ALTHOUGH back pain is most commonly a benign and self-limiting complaint after regional anesthesia, it may require further evaluation and specific treatment. We report a case of lumbar vertebral osteomyelitis presenting as back pain lasting 3 weeks after the removal of an epidural catheter placed for postoperative analgesia. Further investigation resulted in identifying potential causes for the infection.

Case Report

A 76-yr-old, 70-kg man was admitted for a 3-day fever, productive cough, and shortness of breath. His medical history was significant for a longstanding smoking history and bullous emphysema. On admission, a chest x-ray showed bilobar infiltrates consistent with pneumonia. He received intravenous antibiotics for treatment of a Pseudomonas aeruginosa infection identified by sputum cultures. Intravenous methylprednisolone was given to treat his obstructive pulmonary disease.

The patient's hospital course was complicated by an acute massive upper gastrointestinal hemorrhage on hospital day 4, which required emergency surgery. A right pneumothorax subsequently developed, which was decompressed immediately with a chest tube. The patient ultimately underwent a right thoracotomy on hospital day 14 for recurrence of the right pneumothorax secondary to rupture of multiple pulmonary bullae. Before anesthetic induction for the thoracotomy, an epidural catheter was inserted two interspaces above the intercristal line without difficulty using sterile technique. The patient tolerated general anesthesia well, and the trachea was extubated immediately after surgery. He received adequate pain relief with intermittent injections of 3 mg epidural preservative-free morphine every 6–12 h. The epidural catheter was removed on postoperative day 4 without evidence of complications. The patient was discharged on postoperative day 7 on oral antibiotics. The patient had no specific complaints of back pain at that time.

Two weeks after discharge, the patient returned to the hospital for evaluation of persistent low back pain. The pain was described as sharp, localized to the midline near the epidural insertion site, and worsening with activity. He was taking large doses of oral analgesics to help relieve the back pain. The patient denied fever, chills, headache, lower extremity weakness, numbness, tingling, or incontinence of the bladder or bowel. There was no previous history of low back pain. The patient noticed that his gait was "unsteady" that morning, but the symptoms had resolved after a few hours. A thorough neurologic examination revealed no gross or focal neurologic deficits. On examination of his back, there was no evidence of the epidural puncture site, erythema, or discharge. A dorsal spinous process in the upper lumbar region was exquisitely tender and moderately protuberant. Anteroposterior and lateral spine x-rays showed mild degenerative changes of the lumbar vertebrae, but the results were otherwise normal. Magnetic resonance imaging (MRI) of the spine in the sagittal view (Figure 1) demonstrated edematous changes of the L1 spinous process. The L2-L3 interspace showed normal fat without edema. After review of the midline sagittal view, the neuroradiologist noted what appeared to be an epidural tract in the L2-L3 interspace left by the epidural catheter. A radionuclide bone scan of the spine was floridly positive over the spinous process of L1. A computed tomography-guided percutaneous bone biopsy of the L1 spinous process was obtained. Cultures of the bone biopsied resulted in growth of P. aeruginosa. The patient was diagnosed with vertebral osteomyelitis secondary to P. aeruginosa involving the spinous process of the L1 vertebra. He was discharged home with a 4-week course of intravenous gentamicin and oral ciprofloxacin as outpatient therapy. He recovered without evidence of neurologic deficits, and his back pain resolved shortly thereafter.

Figure 1. Magnetic resonance imaging scan of the lumbar spine demonstrating vertebral osteomyelitis of the L1 spinous process. Note edematous tissue at L1. L2-L3 arrow points at probable epidural tract.

Figure 1. Magnetic resonance imaging scan of the lumbar spine demonstrating vertebral osteomyelitis of the L1 spinous process. Note edematous tissue at L1. L2-L3 arrow points at probable epidural tract.


Reports of new-onset backache after epidural anesthesia vary from 2% to 31%. [1]The most common causes of back pain after regional anesthesia are thought to include ligamentous trauma, reflex paraspinous muscle spasm, or ligamentous strain during patient positioning secondary to skeletal muscle relaxation. The use of the preparation 2-chloroprocaine-containing EDTA may contribute to back pain in ambulatory surgical patients undergoing epidural anesthesia. [2]Symptoms related to back pain usually are mild and self-limiting and respond well to conservative therapy. Back pain after regional anesthesia that occurs concomitantly with neurologic dysfunction is rare and should immediately alert the physician to search for other potential causes. Major neurologic dysfunction after regional anesthesia may be caused by neuraxial space-occupying lesions, such as epidural hematoma or abscess, direct or chemically induced neural damage, or ischemic injury to the spinal cord. [3]The presence of preexisting medical conditions should be considered, including inflammatory, degenerative, or malignant disease of the neuraxial structures. Although our patient did not have any associated neurologic findings, the severity and prolonged nature of his back pain was not consistent with what could have been attributed to the common causes aforementioned.

Pyogenic vertebral osteomyelitis is a disease seen predominantly in men older than age 50 yr. [4]In a retrospective review by Patzakis et al., the most common mechanism for vertebral osteomyelitis was hematogenous spread from an infected foci elsewhere within the body. [5]Sources for infection most frequently involve the genitourinary tract, skin, or respiratory tract. [4,6]Complications from surgical instrumentation of the spinous structures may result in vertebral osteomyelitis. The most common presenting features of vertebral osteomyelitis are localized spine tenderness and pain that is aggravated by movement. [7]Fever and leukocytosis are not consistently seen. Approximately one-half of patients with pyogenic vertebral osteomyelitis manifest symptoms of back pain for at least 3 months before visiting a physician for evaluation. [4]Spinal epidural abscess may result as a complication of pyogenic vertebral osteomyelitis in as many as 33% of presenting cases. [8]Epidural abscess formation results in symptoms that progress from localized spine pain and tenderness to weakness and paralysis due to compression of the involved spinal cord. The time course for the manifestations of epidural abscess formation may vary from hours to months. [4]Diagnosis of vertebral osteomyelitis is made by computed tomography, MRI, and radionuclide bone scan. Although management of vertebral osteomyelitis remains controversial regarding duration of therapy, a currently accepted practice includes a 6–8-week course of intravenous antibiotic therapy combined with immobilization for pain reduction. [9,10]Surgical treatment is reserved for patients who demonstrate either marked instability of the spine or new neurologic deficits.

The diagnosis of vertebral osteomyelitis in our patient was made by radionuclide bone scan, MRI, and percutaneous bone biopsy of the vertebral spinous process. In addition, the MRI demonstrated soft tissue changes in the L2-L3 interspace, which suggested the presence of a tract left by the epidural catheter (Figure 1). This intervertebral space was not contiguous with the infected vertebral spinous process of L1. We speculate that the epidural needle and/or catheter did not cause the vertebral osteomyelitis by direct contact with and subsequent contamination of the bone. More likely, the infection was caused by hematogenous spread from the patient's pneumonia, supported by the fact that the same pathogen was identified in both sputum and bone cultures. The presence of the epidural catheter may have caused localized tissue inflammation, thereby establishing a nidus for infection by hematogenous spread. In addition to the patient's clinical history, a predisposing factor that may have contributed to this patient's secondary infection includes immunosuppression resulting from methylprednisolone therapy, which he received during his initial hospitalization.

We cannot state with certainty that the epidural catheter was placed in the L2-L3 interspace. Palpation of the spinous processes alone may not accurately identify an interspace level during epidural placement. The findings on the MRI suggest but do not confirm that the epidural was at L2-L3. Therefore, conclusive statements regarding the proximity of the epidural site to the infection and their relationship remain speculative. It cannot be excluded that the epidural catheter or needle was contaminated and directly responsible for the infection.

This case raises the well debated issue of whether regional anesthesia is safe in patients who are at risk for bacteremia. [11]The catheter, acting as a foreign body, may serve as a nidus for infection in the epidural space or surrounding structures in the pathway of its tract. Inadvertent puncture into the subarachnoid space may lead to meningitis. Epidural abscess, although reported as a rare infectious complication of epidural anesthesia, is most commonly caused by hematogenous spread from an infection arising elsewhere in the body. [12]If epidural anesthesia is given to a patient with bacteremia, a prudent approach may be to administer prophylactic intravenous antibiotics before the anesthetic, as recommended by others. [11,13].

In conclusion, we report a case of pyogenic vertebral osteomyelitis presenting as back pain for 3 weeks after removal of an epidural catheter. The symptoms of vertebral osteomyelitis may mimic more common self-limiting causes of back pain associated with epidural anesthesia. We conclude that hematogenous spread from a coexisting respiratory infection was the most likely cause of infection, although the presence of the epidural catheter may have served as a nidus for infection. We are not aware of any previous reports of vertebral osteomyelitis attributed to an infectious complication of epidural anesthesia. This case underscores the importance of a careful investigation in the evaluation of a symptomatic postanesthetic patient.


Usubiaga JE: Neurological complications following epidural anesthesia. Int Anesthesiol Clin 1975; 13:1-153.
Fibuch EE, Opper SE: Back pain following epidurally administered Nesacaine-MPF. Anesth Analg 1989; 69:113-5.
Kane RE: Neurologic deficits following epidural or spinal anesthesia. Anesth Analg 1981; 60:150-61.
Sapico FL, Montgomerie JZ: Vertebral osteomyelitis. Infect Dis Clin North Am 1990; 4:539-50.
Patzakis MJ, Rao S, Wilkins J, Moore TM, Harvey PJ: Analysis of 61 cases of vertebral osteomyelitis. Clin Orthop 1991; 264:178-83.
Smith AS, Blaser SI: Infectious and inflammatory processes of the spine. Radiol Clin North Am 1991; 29:809-27.
Hitchon PW, Osenbach RK, Yuh WT, Menezes AH: Spinal infections. Clin Neurosurg 1992; 38:373-87.
Roberts WA: Pyogenic vertebral osteomyelitis of a lumbar facet joint with associated epidural abscess: A case report with review of the literature. Spine 1988; 13:948-52.
Cahill DW, Love LC, Rechtine GR: Pyogenic osteomyelitis of the spine in the elderly. J Neurosurg 1991; 74:878-86.
Osenbach RK, Hitchon PW, Menezes AH: Diagnosis and management of pyogenic vertebral osteomyelitis in adults. Surg Neurol 1990; 33:266-75.
Chestnut DH: Spinal anesthesia in the febrile patient. ANESTHESIOLOGY 1992; 76:667-9.
Verner EF, Musher DM: Spinal epidural abscess. Med Clin North Am 1985; 69:375-84.
Strong WE: Epidural abscess associated with epidural catheterization: A rare event? ANESTHESIOLOGY 1991; 74:943-6.