensatory shoulder movements); ◦ only passive range of motion constraints in the glenohumeral joint; ◦ Horner syndrome; ◦ perfectly symmetric weakness distribution; ◦ diabetes mellitus.
• Trauma (benign to gunshot wound), exercise, psychological stress, lumbar puncture, cold exposure, burns. • All kinds of surgery: tonsillectomy, vegetation, appendectomy, hysterectomy, coronary bypass surgery, knee surgery. • Pregnancy, childbirth and post-partum period. • Anti-tetanic serotherapy. • Vaccination: typhoid, diphtheria, tetanus, smallpox, flu, human papillomavirus, etc. • Bacterial infection: upper respiratory tract infection, typhus, typhoid, malaria, diphtheria, pneumonia, malaria, rheumatic fever, dysentery, sepsis, rickettsia coroni, bartonella henselae (cat claw), borrelia burgdorferi? • Viral infection: influenza; cytomegalovirus; herpes virus; varicella-zoster virus; parvovirus B19; Epstein-Barr virus; cocksackie virus A2, B; hepatitis B and E virus (10% vs. 0.9% NA among blood donors); Echo 13/30 virus; smallpox; poliomyelitis.
3. The clinical picture 3.1. Classical clinical pattern The classical clinical pattern of NA includes 3 successive phases: painful phase, then weakness, amyotrophy and sensory complaints, then recovery [1–4,6,8,9,11,14–18]. These 3 phases are encountered in a similar way with idiopathic NA (INA) and hereditary NA (HNA). 3.1.1. The painful phase Pain is the first symptom in 90% of cases; it has an acute onset, during the night in 61% of the cases, and is severe, relentless and neuropathic. Usually, pain on a numeric rating scale is greater than 7/10, and involves the shoulder girdle. Its duration varies from 1 day to 2 months; and is longer for males than females (mean 45 vs. 23 days); only 5% of patients have pain for < 24 h and only 10% for > 2 months. After the acute painful phase, 30% to 70% of patients have persistent pain related to diverse causes. Some have neuropathic pain, others develop pain related to muscular compensation required with weakness of palsied muscles, and others have chronic pain without clear significance. 3.1.2. Phase of weakness, amyotrophy and sensory complaints Weakness may precede the pain in 5% of cases but occurs within 24 h in 34% of cases, after 1–7 days in 39% and after 1–4 weeks in 27%. The amyotrophy appears usually between 2 and 6 weeks and reflects the importance of the axonal loss. It is obvious for superficial muscles (deltoid, supra- and infraspinatus) or is completely invisible for other deep muscles (serratus anterior and pronator quadratus muscles). In descending order of frequency, the most commonly affected muscles are the infraspinatus (72%), serratus anterior (70%), supraspinatus (65%), biceps (61%), rhomboid and pronator teres (53%). The other muscles of the upper limbs are affected in less than 50% of cases, and muscles of the neck (trapezius 20% and sternocleidomastoid 7%) are still less frequently
affected [8]. The distribution of the affected muscles corresponds to a root (C5, C6 or C7 more frequently than C8T1), a part of the brachial plexus (the upper and middle plexus more frequently than the lower plexus), a nerve trunk (suprascapular and long thoracic nerves, etc.) or some fascicules of a nerve (anterior interosseous nerve, a motor branch of the median nerve). Lasegue manoeuvre of the upper limb is frequently positive, and frozen shoulder is reported in 17% of cases as a consequence of the NA [8]. Sensory complaints are present in 66% [6] to 78% [8] of cases, were described in the first reports [1,3,4], and appear with the motor weakness and are an integral part of the classical pattern. Their functional consequences are fewer than those of motor weakness and wasting, consequently, NA has been frequently considered by practitioners as a purely a motor pathology. They usually involve the upper cervical root or brachial plexus but may involve some nerve trunks such as the lateral and medial ante-brachial cutaneous nerves, and exceptionally, the distal median or ulnar nerves [16,19,20]. 3.1.3. Recovery phase Overall, 75% of INA cases show good to complete recovery between 6 months and 3 years [2,6,11,14]. In HNA, the multiple attacks of NA (up to 74% of recurrence) increases the risk of incomplete recovery and frequently leads to progressive weakness and disability of one or both upper limbs (20% to 30%) [8]. When the nerve lesion is partial, the recovery is provided by collateral reinnervation and may be completed in 6 to 12 months. In contrast, when the lesion is severe, recovery is based on nerve growth and direct reinnervation, requires 1 to 3 years, and is not always as good. On the whole, the quality and the delay in recovery depend closely on the severity of the axonal loss and the length of nerve growth. 3.2. Precipitating conditions Precipitating conditions were reported from the very first publication of NA: anti-tetanic serotherapy [15]. Since then, numerous precipitating conditions have been reported (Box 2), in large series and in case reports. Such conditions are common benign trauma, simple or strenuous exercise, pregnancy, any kind of surgery, numerous bacterial or viral infections, and also diverse vaccinations. These conditions are found in 45% to 53% of cases [1,6,8], and occur a few hours to 1 month before the NA attack (within one week 52%) [8,21].
P. Seror / Joint Bone Spine 84 (2017) 153–158
Box 3: A dysmorphic syndrome is found in some patients with hereditary NA (HNA) In considering the facial features of HNA, Dunn et al. claimed that these patients resemble a Modigliani painting [25]. The syndrome combines: • • • • • • • • •
hypotelorism; blepharophimosis; slight ptosis; epicanthal folds; microstomia; dysmorphic ears; cleft uvula; cleft palate; partial syndactyly of fingers or toes.
3.3. Atypical cases or symptoms NA may be atypical, when it is hereditary, non-painful, distal, extensive, extra-brachial, recurrent, purely sensitive or autonomic. 3.3.1. HNA HNA was first described by Taylor, in 1960 [22], and Dunn et al., in 1978 [23], and was mainly investigated by Van Alfen et al. in a national expert center dedicated to NA, that compiled 36 of the 300 HNA families known worldwide. [8,24]. HNA represented almost 20% of the cases reported by Van Alfen et al. but certainly less than 10% in a more common medical practice [13]. Patients with HNA are characterized by other cases in the family, a first attack in childhood (42%), more frequent recurrence (74%), more frequent extra-brachial attack (56%), longer duration of pain, and more frequent disability due to multiple attacks. HNA is related to a mutation of the Septin-9 protein (SEPT9), on chromosome 17q25, with some repeated motifs that altered bind and bundle microtubules [25]. This autonomic-dominant form [26–28] represents 55% to 85% of the cases in The Netherlands and the United States, respectively. HNA patients may have a dysmorphic syndrome (Box 3) [9,23,26,28]. 3.3.2. Non-painful NA Less than 10% of cases show no pain or little pain at onset of the disease [8], and NA should not be ruled out if other positive signs are present (Box 1). In fact, Parsonage and Turner reported 2 of 136 cases with no pain [1]. 3.3.3. Distal NA Patients with motor and/or sensory symptoms distal to the elbow joint are considered to have distal NA. England and Sumner in 1987 described 9 cases and considered NA as “an increasingly diverse entity” [16]; later, 4 cases were reported with distal sensory and mild motor impairment [29]. However, many cases of distal NA involving the anterior interosseous nerve (AIN) and posterior interosseous nerve (PIN) or distal median and ulnar nerves were reported with the first descriptions and confirmed by the large series of Van Alfen et al. [8]. The AIN syndrome was frequently described as “isolated neuritis” [5,30,31]. Currently, the syndrome is largely considered part of NA and was more clinically and electrodiagnostically investigated by Seror [32], who pointed out the frequency of partial AIN palsies (43%) [33]. 3.3.4. Extensive NA In some rare cases, NA may involve many roots, division cords or nerves of both brachial plexus, leading to a severe, extensive asymmetric palsy of both upper limbs. The disease affects proximal and distal muscles, is very painful, and usually recovers
155
incompletely. The diagnosis requires, more than any other, all additional investigations, NA always being an elimination diagnosis. Some of these cases are associated with abnormally elevated liver enzymes, phrenic nerve palsy [8,34] and hepatitis E [35]. 3.3.5. NA with extra-brachial nerve lesions In some cases, NA may involve nerves not in the brachial plexus. Although some were described by Parsonage and Turner (9 involved trapezius muscles: 11th cranial nerve to C4 roots), extra-brachial forms are particularly unusual, and some have been slow to emerge as NA, particularly those affecting the lumbosacral plexus [36–39], and the phrenic nerve [17,34]. The most frequently involved extra-brachial nerve is the spinal accessory nerve [40,41] and the most exceptional nerves are the recurrent nerve and the cranial nerves [8,14,42]. These lesions, which are rare in INA (17%), are quite frequent in HNA (56%) [8]. All extra-brachial NA may be an inaugural presentation, are always an elimination diagnosis, and require additional examinations, except perhaps for some unilateral spinal accessory nerve lesion [40]. 3.3.6. Recurrent NA The rate of recurrence varies from 5% [6] to 26% [8] in INA patients as compared with 74% in HNA patients. In one HNA patient, the number of attacks varies from ×2 (19%) to ×5 (23%) and up to ×20! 3.3.7. Sensory NA Pure sensory cases are exceptionally reported [19,43] but are certainly underdiagnosed. In NA, sensory complaints are underestimated, and for practitioners not familiar with NA, the presence of persistent sensory complaints rules out the NA diagnosis. Altered sensory nerve action potentials of sensory-motor nerves or purely sensory nerves such as radial, median, ulnar, lateral and medial ante-brachial cutaneous nerves are rarely documented and reported [16,19,20,43]. 3.3.8. Autonomic symptoms Vasomotor symptoms (oedema, temperature dysregulation, and increased sweating) are rarely reported [8], and are more frequent (15% to 32%) in patients with motor involvement of the posterior cord and lower part of the brachial plexus. When autonomic dysfunction is present, sensory symptoms are also present in 94% of cases. 4. Additional investigations No additional investigations can confirm the diagnosis of NA. Its diagnosis is essentially clinical with a typical pattern; with an atypical pattern, NA diagnosis is an elimination diagnosis. Some investigations are frequently used to confirm the peripheral nature of the motor and sensory deficit and others are used to rule out other diagnoses such as proximal entrapment neuropathies of the upper limb, compressive cervical root disorder, rotator cuff syndrome, meningo-radiculitis, facio-scapulo-humeral myopathy, infiltrative or post-radiation plexopathy, and vasculitis. 4.1. Electrodiagnosis (EDX) EDX is the first examination required when a NA is suspected. It demonstrates that the palsy is of peripheral origin, due to an important to severe axonal damage [16,18]. The distribution is patchy, according to a root, division, cord, branch of the plexus, or nerve, separately or in combination, and has very unequal severity from one nerve to another. The needle examination provides the most characteristic features for NA diagnosis, when it is performed at least 10 to 15 days after onset. Then, it reveals
156
P. Seror / Joint Bone Spine 84 (2017) 153–158
an obvious neurogenic pattern: single unit pattern to only simplified interference pattern with high firing rate of motor units (according to the severity of the axonal loss) and numerous denervation signs (2+ to 3+ fibrillations and positive sharp waves). The motor and sensory nerve conduction velocities remain normal, thereby eliminating demyelinating disorders such as chronic idiopathic demyelinating polyneuropathy and hereditary neuropathy with liability to pressure palsy. The motor nerve conduction study demonstrates low or dramatically decreased amplitude of compound motor action potential in palsied muscles. The sensory nerve conduction study may demonstrate normal or decreased amplitude of sensory nerve action potential (SNAP) according to the level of nerve involvement: normal for pre-ganglionic damage (root); decreased for post-ganglionic damage [44,45]. Because NA involves mainly the upper plexus and SNAP tests mainly the middle and lower plexus, the proportion of abnormal SNAP remains low, about 17% [16,19,20,29]. Exceptionally, conduction block aspects were observed [46].
Histological examination of peripheral nerves during the NA is never typically done. However, some authors have performed surgery for certain forms of NA that could be diagnosed as nerve entrapment of AIN, PIN, lateral ante-brachial cutaneous and other nerves [51–53]. The authors found, macroscopically, fascicular hourglass-like constrictions and microscopically, infiltrates of CD8positive T lymphocytes in all excised specimens, with thickening of the perineurium at the distal and proximal portions of the constriction [52]. Other authors, concerned about malignancy, found mononuclear inflammatory infiltrates (T lymphocytes and rarely CD20+ B lymphocytes), surrounding epineurial and endoneurial vessels, without signs of true necrotizing vasculitis [7]. Histological study of HNA patients showed that “the sensory nerves are definitely involved despite the mainly motor symptoms, and that the lesions in nerves and muscles are more widespread than clinically presumed” [24].
4.2. Laboratory findings
5. Pathophysiology of NA
No routine immunological test searching for specific antibodies (anti-myelin, anti-axon, or anti-ganglioside), proving cross-immune reaction, is available to assist in diagnosis of NA [8,47]. Blood investigations may be useful to define the recent viral infection (IgM antibodies), that may have triggered the NA attack, before some specific clinical signs (Box 2) [21]. Hepatitis E infection should be searched when NA is bilateral, with elevated serum transaminase levels, diaphragm palsy, or extensive NA [35]. Searching for mellitus diabetes and antibodies such as ANCAs, FAN, Latex and others according to the clinical pattern (lumbosacral plexus or extensive NA) is useful to rule out some auto-immune disorders. Lumbar puncture findings are usually normal in NA and the procedure is rarely performed but is necessary with extensive NA, with involvement of the lumbosacral plexus or the cranial nerves, or with chronic erythema migrans to search for Lyme disease and other causes of meningo-radiculitis. Genetic investigations of the SEPT9 protein will be performed when HNA is suspected. In some rare cases a deletion of the gene of the PMP22 can be searched, when the clinical and EDX pattern is evocative of hereditary neuropathy with liability to pressure palsy [8].
The pathophysiology of NA remains uncertain, although it often suggests an auto-immune origin such as in Guillain-Barré syndrome [7,35,47,54,55]. Indeed, NA often occurs after similar triggering such as a viral or bacterial infection, vaccination, surgery, or childbirth; these circumstances were noted by all authors since its first description. The other hypothesis is the possibility of a direct effect of the virus on the nerve, for viral radiculitis, mononeuritis, or plexitis. In any case, these two mechanisms, which are not mutually exclusive and can coexist, remain until now simple presumptions. Regardless, an intense, focal inflammation [7,52] of vessels of the nerve causes acute axonal damage, with very unequal severity from one nerve to another. A concept of unique or multiple focal inflammation immune-triggered, best defines the pathophysiology of NA. So the use of strenuous anti-inflammatory and immunomodulating drugs, at the acute phase of NA is actually discussed, to minimize the focal inflammation, and perhaps the fibrous, stenotic evolution of some fascicular hourglass-like constrictions and finally prevent the axonal damage.
4.3. Imaging studies Cervical spine imaging must be systematically performed when NA has a radicular distribution. Then, the pitfall is to find disc and osteophytic abnormalities in patients ≥ 45 years old and to consider that this imaging abnormality is responsible for the clinical involvement, actually related to the NA. Magnetic resonance imaging (MRI) of the brachial plexus may reveal abnormal T2 hypersignals of some trunk, division, or cord of the brachial plexus in 6% of NA cases [8,48] but their specificity is not demonstrated, as far as they may be frequently found in association with cervical root compression [49] and other causes of brachial plexopathy such as inflammatory, invasive, and postradiation plexus lesions. Recently, MRI performed in 20 cases of AIN syndrome found an abnormal T2 signal restricted to AIN fascicules inside the main trunk of the median nerve in all cases [30]. MRI may also show abnormal T2 hypersignal of palsied muscles [50], but they are usually discovered by chance, during an MRI of the shoulder for another reason, and thus may allow for rectifying the diagnosis, if the radiologist is aware of NA.
4.4. Histological examination
6. Treatment The treatment for NA is now fairly well codified and varies by the phase when the patient is seen, even if treatment strategies are still empirical and based on clinical series. Painful phase, if the patient is seen rapidly after onset of the attack (< 1 month), corticosteroids for 2 weeks is the treatment of choice as it seems to shorten the painful phase and provide more frequent recovery after 1 year (56% vs. 11%); the regimen of oral corticosteroids proposed by Van Eijk et al. was “60 mg/day (prednisolone) in the first week, tapered by 10 mg every day during the next 5 days, ending with 5 mg on day 13” [56]. The major limitation of this treatment is that patients are rarely seen so quickly by specialists who are familiar with NA; in fact NA is little known and is usually diagnosed late (cf. epidemiology) in common medical practice. In fact, only patients with NA recurrence have a “chance” to benefit from this treatment. In the most painful cases, opioids may be added, and in a few cases of extensive NA, intravenous corticosteroids and immunoglobulins have been used, with apparent efficacy on functional outcome [48]. With NA diagnosed later than 1 or 2 months after onset, pain is usually less, and symptomatic treatment combining analgesics and/or non-steroidal anti-inflammatory drugs are usually sufficient, but in the most severe cases, opioids, corticosteroids and intravenous immunoglobulins may still be discussed [57]. Later,
P. Seror / Joint Bone Spine 84 (2017) 153–158
during the palsy period, pharmacologic and non-pharmacologic treatments are used, according to the complaints of the patient, during recovery. Analgesics are not required in most cases, but with neuropathic pain, specific neuropathic pain-relief medications may be required. In some other cases, residual pain is related to muscular compensation due to the weakness of palsied muscles, then rehabilitation is the logical treatment. More exceptionally, because the weakness does not recover, a compensation orthotic or a surgical procedure may be considered. Surgery is quite never required, and Pan et al., analysing results of surgery for different nerve palsies with fascicular hourglass-like constriction, or stenosis, concluded that “the clinical presentation of patients with hourglass-like constrictions in their nerves is similar to that of patients with neuralgic amyotrophy, and the role of surgery in this condition is uncertain” [52]. Furthermore, “a neurolysed fascicular constriction may improve or disappear soon after neurolysis, or after deflation of the air tourniquet” [31], and then, has not the significance of definitive fibrous stenosis absolutely requiring surgery [52]. 7. The differential diagnosis There is no differential diagnosis when NA is responsible for a palsy of a right long thoracic nerve and a left suprascapular nerve in the same time. Otherwise, the main differential diagnoses are hyperalgesic cervical root disorder and shoulder joint disorders [8]. In cases of mononeuropathies, an entrapment neuropathy related to a fibrous arcade, synovial cyst, or lipoma may be considered. In other cases, according to the clinical and additional investigations, meningo-radiculitis, neoplastic plexopathy, or vasculitis, for example, may be considered. With no-pain “NA”, a chronic idiopathic demyelinating polyneuropathy, a multifocal motor neuropathy [58], a Lewis Sumner syndrome, and an hereditary neuropathy with liability to pressure palsy [59] or a facio-scapulo-humeral myopathy (unilateral accessory spinal palsy) may be considered. 8. Conclusions On the whole, in 2016, NA is still diagnosed on presumptions that combine evocative history, physical examination, and EDX examination (and without any identifiable cause on imaging). It corresponds neurophysiologically to an acute axonal damage that may involve alone or in combination a root, plexus or nerve lesion. It seems to best match an immune-triggered peripheral neuropathy that mostly involves the brachial plexus and more especially the long thoracic and the suprascapular nerves but may also involve the cervical and lumbosacral plexus. HNA is better known, and such cases more frequently exhibit a longer duration of pain, an attack in childhood, recurrent attacks, extra-brachial attacks, and a final disability. Hepatitis E should be searched when NA is bilateral, with modestly elevated serum transaminase levels, or phrenic nerve lesions or in extensive cases. When by chance the patient is seen early after the attack, treatment with corticosteroids must be promptly administered for 2 weeks. NA evolution is usually favourable and never progresses to a generalized neuropathy. Disclosure of interest The author declares that he has no competing interest. References [1] Parsonage MJ, Turner JW. Neuralgic amyotrophy. The shoulder girdle syndrome. Lancet 1948:973–8. [2] Turner JW, Parsonage MJ. Neuralgic amyotrophy (paralytic brachial neuritis); with special reference to prognosis. Lancet 1957;273:209–12.
157
[3] Wyburn-Mason. Brachial plexus neuritis occurring in epidemic form. Lancet 1941:662–4. [4] Spillane. Localised neuritis of the shoulder girdle. A report of 46 cases in the MEF. Lancet 1943:532–5. [5] Kiloh LG, Nevin S. Isolated neuritis of the anterior interosseous nerve. Br Med J 1952;1:850–1. [6] Tsairis P, Dyck PJ, Mulder DW. Natural history of brachial plexus neuropathy. Report on 99 patients. Arch Neurol 1972;27:109–17. [7] Suarez GA, Giannini C, Bosch EP, et al. Immune brachial plexus neuropathy: suggestive evidence for an inflammatory-immune pathogenesis. Neurology 1996;46:559–61. [8] van Alfen N, van Engelen BG. The clinical spectrum of neuralgic amyotrophy in 246 cases. Brain 2006;129:438–50. [9] Serratrice G, Baudoin D, Pouget J, et al. [Typical and atypical forms of neuralgic amyotrophy of the shoulder: 86 cases]. Rev Neurol 1992;148:47–50. [10] Seror P. Le syndrome de Parsonage et Turner. In: Brunon-Martinez A, Codine P, Herisson C, editors. Épaule et médecine de rééducation. Paris: Masson; 2009. p. 81–8. [11] Beghi E, Kurland LT, Mulder DW, et al. Brachial plexus neuropathy in the population of Rochester, Minnesota, 1970–1981. Ann Neurol 1985;18: 320–3. [12] MacDonald BK, Cockerell OC, Sander JW, et al. The incidence and lifetime prevalence of neurological disorders in a prospective community-based study in the UK. Brain 2000;123:665–76. [13] van Alfen N, van Eijk JJ, Ennik T, et al. Incidence of neuralgic amyotrophy (Parsonage Turner syndrome) in a primary care setting – a prospective cohort study. PloS One 2015;10:e0128361. [14] Cruz-Martinez A, Barrio M, Arpa J. Neuralgic amyotrophy: variable expression in 40 patients. J Peripher Nerv Syst 2002;7:198–204. [15] Dyke SC. Peripheral nerve lesions after anti-tetanic serum. Lancet 1918;1:570. [16] England JD, Sumner AJ. Neuralgic amyotrophy: an increasingly diverse entity. Muscle Nerve 1987;10:60–8. [17] Lahrmann H, Grisold W, Authier FJ, et al. Neuralgic amyotrophy with phrenic nerve involvement. Muscle Nerve 1999;22:437–42. [18] Subramony SH. AAEE case report #14: neuralgic amyotrophy (acute brachial neuropathy). Muscle Nerve 1988;11:39–44. [19] Seror P, Kuntz P, Maisonobe T, et al. Sensory nerve action potential abnormalities in neuralgic amyotrophy: a report of 18 cases. J Clin Neuromusc Dis 2002:445–9. [20] van Alfen N, Huisman WJ, Overeem S, et al. Sensory nerve conduction studies in neuralgic amyotrophy. Am J Phys Med Rehabil 2009;88:941–6. [21] Winer JB, Hughes RA, Anderson MJ, et al. A prospective study of acute idiopathic neuropathy. II. Antecedent events. J Neurol Neurosurg Psychiatry 1988;51:613–8. [22] Taylor RA. Heredofamilial mononeuritis multiplex with brachial predilection. Brain 1960;83:113–37. [23] Dunn HG, Daube JR, Gomez MR. Heredofamilial brachial plexus neuropathy (hereditary neuralgic amyotrophy with brachial predilection) in childhood. Dev Med Child Neurol 1978;20:28–46. [24] van Alfen N, Gabreels-Festen AA, Ter Laak HJ, et al. Histology of hereditary neuralgic amyotrophy. J Neurol Neurosurg Psychiatry 2005;76:445–7. [25] Bai X, Bowen JR, Knox TK, et al. Novel septin 9 repeat motifs altered in neuralgic amyotrophy bind and bundle microtubules. J Cell Biol 2013;203: 895–905. [26] Laccone F, Hannibal M, Neesen J, et al. Dysmorphic syndrome of hereditary neuralgic amyotrophy associated with a SEPT9 gene mutation – a family study. Clin Genet 2008;74:279–83. [27] Kuhlenbaumer G, Hannibal MC, Nelis E, et al. Mutations in SEPT9 cause hereditary neuralgic amyotrophy. Nat Genet 2005;37:1044–6. [28] van Alfen N, Hannibal MC, Chance PF, et al. In: Pagon RA, Adam MP, Ardinger HH, et al., editors. Hereditary neuralgic amyotrophy. Seattle (WA): University of Washington; 1993–2016. [Source: GeneReviews® (Internet). 2008 Feb 27. Updated 2012 Dec 6]. [29] Vanneste JA, Bronner IM, Laman DM, et al. Distal neuralgic amyotrophy. J Neurol 1999;246:399–402. [30] Pham M, Baumer P, Meinck HM, et al. Anterior interosseous nerve syndrome: fascicular motor lesions of median nerve trunk. Neurology 2014;82:598–606. [31] Ochi K, Horiuchi Y, Tazaki K, et al. Fascicular constrictions in patients with spontaneous palsy of the anterior interosseous nerve and the posterior interosseous nerve. J Plast Surg Hand Surg 2012;46:19–24. [32] Seror P. Electrodiagnostic examination of anterior interosseous nerve. Normal and pathological data (21 cases). Electromyogr Clin Neurophysiol 1999;39:183–9. [33] Seror P. [Anterior interosseous nerve lesions revisited in 2004]. Rev Med Liege 2004;59:58–66. [34] van de Ven AC, van Alfen N, Heijdra YF. [An unusual cause of acute dyspnoea: neuralgic amyotrophy]. Ned Tijdschr Geneeskd 2009;153:A181. [35] van Eijk JJ, Madden RG, van der Eijk AA, et al. Neuralgic amyotrophy and hepatitis E virus infection. Neurology 2014;82:498–503. [36] Bradley WG, Chad D, Verghese JP, et al. Painful lumbosacral plexopathy with elevated erythrocyte sedimentation rate: a treatable inflammatory syndrome. Ann Neurol 1984;15:457–64. [37] Evans BA, Stevens JC, Dyck PJ. Lumbosacral plexus neuropathy. Neurology 1981;31:1327–30. [38] Seror P, Maisonobe T, Viala K, et al. [Idiopathic lumbosacral plexopathy]. Presse Med 2005;34:856–8.
158
P. Seror / Joint Bone Spine 84 (2017) 153–158
[39] Dyck PJ, Windebank AJ. Diabetic and nondiabetic lumbosacral radiculoplexus neuropathies: new insights into pathophysiology and treatment. Muscle Nerve 2002;25:477–91. [40] Seror P. [Injury of the external branch of the accessory nerve]. Rev Neurol 2002;158:759–62. [41] Friedenberg SM, Zimprich T, Harper CM. The natural history of long thoracic and spinal accessory neuropathies. Muscle Nerve 2002;25:535–9. [42] Pierre PA, Laterre CE, Van den Bergh PY. Neuralgic amyotrophy with involvement of cranial nerves IX, X, XI and XII. Muscle Nerve 1990;13:704–7. [43] Seror P. Isolated sensory manifestations in neuralgic amyotrophy: report of eight cases. Muscle Nerve 2004;29:134–8. [44] Ferrante MA, Wilbourn AJ. The utility of various sensory nerve conduction responses in assessing brachial plexopathies. Muscle Nerve 1995;18:879–89. [45] Aminoff MJ, Olney RK, Parry GJ, et al. Relative utility of different electrophysiologic techniques in the evaluation of brachial plexopathies. Neurology 1988;38:546–50. [46] Watson BV, Nicolle MW, Brown JD. Conduction block in neuralgic amyotrophy. Muscle Nerve 2001;24:559–63. [47] Vriesendorp FJ, Dmytrenko GS, Dietrich T, et al. Anti-peripheral nerve myelin antibodies and terminal activation products of complement in serum of patients with acute brachial plexus neuropathy. Arch Neurol 1993;50:1301–3. [48] Johnson NE, Petraglia AL, Huang JH, et al. Rapid resolution of severe neuralgic amyotrophy after treatment with corticosteroids and intravenous immunoglobulin. Muscle Nerve 2011;44:304–5. [49] Yoshida T, Sueyoshi T, Suwazono S, et al. Three-tesla magnetic resonance neurography of the brachial plexus in cervical radiculopathy. Muscle Nerve 2015;52:392–6.
[50] Gaskin CM, Helms CA. Parsonage-Turner syndrome: MR imaging findings and clinical information of 27 patients. Radiology 2006;240:501–7. [51] Yamamoto S, Nagano A, Mikami Y, et al. Fascicular constriction in the anterior interosseous nerve and other motor branches of the median nerve. Muscle Nerve 1999;22:547–8. [52] Pan Y, Wang S, Zheng D, et al. Hourglass-like constrictions of peripheral nerve in the upper extremity: a clinical review and pathological study. Neurosurgery 2014;75:10–22. [53] Comtet JJ, Chambaud D. [“Spontaneous” paralysis of the posterior interosseous nerve by unusual lesions]. Rev Chir Orthop Reparatrice Appar Mot 1975;61:533–41. [54] van Eijk JJ, van Alfen N, Tio-Gillen AP, et al. Screening for antecedent Campylobacter jejuni infections and anti-ganglioside antibodies in idiopathic neuralgic amyotrophy. J Peripher Nerv Syst 2011;16:153–6. [55] Vighetto A, Grand C, Confavreux C, et al. [The Parsonage-Turner syndrome and similar diseases, 29 cases]. Rev Neurol 1988;144:494–8. [56] van Eijk JJ, van Alfen N, Berrevoets M, et al. Evaluation of prednisolone treatment in the acute phase of neuralgic amyotrophy: an observational study. J Neurol Neurosurg Psychiatry 2009;80:1120–4. [57] Naito KS, Fukushima K, Suzuki S, et al. Intravenous immunoglobulin (IVIg) with methylprednisolone pulse therapy for motor impairment of neuralgic amyotrophy: clinical observations in 10 cases. Intern Med 2012;51:1493–500. [58] Seror P, Leger JM, Maisonobe T. Anterior interosseous nerve and multifocal motor neuropathy. Muscle Nerve 2002;26:841–4. [59] Amato AA, Jackson CE, Kim JY, et al. Chronic relapsing brachial plexus neuropathy with persistent conduction block. Muscle Nerve 1997;20:1303–7.