Diabetic Amyotrophy (Bruns-Garland Syndrome): A Narrative Review

Abstract

Background and Aims:

Bruns-Garland syndrome (BGS) continues to be a contentious topic even 100 years after its discovery. Its lifelong incidence is 1% amongst diabetic individuals and affects middle aged-elderly individuals with type 2 diabetes mellitus (usually not poorly controlled).

Methods:

The data presented in this review was collated from studies published on PubMed, MEDLINE and Google Scholar in October 2021. The search words included: “Bruns-Garland syndrome”, “diabetic amyotrophy” and “diabetic lumbosacral radiculoplexus neuropathy” and “proximal diabetic neuropathy”.

Results:

The exact pathophysiology is debatable but an ischemic pathology (non-systemic microvasculitis) is most plausible. Its cardinal symptoms include acute onset of severe proximal lower extremity pain followed by weakness and wasting, some sensory loss, weight loss and autonomic symptoms.

Conclusion:

The prognosis is good as most patients improve to near- normal strength with pain cessation within 18 months of onset.

INTRODUCTION

Bruns-Garland syndrome (BGS) continues to be a contentious and much debated topic even more than 100 years after its initial discovery. It was originally described by Bruns[1] in 1890 and later by Garland and Taverner in a case series in 1955.[2,3,4,5] They described this syndrome as one characterized by asymmetric, lower extremity-predominant symptom complex occurring in patients with short and “nonsevere” diabetes and coined the often used but erroneous term “diabetic amyotrophy.” The term amyotrophy makes one refer the localization to muscle or anterior horn cell. For this reason, multiple other nomenclature have been described for it: Proximal diabetic neuropathy,[6,7,8] diabetic proximal amyotrophy,[9] diabetic lumbosacral plexopathy,[10,11] femoral-sciatic neuropathy[12] and femoral neuropathy,[13,14] and diabetic lumbosacral radioculo-plexo-neuropathy.[15] There is some correctness in all of these, but we support the eponym Bruns-Garland syndrome. This not only prevents future confusion, but also follows the working precedent established for diseases like Guillain-Barré syndrome and Emery-Dreifuss muscular dystrophy, where descriptive terms are misleading.[16]

The lifelong incidence of BGS is approximately 1% among diabetic individuals[17] and generally affects middle-aged elderly individuals with type 2 diabetes mellitus. The diabetes mellitus is usually not poorly controlled may occur soon after diagnosis or may be the presenting feature leading to its diagnosis.[16]

MATERIAL AND METHODS

The data presented in this review were collated from studies published on PubMed, MEDLINE, and Google Scholar in October 2021. The search was conducted by AA and VVY and the search words included the following: “Bruns-Garland syndrome,” “diabetic amyotrophy” and “diabetic lumbosacral radiculoplexus neuropathy,” and “proximal diabetic neuropathy.” These articles were thereafter screened for relevance and all articles with human patients and published in English till December 2021 were included. Only one unpublished randomized controlled trial (RCT; abstract published) by Dyck et al.[18] was found, with the remainder being case series or reports.

Etiopathogenesis, clinical presentation, and evaluation

The exact pathophysiology of BGS is debatable but a painful onset and subacute, progressive course suggest an ischemic pathology in the form of a nonsystemic microvasculitis. Many studies conducted biopsies on the sural and intermediate cutaneous nerve of thigh and showed features supportive of this hypothesis.[19,20,21] Inflammatory changes were found involving the epineural vessels, including vasculitis or inflammatory infiltrates. Immunostaining also revealed increased nuclear factor-κB staining and ICAM-1 cells in the blood vessels supporting a dysimmune vascular pathogenesis.[22]

The usual age of presentation of BGS is the sixth decade of life (median onset: 60 years) with a slight male predilection (male: female-3:2). It occurs within a few years of diagnosis of diabetes mellitus (median: 3 years).[15] Occasionally, its presence sometimes led to the diabetes diagnosis and is commoner with type 2 diabetes. The blood sugar control is usually good with no concomitant microvascular complications (nephropathy and retinopathy) are present. However, some patients can present with poor glycemic control. There was no correlation with the use of any particular hypoglycemic agent (oral drugs vs insulin).

Its cardinal symptoms included acute onset of severe proximal lower extremity pain followed by weakness, mild-to-moderate sensory loss, weight loss, and autonomic symptoms.[15] Pain involves the thigh, hip, or back and is unilateral in most cases (approximately 70%) and is asymmetric in bilateral cases. It is severe in intensity and deep aching in character (with or without paresthesia). This is succeeded in a few weeks by atrophy and weakness of the afflicted muscles (predominantly: quadriceps, hip adductors, and iliopsoas),[23] which is again proximal predominant. It is unilateral in 25% cases and bilateral but asymmetric in the remaining cases. This contralateral spread occurs within 6 months. A small proportion developed can progress to develop diffuse weakness leading to quadriparesis or cause upper limb involvement as well. Progression in the involved limb to cause distal weakness occurs in 60% of cases.[15] The afflicted region has loss of reflexes. Distal symmetric sensory involvement is also common, but not clinically useful because of its nonlocalizing value. Approximately, a quarter of cases have associated autonomic dysfunction. Significant weight loss is a common accompaniment, and usually begins with symptom onset. Rarely, symptom onset suggestive of cervical radiculoplexus neuropathy may occur.[24] These patients have similar findings as mentioned above but in the upper limbs. Table 1 lists the differences between the 2 presentations.

Table 1.

Comparison between diabetic lumbosacral and cervical radiculoplexus neuropathies

	Diabetic lumbosacral radiculoplexus neuropathy	Diabetic cervical radiculoplexus neuropathy
Male:female	3:2	2:1
Median age at onset (in years)	60	62
Most common presenting symptom	Pain	Pain
	No pain-10%	No pain-20%
Progression	Chronic for 12-18 months in the majority	Chronic progression (>1 month) in 15% cases
Bilateral involvement	85%	47%
Weight loss	>50%	35%
CSF protein elevation	85%	90%
NCS/EMG	Sciatic/femoral/obturator nerve and lumbar roots	Upper plexus- 52%
		Middle plexus- 45%

CSF=cerebrospinal fluid, NCS=nerve conduction studies, EMG=electromyography

Examination characteristically reveals weakness in the proximal muscles of the affected limb with concomitant atrophy of hip adductors, quadriceps, and iliopsoas. Mild distal sensory loss may be present due to coexistent diabetic distal symmetrical polyneuropathy (DSPN). The knee jerks are absent and the ankle jerks are usually preserved. However, the latter can be lost with underlying DSPN.[23]

Nonetheless, the eventual prognosis is good as most patients tend to improve to near- normal strength with pain cessation within 18 months of symptom onset. Some distal weakness may persist in 50% cases and around 10% to 15% can develop relapsing symptoms.

However, approximately 10% patients might continue to be dependent for ambulation with incomplete recovery after 2 years of onset.[15]

Since the diagnosis is mainly clinical, a few similarly presenting differential diagnoses need to be excluded. Neoplastic lumbosacral plexopathy needs to always be excluded and the role of neuroimaging in BGS is predominantly for this purpose. If present, it reveals infiltration of lumbosacral nerve roots by the tumor leading to this presentation.

Cauda equina syndrome and spinal canal stenosis can rarely present in this fashion and can also be diagnosed by neuroimaging. GBS and chronic inflammatory demyelinating polyneuropathy can also mimic BGS but clinical clues like progression for <1 month, nerve thickenings, and demyelinating pattern on nerve conduction studies (NCS) help clinch the diagnosis.[23,24]

Diagnostic evaluation for BGS is supportive and no gold standard test is available at present. The diagnosis is made on the basis of a thorough history and examination. Magnetic resonance imaging of the lumbosacral spine is usually performed to rule out structural causes that can manifest similarly. Magnetic resonance imaging findings in BGS include lumbosacral enhancement in the roots, plexus, and nerves with T2 hyperintensities in the involved muscles.[25] The cross- sectional area of these structures is also increased compared with controls.[26] However, these changes are nonspecific and can be seen in inflammatory radiculoneuropathies and radiculitis. Hlis et al.[27] showed that neuromuscular lesions secondary to BGS can be qualitatively and quantitatively identified by magnetic resonance neurography and diffusion tensor imaging, and that the most commonly affected structures were the femoral and sciatic nerves followed by lumbosacral nerve roots and finally obturator nerves. They also revealed a positive correlation of apparent diffusion coefficient values with the patient's hemoglobin A1c levels.

Therefore, these tests can be employed when BGS is suspected. NCS reveal an amplitude drop of the nerves of the affected region suggestive of axonal neuropathy and electromyography (EMG) findings are those of the involved muscles showing active denervation with or without reinnervation suggestive of a lumbosacral plexopathy. However, paraspinal involvement is common, which is indicative of concomitant radicular involvement, and should always be tested. These EMG findings are commonly asymmetric, with more severe affliction of one side. A normal NCS and EMG study in a suspected BGS patient is a red flag and should prompt a diagnostic reconsideration.

Cerebrospinal fluid analysis may sometimes be done classically, which reveals elevated protein (median: 90 mg/dL) and normal cells (albumin-cytological dissociation) in approximately 85% patients.[15] This finding is similar to that found in GBS or chronic inflammatory demyelinating polyneuropathy and does not add to diagnosis and is not routinely advocated for diagnostic evaluation for this reason. The presence of pleocytosis is a red flag and should prompt a search for mimics.

Testing for other autoimmune diseases is unlikely to add to the diagnostic yield. Inflammatory markers like erythrocyte sedimentation rate and C-reactive protein tend to be raised in 20% cases but are nonspecific. Glycosylated hemoglobin A1c values tend to be mildly elevated with median value of 8.0%. Nerve biopsies are not recommended. Nerve biopsy studies done till date were done either on the intermediate cutaneous nerve of the thigh or the sural nerve. Those done on the latter are of unclear relevance, because the sural nerve contains axons derived from the S1 nerve root, whereas clinical findings predominantly localize to the axons from the L2-L4 nerve roots.[28]

The treatment of BGS focusses on pain control and improvement in strength above and beyond good euglycemic control. Pain is treated with drugs that have been proven to be effective in managing painful DSPN like selective serotonin reuptake inhibitors, tricyclic antidepressants, anticonvulsants, and selective noradrenaline reuptake inhibitors.[15,23]

Some improvement in pain is also seen with immunotherapy agents. The prototypical treatment protocol for motor improvement remains controversial in the absence of evidence from RCTs. Due to a postulated microvasculitic etiology, treatment with immunotherapy seems prudent apart from basic supportive measures. The latter includes active physiotherapy and usage of assistive devices to enhance motility. Despite widespread interest in immunotherapy, results from various studies have been nonconsistent and conflicting.[29] Corticosteroids are often used as a first-line agent. Dyck et al.[18] showed improvement in neuropathic pain and secondary endpoints but no significant improvement in motor power with intravenous methylprednisolone. Krendal et al.[30] showed some improvement in strength with intravenous immunoglobulin. Most of these patients, however, also received intravenous or oral steroids. Pascoe et al.[31] showed improvement in 4 out of 5 patients treated with plasma exchange with or without intravenous immunoglobulin. Similar findings were also exhibited by Jaradeh et al.[32] in 9 patients. However, considering all these studies were not RCTs and a natural tendency toward improvement after 12 to 18 months, these results need to be interpreted with caution. Patients can be depressed considering the background of acute/subacute onset painful weakness and may need reassurance and antidepressants.

CONCLUSION

BGS is a rare diabetic complication but considering the humongous number of patients living with diabetes, it is likely to encounter frequently. This warrants increasing awareness about it and its management among neurologists and nonneurologists (endocrinologists and primary physicians). Immunotherapy is controversial and their use should be weighed based on pros and cons of treatment. Earlier recognition will lead to rapid diagnosis, which might provide new insights into pathophysiology and progression, and an RCT is needed to answer the perpetual query whether to treat or not to treat with immunotherapy.

Research highlights

1. Cardinal symptoms of BGS are acute onset of severe proximal lower extremity pain followed by weakness and wasting

2. Exact pathophysiology is debatable but nonsystemic microvasculitis is most plausible

3. Usually self-limiting with pain cessation and near normal power attainment within 18 months of onset

4. Immunotherapy is controversial.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.