Introduction

Spasticity is a motor disorder characterised by a velocity dependent increase in the tonic stretch reflex [3]. It is usually accompanied by increased tone, brisk reflexes and weakness, all representing the upper motor neuron syndrome, found in a number of neurological conditions [4, 5]. Spasticity becomes clinically relevant when it interferes with function or care of patients, and is commonly treated with physical modalities or oral medications.

However, it is also frequently treated with local injections of either Botulinum neurotoxin, which inhibits the release of acetylcholine at the neuromuscular junction, causing muscle weakness, or by local chemical neurolysis involving application of local anaesthetics, phenol and alcohol to a local nerve [5]. Prior to treatment by local injection, informed consent is obtained from the patient or carers, including the possible side effects of medication discussed, using the product disclosure information [6].

Myositis is the collective term for illnesses that involve muscle inflammation with elevated creatine kinase (CK) and C reactive protein (CRP) [7]. Dermatomyositis, polymyositis, necrotizing myopathy and inclusion body myositis are four distinct subtypes of idiopathic inflammatory myopathies—in short, myositis. Recent studies have shed some light on the unique pathogenesis of each entity. While some of the clinical features are distinct, muscle biopsy is indispensable for making a reliable diagnosis. The use of magnetic resonance imaging of skeletal muscles and detection of myositis‐specific autoantibodies have become useful additions to the diagnostic repertoire [7]. Only few controlled trials are available to substantiate current treatment approaches for myositis, although novel approaches such as rituximab in patients with certain myositis-specific autoantibodies, and other biologic agents are currently being investigated [8].

In order to see if there is a relationship between Botox and myositis, an electronic literature search was performed using the Allergan Product Literature database and this did not identify any peer-reviewed publications specific to this topic. Please see Appendix B for the search strategy.

The search yielded two multicentre, randomized, double-blind, placebo-controlled studies (Studies 142 and 143) which evaluated Botox (n = 921) vs placebo (n = 257) for the temporary improvement in the appearance of moderate to severe forehead lines (FHL). There were no treatment-emergent adverse events (TEAEs) of myositis reported in Study 142. In Study 143, a TEAE of myositis was reported in one patient in the Botox 40 Unit (U) group (n = 318) and one patient in the Botox 64 U group (n = 746). All TEAEs of myositis in Study 143 were considered unrelated to the treatment by the investigators [9, 10]. The other search results did not report the complication of myositis following injections of onabotulinum toxin A [11, 12].

Case presentation

Rehabilitation ward

A 17-year-old man with sensory incomplete SCI (C4 AIS B) presented on admission to rehabilitation with severe spasticity in bilateral adductors, iliopsoas, rectus femoris and the right hamstrings, with a pattern of right knee extension and plantar flexion and left knee and hip extreme flexion. However, the Tardieu scale did not truly show the severity of the patient’s spasms, which occurred throughout the day and when he slept in supine position. His spasms were aggravated by light touch, pressure and vibration, rather than physical movement of the limbs.

He was trialled on increasing doses of oral Baclofen (which was subsequently changed over to Tizanidine) and Pregabalin, without any noticeable effect on his spasticity, and with significant side effects (drowsiness and altered liver function tests). Given the extent of the patient’s spasticity, intrathecal baclofen treatment was considered, but due to the fact that this was considered to be too invasive and permanent by the patient and his parents, decision was made to inject Onabotulinum toxin A (Botox) into the most affected muscles: bilateral adductor magnus and ilio-psoas. The adductor magnus muscles were injected with 100units (u) of Botox each diluted with 2mls of sterile normal saline (in 2 different sites, 50 u = 1 ml solution each), using electro-stimulation. The next day the ilio-psoas muscles were injected by a radiologist, using sterile technique and local anaesthetic, with CT scan guidance, after injection of contrast for good localisation (100 u Botox diluted in 5mls of sterile normal saline in each muscle, in one site). All four vials of Botox were from the same batch, having an identical serial number and expiry date.

After having the Botulinum Toxin injections, there was no significant improvement observed with this pattern of spasticity, in terms of his wheelchair positioning, lying in bed and functional training. In total, 40 days following the injections, he started to feel unwell, with episodes of autonomic dysreflexia, which were thought to be due to a urinary tract infection (UTI). The microbiology showed Klebsiella species sensitive to Gentamicin and Trimethoprim, so he was given for two consecutive days intramuscular Gentamicin 160 mg per day (in two sites on anterior aspect of the thighs), followed by oral Trimethoprim 300 mg per day. However, his white cell count (WCC) continued to rise to 12.1×10^9/L (normal value range: 4–11x^9/L), as did his C reactive protein (CRP) at 140 mg/L (normal value range: 0–5 mg/L). He had no abdominal pain, no vomiting, was afebrile, and all other observations were stable. Two days later, his CRP continued to rise, to 190 mg/L. He was still feeling unwell, and started to experience vomiting, a high-grade fever of 39 °C, regular tachycardia at 130 beats/minute, rigors and showed peritonism on clinical examination. As a consequence, he was transferred to an acute facility for investigation and management.

First acute hospital admission

The young man was consulted by different specialists, given that his creatine kinase (CK) was 3936units/L (normal value range: 45–250 u/L). A CT scan of his abdomen and pelvis was showing extensive intramuscular oedema of the bilateral adductor muscle groups and oedema tracking along the left psoas muscle. An ultrasound showed a small volume of inter-muscular fluid in the proximal medial left thigh, inflammatory thickening and oedema of the fascia next to the femoral vessels, with no drainable collection being detected. The most likely diagnosis considered was septic myositis, so he was started on intravenous broad-spectrum antibiotics. With this treatment, bed rest and intravenous fluids, after 6 days, his clinical status has significantly improved, CK was 266 u/L and CRP 68 mg/L, and was discharged back to rehabilitation.

The differential diagnoses considered were: autoimmune myositis (however, this does not cause severe pyrexia and rigors [7]), focal myositis (but this is well circumscribed and while an MRI would be of value for diagnosis [7], it was not considered as the patient would have needed sedation) or limb limited vasculitis of the leg muscles (this is usually unilateral and with lower CK). All auto-immune screening tests were negative. Muscle biopsy was not considered at this stage, as the patient improved.

Back in rehabilitation, after three days, there was worsening of his CK and CRP: CK 822 u/L and CRP 83 mg/L respectively, with normal full blood count and electrolytes.

Second acute hospital admission

The patient was transferred back for further investigations of cause and treatment of myositis. An MRI (T2 sequence below) showed extensive myofascial inflammatory changes, consistent with myositis of the left thigh adductors and pectineus (1), right anterior compartment myofascial oedema (2) and right femoral neck and lesser trochanter changes (3) suspicious of early osteomyelitis (Fig. 1).

Fig. 1: Creatine kinase and C reactive protein trends.
figure 1

CK and CRP trend over time.

Full size image

The orthopaedic team performed a left adductor muscle biopsy which showed prominent infarcted necrotic muscle without significant inflammation, ruling out acute infective pyomyositis (Fig. 2).

Fig. 2: Lower limbs Magnetic Resonance Imaging T2 sequence.
figure 2

MRI T2 sequence.

Full size image

The histological appearance was non-conclusive to the underlying cause, and possible considered events were: localised response to drug or toxic causes, localised trauma, thrombus/emboli, vasculitis, or vascular spasm. An arterial Doppler of the abdomen and lower extremities was normal. Auto-antibodies screen remained negative and the patient remained clinically well despite the concerning MRI and biochemical changes. Only conservative treatment was given (bed rest and intravenous fluids) and he returned to the rehabilitation ward after five days and continued inpatient therapy consisting of daily physiotherapy and occupational therapy. An occasional rise in CK and CRP was later noted, but with increase in fluid intake and limiting intensive exercises, these levels returned to normal (Fig. 3).

Fig. 3: Frozen sections with haematinin-eosin staining showing infarcted necrotic muscle, without inflammation.
figure 3

A Frozen section haematinin-eosin staining showing infarcted necrotic muscle, without inflammation. Neutrophils and histocytes (highlighted) appear to be a secondary reaction to the necrotic muscle. B Haematinin-eosin staining of infarcted muscle without significant inflammation. A fibrin thrombus seen in a degenerate vessel which is a secondary feature associated with the necrotic muscle.

Full size image

Applying the Tardieu scale of spasticity evaluation, the patient’s spasticity was worse at discharge than on admission, with main worsening noted in the left hamstring and bilateral adductors. Rectus femoris and iliopsoas measurements remained the same as on admission. He was successfully discharged home with the recommendation of fortnightly monitoring of CK and CRP. As the severe spasticity was affecting the patient’s function and quality of life, he was referred for consideration of intrathecal baclofen therapy.

Discussion

Given the severity of the patient’s symptoms, Allergan Medical Section was contacted and a severe adverse reaction report was completed. This service provided advice regarding a vast literature search looking for myositis following Botox injections. Only one study showed myositis in two patients who received Botox in facial muscles, but both cases were subsequently not thought to be related to Botox [9, 10]. In addition, there are a number of studies that indicate successful treatment of myositis with Botulinum Toxin [7, 13]. Mancini et al. [14] describes a subset of patients who experienced flu-like syndrome and oedema of the Botox injected leg, which some patients endured for more than 4 weeks, within a group of patients who had a higher dose (i.e., 100 u into soleus or 200 u into the gastrocnemius muscles).

The potentiating effects in our presented case might include the use of intramuscular Gentamicin also into the anterior thigh compartment, after the onset of symptoms, as well as influenza vaccination 10 days prior to onset of symptoms. The inability of our patient to localise his symptoms added to the severity and complexity of his presentation, and possibly delayed his diagnosis.

In summary, a dose-related effect (2 successive days of 200 units) of the onabotulinum A cannot be ruled out. This case remains open for discussion and no conclusions can be drawn, except perhaps to include myositis in the possible side-effects following injections of Botox for spasticity, when asking the patient for informed consent. Another potential consideration would be avoiding intramuscular aminoglycosides around sites of potential botulinum diffusion. There will be further scope to explore if point of care ultrasound may be of value for routine post botulinum toxin surveillance in the setting of treating patients with severe spasticity and concomitant sensory loss.