Clinical Neurology and Neurosurgery 128 (2015) 84–89
Contents lists available at ScienceDirect
Clinical Neurology and Neurosurgery journal homepage: www.elsevier.com/locate/clineuro
Tropical pyomyositis of erector spinae complicated with spinal epidural abscess Yun-Cong Zheng, Ching-Chang Chen, Kuo-Chen Wei, Jyi-Feng Chen, Shih-Tseng Lee, Cheng-Chih Liao ∗ Department of Neurosurgery, Chang Gung Memorial Hospital, Chang Gung University, Kweishan, Taoyuan, Taiwan
a r t i c l e
i n f o
Article history: Received 3 January 2014 Received in revised form 17 October 2014 Accepted 27 October 2014 Available online 11 November 2014 Keywords: Tropical pyomyositis Erector spinae Spinal epidural abscess Laminectomy Laminotomy
a b s t r a c t Study design: This is a retrospective case series. Objective: Tropical pyomyositis of erector spinae muscle (ESPM) is a rare muscular infection which may extend into the intraspinal canal to become spinal epidural abscess (ESPM-SEA). If left untreated, it may cause catatrophic spinal cord dysfunction and lead to irreversible paralysis. A series of eight such cases is presented, in order to provide proper surgical options and clarify the prognostic factors of the disease. Summary of background data: Merely six sporadic case reports had been found in the literature. Surgical debridement and laminectomy to drain the intraspinal abscess combined with systemic antibiotics were the choice of treatment to treat the disease with good therapeutic effect. Methods: Inpatient charts of the patients were reviewed. The therapeutic effect and functional neurological recovery are correlated with the demographic characteristics of the patients, neurological deﬁcits before drainage, and the different procedures of drainage. Results: Old age, long ESPM-SEA (>6.5 vertebral segments), spinal cord dysfunction as well as complete paralysis before the interventional procedures are signiﬁcantly correlated with poor functional neurological recovery (Sperman’s coefﬁcient correlation, all p < 0.05). Pig-tail drainage of ESPM combined with adequate systemic antibiotics could cure if infection presents with lumbar radiculopathy only, but it failed to rescue the spinal cord dysfunction in two patients present with complete paralysis. Surgical drainage of ESPM with mini-laminotomy to drain ESPM-SEA combined with systemic antibiotics provided good functional recovery of patients, despite of prolonged pre-operative complete paralysis. Conclusion: Early drainage of the ESPM and related epidural abscess combined with systemic antibiotics can provide excellent therapeutic effect of ESPM-SEA. Open drainage with mini-laminotomy is superior to pig-tail drainage when spinal cord dysfunction occurred associated with ESPM-SEA. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Tropical pyomyositis (PM) is an infectious disease of the skeletal muscles that presents with pain in the involved area and is accompanied by abscess formation . PM is usually more predominant in males and involves the large muscles of the pelvic girdle and lower extremities . During the early invasive stage of PM, an infectious myositis without purulent abscess usually mimics
∗ Corresponding author at: Department of Neurosurgery, Chang Gung Memorial Hospital, 5 Fu-shing Street, 333 Kweishan, Taoyuan, Taiwan. Tel.: +886 3281200x2119; fax: +886 3285818. E-mail addresses: [email protected]
, [email protected]
(C.-C. Liao). http://dx.doi.org/10.1016/j.clineuro.2014.10.022 0303-8467/© 2014 Elsevier B.V. All rights reserved.
regional myofascial pain which is difﬁcult to diagnose. The invasive stage of PM can last for several weeks and progresses to the second, purulent stage when necrosis of the infection focus within the involved muscle occurs. The usual clinical presentation is during the purulent stage, best diagnosed by contrast enhanced computed tomography (CT) or with magnetic resonance image (MRI) when suspecting epidural involvement. Intramuscular abscess formation can be conﬁrmed through needle aspiration of the abscess. In the ﬁnal, septic stage of PM, general reactions to sepsis such as fever, leucocytosis, tachycardia, and hypotension may occur . Emergency surgical drainage of the PM is indicated in life threatening septic shock. Although repeated minor trauma of the muscle and immunodeﬁciency have been postulated to precede PM, the real etiology remains unknown [4,5]. PM involving the posterior spinal muscle: erector spinae (ESPM) is rarely reported in the
Y.-C. Zheng et al. / Clinical Neurology and Neurosurgery 128 (2015) 84–89 Table 1 American Spinal Injury Association (ASIA) impairment score (AIS).
Summary of paralysis
3.1. Demographic characteristics of the patients
Motor and sensory complete
No sensory or motor function is preserved in the sacral segments S4–S5.
Motor complete and sensory incomplete
Sensory function is preserved below the neurological level of injury (NLI) including the sacral segments S4–S5, but no motor function is preserved more than three levels below the NLI on either side of the body.
Motor incomplete Major
Motor function is preserved below the neurological level, and more than half of key muscle functions below the NLI have a muscle grade less than 3 (active movement with full range of motion against gravity).
Motor incomplete Minor
Motor function is preserved below the neurological level, and at least half (half or more) of key muscle functions below the NLI have a muscle grade >3.
The sensation and motor function as tested to be normal in all segments below the NLI.
During 2007–2013, eight patients with conﬁrmed ESPM-SEAs were identiﬁed from the medical database of our institute. All of them were male. Table 3 describes the demographic characteristics, treatment procedures, and outcomes for all 8 patients with ESPM-SEA. The median age for the patients was 56.5 years old (range: 13–72) with an interquartile range (IQR) of 20.5 years old. Six patients (75%) recalled a mild injury to the back or neck prior to the painful onset of the disease. Two patients had chronic C-viral hepatitis, two had chronic alcoholism associated with liver function impairment, two had diabetes mellitus (DM), and two patients had no medical diseases that would impact immunocompetence. The duration from the ﬁrst incident of painful attack caused by ESPM to conﬁrmed veriﬁed diagnosis of ESPM ranged from 3 to 25 days with a median of 8 days (IQR = 11.75). The ESPM-SEA presented in this series involved multiple vertebral segments ranging from 2 to 23 segments with a median segment length of 6.5 (IQR = 11.5). Seven patients had leukocytosis with a white blood cell count higher than 10,000/dL. All patients had elevated C-reactive protein (CRP) ranging from 24.63 to 380 mg/L (median = 136.49 and IQR = 130.9) which progressively decreased after treatment with systemic antibiotics with or without drainage. Bacteriological study of the pus from the PM revealed one oxacillin-resistant Staphylococcus aureus (ORSA), six oxacillin-sensitive Staphylococcus aureus (OSSA), and one Streptococcus anginosus. Seven patients had an ESPM-SEA involving the vertebral region containing the spinal cord with four suffering from myelopathy (50%) prior to either CT-guided pig-tail drainage or surgical debridement. The period between onset and diagnosis in these four myelopathy patients ranged from 3 to 19 days with a median of 6.5 days (IQR = 12.25). Two patients (patients 1 and 2 in Table 3) whose ESPM involved the cervical region, developed AIS-A cervical myelopathy with respiratory failure and cardiac arrest. Patient 1 underwent surgical drainage of the ESPM as well as a partial C1 laminotomy to drain the ESPM-SEA three days after successful resuscitation (Fig. 1). Patient 1 made a slow recovery from the cervical myelopathy and could ambulate without assistance 15 months after the surgical treatment of the ESPM-SEA (Nurick-2 and AIS-D). Patient 2 underwent a CT-guided needle aspiration of the cervical ESPM but not the SEA due to persisting poor cognitive functioning after resuscitation (Fig. 2). Patient 2 suffered from persistent quadriplegia and died of uncontrolled sepsis three months after AIS-A cervical myelopathy developed. Both patients received three months of systemic antibiotics after the drainage procedures. Patient 3 suffered from thoracic ESPM and developed a C3 to T8 SEA associated with an AIS-A thoracic myelopathy lasting for three days prior to being transferred to our hospital. He underwent CT-guided pig-tail drainage of the ESPM with systemic antibiotics. Although the infection was controlled by systemic antibiotics and the CRP level returned to normal 3 months later, the neurological deﬁcits remained at AIS-A and the patient was wheel chair bound (Nurick-5) after 12 months of follow up. Patient 4 had alcoholic liver disease and suffered from right thoracolumbar ESPM and developed an AIS-D lumbar (L1) myelopathy with ESPM-SEA extending from C7 to L4-5 where asymptomatic grade 2 spondylolithesis could be identiﬁed (Fig. 3). Although the ESPM-SEA involved the whole spinal canal, he underwent surgical debridement to drain the ESPM and right L1 mini-laminotomy to drainage the purulent abscess. Combined with post-operative antibiotics for 3 months, the patient recovered to AIS-D and Nurick1 with some parasthesia bilaterally in the distal lower limbs. Four patients suffered from unilateral lumbar radiculopathy limited to the lumbar region where the ESPM entered the spinal canal through the intervertebral neuroforamen (Fig. 4). In these four patients, one
English literature because the muscle, surrounded by tight fascia, is resistant to the invasion of blood born microorganisms . Even more rarely will the ESMP extend into the spinal canal to cause an epidural abscess (SEA) [5,7–10]. Due to the rarity of ESPM-SEA, the authors present a case series detailing their treatment experiences and compare those results with the currently available literature in order to clarify the clinical presentation, pathogenesis, and proper treatment strategies for this rare infectious disorder. 2. Materials and methods This is a retrospective chart review study. Demographic characteristics, clinical manifestations, location of the ESPM and extension of the ESPM-SEA, related neurological deﬁcits, therapeutic strategies, results from bacterial cultures, and treatment outcomes were reviewed and analyzed. For patients that received therapy for multiple stages, the initial therapeutic plan and the supplemental procedures in response to the ineffectiveness of the initial treatment were documented as primary and secondary treatments. The neurological status before and after the treatment was recorded by American Spinal Injury Association (ASIA) impairment scores . The summary and description of ASIA impairment scores (AIS) are listed in Table 1. A modiﬁed Nurick scale was used to assess the functional outcome of the patients . Table 2 displays the modiﬁed Nurick classiﬁcation scale with grade 6 representing disease-related in-hospital mortality. The lower the Nurick grade, the better the functional outcome. Spearman’s correlation coefﬁcient was used to correlate the ﬁnal Nurick grade with the disease characteristics and the therapeutic procedure. Nonparametric and Fisher’s exact tests were used to compare the ﬁnal functional neurological grades for ordinal and binary prognostic factors, respectively. Table 2 Nurick grades with a modiﬁed grade 6 implying disease related in-hospital mortality. Grade
1 2 3 4 5 6
Normal walking with possible clinical spinal irritation Slight difﬁculties in walking with normal domestic and working life Functional disability limiting normal work and domestic activities Signiﬁcant weakness making walking impossible without help Bedridden or wheelchair-bound Disease related in hospital mortality
Y.-C. Zheng et al. / Clinical Neurology and Neurosurgery 128 (2015) 84–89
Fig. 1. The T2 MRI (A) and enhanced T1 MRI (B), acquired by 1.5-Tesla MR scanner, of a cervical tropical pyomyositis of erector spinae (ESPM) with spinal epidural abscess (SEA) showed severe spinal cord compression occurring at the C1-2 region resulting in complete high cervical paralysis with respiratory compromise. The ring enhanced ESPM (B, arrow) penetrated the epidural space through the posterior atlanto-axial ligament and was drained by surgical debridement as well as C1 laminotomy. The immediate post-operative cervical MRI (C) showed good decompression of the C1-2 spinal cord with a residual SEA in the posterior epidural space at the cervical thoracic junction. One year follow up MRI (D) showed complete resolution of the SEA with myelomalacia (D, arrow) at the C34 spinal cord but good recovery of neurological function with slight difﬁculty in walking without assistance.
Fig. 2. The post-contrast T1 sagital (A) and axial (B) MRI showed multiple tropical pyomyositis of erector spinae (ESPM) at the lumbar region (A and B, black arrow). A ring-enhanced C1-2 ESPM (C, white arrow) was noted. The abscess extended through the neuroforamen to form an epidural abscess and cause diffuse inﬂammation of the leptomeninge and spinal cord (D, white arrow). The patient suffered from prolonged quadriplegia and respiratory failure and died of sepsis despite pig-tail drainage of the C1-2 ESPM.
Fig. 3. The coronal section of enhanced MRI (A) showed the lateral extent of thoraco-lumbar tropical pyomyositis of erector spinae (ESPM) which penetrate the spinal canal through neuroforamen (A, arrow). The sagital lumbar (B), thoracic (C) and cervical (D) MRI showed the whole spine extension of the spinal epidural abscess (C and D, short arrow head) with maximal spinal cord compression at the L1-2 segment due to coexisting protrusion disk (B and C, arrow). Spinal ﬁxation screws could be visualized at the upper cervical spine. The patient developed L1 myelopathy with decreased muscle powers of both lower limbs which could elevate against gravity. The motor deﬁcit completely recovered after open debridement of ESPM and L1 laminotomy to drain the liquid portion of associated spinal epidural abscess.
22 16 38 36
Monoplegia D OSSA
D/1 E/0 E/0 E/0 A+C A+C A A+B L1 myelopathy L3 radiculopathy L4 radiculopathy L5 radiculopathy C7-L4 L1-4 T12-L5 T12-L3 T12-L4 L1-5 L4-S1 L3-5 234.05 120.15 139.99 24.63
N 52 8
Y Y Y N 64 40 16 53 4 5 6 7
3 7 9 25
Y 60 3
Viral hepatitis Alcoholism Alcoholism – Viral hepatitis –
Paraplegia Monoplegia Monoplegis Monoplegia D D D D OSSA OSSA OSSA OSSA
A/5 Paraplegia A T7 myelopathy C3-T8
C2 myelopathy C2 myelopathy
A+C A+B SEA
C1-3 C1-3 and L3-S1 T5-8 260.41 380 DM DM Y Y 72 62 1 2
Quadriplegia Quadriplegia A A
OSSA Streptococcus anginosus ORSA
Follow up (months) Outcome (AIS/Nurick) Paralysis before drainage Bacteria study Worst AIS Treatment Neurological deﬁcits Disease distribution
Initial CRP (mg/L) Chronic disease Pain attack to diagnosis (days) Age
underwent surgical debridement and L3 mini-laminotomy to drain the ESPM-SEA, two patients underwent CT-guided pig-tail drainage of the ESPM, and one patient received three months of systemic antibiotics treatment. All four of the patients recovered to AIS-E with Nurick-0. 3.2. Prognostic factor evaluation
Table 3 Demographic characteristics of patients (N = 8). The treatment is expressed as (A) antibiotics, (B) CT guided pig-tail insertion, or (C) surgical debridement with mini-laminotomy. All drainage procedures were coupled with adequate antibiotics treatment.
Y.-C. Zheng et al. / Clinical Neurology and Neurosurgery 128 (2015) 84–89
The functional outcome represented by Nurick grades at the last follow up was negatively correlated with increased age (Spearman’s correlation coefﬁcient = −0.873, p = 0.005), but not correlated with initial body temperature, white blood cell count, and Creactive protein (all p > 0.05). Meanwhile, the length of ESPM-SEA >6.5 segments, the presence of spinal cord dysfunction (myelopathy), and complete paralysis occurring before abscess drainage are adverse prognostic factors for ﬁnal neurological outcome (Table 4). Of the three patients with complete paralysis before the drainage procedure, surgical debridement of the ESPM with minilaminotomy to drain the SEA provided good functional recovery for one patient, but CT-guided needle drainage of the ESPM was unable to recover any spinal cord function for the other two. 4. Discussion ESPM accounts for only 4–6% of the total occurrences of tropical pyomyositis [2,4] and was postulated to be associated with micro-trauma of the erector spinae caused by unintended sprains, vigorous exercise, as well as Chinese acupuncture [13,14]. In the past decade, immune-deﬁciency, such as AIDS, DM, cirrhosis, and steroid therapy was also postulated to be a reason contributing to the occurrence of PM [15,16]. In the present study, seven out of eight patients (87.5%) recalled a minor injury to the neck or back preceding the subsequent chronic regional ache. Meanwhile, two out of eight patients in the study had DM, two had chronic viral hepatitis, and two had chronic alcoholic liver disorders which are all well known as diseases with immunocompromising features. High prevalence of minor back injury and possible immune-deﬁciency in this small patient series does support the previously reported provocative factors of ESPM. Computed tomography (CT) with contrast can provide a prompt diagnosis of PM in the limbs or trunk but not intraspinal abscesses . MRI with gadolinium enhancement remains the preferred diagnostic tool to identify ESPM-SEA [17,18]. In the present study, all of the SEAs could be diagnosed with sagital and axial MRI, but the extent of the abscess in the erector spinae could be better visualized with axial or coronal images (Fig. 1A). Compared to anteriorly located spinal abscesses which result from spondylodiscitis, SEA secondary to ESPM is mainly located posterior-laterally to the dural sac . When the ESPM is not managed by appropriate treatment, the infection might spread to the internal venous plexus or directly pass through the intervertebral foramens and ligmentum ﬂavum. After penetration into the epidural space, the abscess may extensively spread through long segments due to the negative pressure in the epidural space. Neurological deﬁcits, an indicator for emergent drainage, occur in 30% of SEAs [17,19]. The longer the symptoms persist, the higher the possibility of permanent disability. Short segmented ESPMSEA in the lumbar region can irritate the lumbar nerve root and result in radiculopathy, but patients with ESPM-SEA that extends longer than 6.5 spinal segments were likely to develop potentially irreversible myelopathy due to the compressing of the spinal cord or the resulting epidural vasculitis leading to spinal cord ischemia. In the present study, there were four patients (50%) whose ESPMSEA extended no less than 6.5 spinal segments who suffered from myelopathy and three of them (75%) had complete paralysis with a high probability of mortality and morbidity (one died of
Y.-C. Zheng et al. / Clinical Neurology and Neurosurgery 128 (2015) 84–89
Fig. 4. The sagital (A) and axial (B) T1 MRI of lumbar spine showed a multiple ring enhanced tropical pyomyositis of erector spinae (ESPM) at the lumbo-sacral region with anterior extension to the left psoas muscle (B, *) and pre-vertebral soft tissue. The ESPM extended through the left neuroforamen (A and B, arrow) to form left posterior lateral spinal epidural abscess (C, arrow) which results in left L5 radiculopathy. CT-guided pig-tail drainage of the lumbo-sacral ESPM combined with three months systemic antibiotics effectively eradicate the infection with complete recovery of the neurological irritation.
the disease, one became permanent paraplegic, and two partially regained functional independence). In comparison to the other four patients having lumbar ESPM-SEA and radiculopathy only, all the latter patients could achieve independent functional recovery without residual neurological deﬁcit (Nurick-0). Therefore, the longer ESPM-SEA (>6.5) did contribute to a signiﬁcant worse recovery of the neurological deﬁcits (p = 0.014, Mann–Whitney U test). In comparison to the other four patients having lumbar ESPM-SEA and radiculopathy only, all of latter patients could achieve independent functional recovery without residual neurological deﬁcit (Nurick-0). Therefore, the longer ESPM-SEA (>6.5) did contribute to a signiﬁcant worse recovery of the neurological deﬁcits (p = 0.014, Mann–Whitney U test). However, we did not observe any transitional symptom evolution from radiculopathy to myelopathy in the study to conclude that the radiculopathy is the early stage, and myelopathy represents the late stage of ESPM-SEA. In the previous literature, the patients with ESPM-SEAs were all treated with systemic antibiotic infusion combined with immediate surgical debridement of the muscular abscess and extensive laminectomy to drain the associated SEA [4,7,9] in case that the spinal cord dysfunction may deteriorate suddenly to cause complete paralysis which is generally accepted to be a major adverse prognostic factor . For patients suffering from complete paralysis for longer than 24–36 h, antibiotic infusion with or without CT-guided pig-tail drainage of the epidural abscess is suggested due to the high peri-operative morbidity and limited neurological recovery . In the present study, complete and prolonged paralysis occurred in three patients (two cervical and one thoracic). In one patient, C1 laminectomy and surgical debridement of the upper cervical ESPM-SEA resulted in the good recovery of functional performance (from AIS-A to AIS-D and Nurick-2), but
CT-guided pig-tail drainage was not efﬁcacious to produce neurological recovery in the other two patients. Bowen reported a pediatric thoracic ESPM-SEA patient that presented with complete paralysis was successfully treated with open surgery and laminoplasty to achieve a complete recovery . Olson also reported an ESPM-SEA patient who suffered from delayed neurological deterioration during CT-guided pig-tail drainage of the ESPM that required secondary laminectomy to reverse the deteriorating myelopathy . Therefore, we recommend that, even when complete paralysis lasts longer than 24–36 h, open surgery to drain the ESPM-SEA combined with adequate systemic antibiotics shall be performed as soon as possible, if the vital signs of patient can be stabilized after resuscitation. However, in long segmented ESPM-SEA, extensive laminectomy to drain the intraspinal abscess is impractical and may induce post-operative spinal instability and deformity . In the present study, the three extensive ESPM-SEA patients underwent limited laminotomy focusing on the location where the ESPM entered the intraspinal canal and exerted the most severe compression of spinal cord based on the MRI. Although the limited laminotomy could only drain the liquid portion of the epidural abscess with large portion of infectious epidural soft tissue left behind (Fig. 1), post-operative systemic antibiotics for three months provided a good therapeutic effect. For those four patients with short segment ESPM-SEA, restricting the thoracolumbar region without spinal cord dysfunction, early systemic antibiotics combined with or without CT-guided pig-tail drainage provided good therapeutic effect and preserve neurological function. If late deterioration occurred after the minimal invasive drainage procedure, open surgery with laminectomy to ensure maximal decompression of spinal cord and drainage of the abscess should be performed in case of irreversible spinal cord dysfunction might occurred and predispose to poor prognosis.
Table 4 Signiﬁcant prognostic factors of ESPM-SEA (Mann–Whitney U test). Prognostic factor (N)
Final Nurick grade Median
Radiculopathy (4) Myelopathy (4)
0 [−0.29, 7.28]
Incomplete (5) Complete (3)
[−0.83, 9.5] [−0.35, 0.75]
Length of ESPM-SEA
>6.5 (4) 6.5 vertebral segments), neurological deﬁcits frequently occur as an indication for emergent drainage of the abscess. CT-guided pig-tail drainage of the ESPM combined with systemic antibiotics may provide good therapeutic effect when the intraspinal abscess causes nerve root irritation and radiculopathy only. If there is spinal cord dysfunction or myelopathy that developed secondary to the ESPM-SEA, open surgical drainage of the ESPM with minimal laminotomy to drain the liquid portion of intraspinal abscess combined with systemic antibiotics provides good recovery from the neurological deﬁcits even in the condition of complete paralysis lasting for 24–36 h. References  Chiedozi LC. Pyomyositis review of 205 cases in 112 patients. Am J Surg 1979;137:255–9.  Bickels J, Ben-Sira L, Kessler A, Wientroub S. Primary pyomyositis. J Bone Joint Surg Am 2002;84A:2277–86.  Yu CW, Hsiao JK, Hsu CY, Shih TT. Bacterial pyomyositis: MRI and clinical correlation. Magn Reson Imaging 2004;22:1233–41.  Crum NF. Bacterial pyomyositis in the United States. Am J Med 2004;117:420–8.
 Olson DP, Soares S, Kanade SV. Community-acquired MRSA pyomyositis: case report and review of the literature. J Trop Med 2011:970848 http://dx.doi. org/10.1155/2011/970848  Grose C. Bacterial myositis and myositis. In: Feigin RD, Cherry JD, editors. Textbook of pediatric infectious disease. 4th ed. Philadelphia: Saunders; 1998. p. 704–8.  Bowen DK, Mitchell LA, Burnett MW, Rooks VJ, Martin JE. Spinal epidural abscess due to tropical pyomyositis in immunocompetent adolescents: report of 2 cases. J Neurosurg Pediatr 2010;6:33–7.  Mitchell LA, Rooks VJ, Martin JE, Brugos RM. Paraspinal tropical pyomyositis and epidural abscesses presenting as low back pain. Radiol Case Rep 2009;4:1–5.  Marshman LAG, Bhatia CK, Krishna M, Friesem T. Primary erector spinae pyomyositis causing an epidural abscess: case report and literature review. Spine J 2008;8:548–51.  Srinivasan US, Jagadeesh PC, Senthil KG. Pyogenic epidural abscess with primary erector spinae pyomyositis – case report. Open Spine J 2010;2:17–20.  Kirshblum SC, Burns SP, Biering-Sorensen F, Donovan W, Graves DE, Jha A, et al. International standards for neurological classiﬁcation of spinal cord injury (revised 2011). J Spinal Cord Med 2011;34:535–46.  Nurick S. The pathogenesis of the spinal cord disorder associated with cervical spondylosis. Brain 1972;95:87–100.  Romeo S, Sunshine S. Pyomyositis in a 5-year-old child. Arch Fam Med 2000;9:653–6.  Yuksel H, Yilmaz O, Orguc S, Yercan HS, Aydogan D. A pediatric case of pyomyositis presenting with septic pulmonary emboli. Joint Bone Spine 2007;74:491–4.  Christin L, Sarosi GA. Pyomyositis in North America: case reports and review. Clin Infect Dis 1992;15:668–77.  Hall RL, Callaghan JJ, Moloney E, Martinez S, Harrelson JM. Pyomyositis in a temperate climate: presentation, diagnosis, and treatment. J Bone Joint Surg Am 1990;72:1240–4.  Darouiche RO. Spinal epidural abscess. N Engl J Med 2006;355:2012–20.  Soler R. Magnetic resonance imaging of pyomyositis in 43 cases. Eur J Radiol 2000;35:59–64.  Govender S. Spinal infections. J Bone Joint Surg Br 2005;87:1454–8.