To the Editor:—
Spinal tuberculosis highlighted by epidural anesthesia has been scarcely reported in the literature, and in most cases, only the anterior wall of the spine was involved.1,2Here, we report a case of unknown tuberculosis involving the posterior spinal process with epidural abscess drained by obstetric combined spinal–epidural anesthesia.
A 35-yr-old Cambodian woman, gravida 2, was admitted during labor at 39 weeks’ gestation, at term of a normal pregnancy. Her medical history revealed a postpartum hemorrhage during her first delivery, which had necessitated blood transfusion with inactive hepatitis C as a sequela. At 5 cm of cervical dilatation, she benefitted from combined spinal–epidural analgesia with strict aseptic conditions. The spinal injection included 2.5 mg bupivacaine and 5 μg sufentanil. After a test dose, an epidural continuous infusion of 0.1% ropivacaine was started through a filter. The patient did not experience motor block, decrease in blood pressure, fetal bradycardia, or pruritus. The onset of analgesia was rapid, and maternal satisfaction was high. Thirty minutes later, she delivered a healthy boy without complications. The epidural catheter was removed 2 h after delivery, and the epidural tip did not show any abnormality. Before discharge from the hospital, the patient reported a small tumefaction on her back, without fever or pain. The site of the epidural puncture was slightly swollen but without signs of inflammation or neurologic abnormality. She was discharged with local antiseptic care.
Fifteen days later, the patient was treated for a back skin abscess with surgical incision and oral antibiotics in view of the increase of the tumefaction. The results of bacteriologic investigations were negative. The patient did not experience any pain or fever. The results of clinical and neurologic examinations were unremarkable. Biologic samples were within the reference ranges, except for the inflammation test result. Two months after her delivery, the patient was referred to the hospital because the wound had still not healed, despite her good health. Magnetic resonance imaging (fig. 1) and computed tomography were performed and revealed a displacement of the nerve roots by an expansive mass in the posterior extradural space, which took up more than 50% of the spinal canal. This mass began at the L2–L3 level with an L2 neural arch lysis and continued down to L4–L5. Inflammatory infiltrate extended in both psoas and sacroiliac joints. Soft tissues were involved with collections in both paraspinous muscles. An emergency neurosurgical procedure was performed with laminectomy at L2 and L3, and a solid abscess, adherent to the dura mater, was removed from the epidural space. No pus was found in the spinal canal. The soft tissues and paravertebral muscles were dissected. The patient made an uneventful recovery. All of the surgical bacteriologic samples were sterile. No alcohol-fast bacilli were identified in the surgical specimen. However, pathologic examination of the abscess showed granulomatous inflammation with epithelioid and gigantocellular cells and necrosis, which was poorly compatible with classic epidural abscess. A skin tuberculin test result was strongly positive, and a chest radiograph showed a right mediastinal mass. Empirically, double therapy was started with antibiotics and antituberculous treatment (rifampin, isoniazid, and pyrazinamide). Six weeks later, the diagnosis of tuberculosis was confirmed when Mycobacterium tuberculosis was cultured from intraoperative specimens of bone fragments and necrotic soft tissues. Moreover, a thoracic computed tomography image showed a voluminous tissue mass in the lower right laterotracheal area and pleural nodules in the upper and lower right lung. The final diagnosis was tuberculosis of the pulmonary area, mediastinal and lumboaortic lymph nodes, and vertebral areas resulting in lysis of the L2 spinal process. Epidural puncture led to revealing the disease. The outcome of the patient was good, without any sequelae, and antituberculous treatment is being continued for 18 months.
Spinal tuberculosis highlighted by epidural anesthesia has been reported twice in the literature.1,2However, this is the first report of involvement of the posterior wall of the lumbar spine. Tuberculosis of posterior bony elements is an uncommon entity.3Spinal tuberculosis usually involves the anterior wall of the spine with osteomyelitis or discitis. The location of the posterior elements of the lumbar spine (both bone and epidural space) and the preceding history of epidural catheterization led us to suspect septic complications of anesthetic management in this case. Such catheter-related epidural abscesses are well described in the obstetric population.4In theses cases, Staphylococcus aureus is the most common infecting organism. Patients generally present with fever, leucocytosis, and neurologic symptoms in addition to backaches, and diagnosis is fast because of the symptoms. On the contrary, tuberculosis is known to be a latent illness. It can be well advanced before any symptoms are sufficient for warning the medical staff. In our case, the patient was poorly symptomatic, but the extent of the lesion was very important, as shown in figure 1. The clinical presentation did not argue against usual germ epidural abscess, and the diagnosis was delayed. Nevertheless, the positive inflammation test result, in addition to a cutaneous wound that did not heal with antibiotics, in an immigrant parturient should have led to the diagnosis of tuberculosis. Immigrants coming from endemic regions frequently develop bone and joint tuberculosis as a reactivation of endogenous tuberculosis.5Furthermore, southeast Asia has the highest prevalence of tuberculosis infection, with nearly 50% of all reported cases of tuberculosis.6
Despite the decrease in immunity status during pregnancy, pregnancy does not seem to worsen tuberculosis.7Nevertheless, the vertebral hematogenous dissemination might have been supported by the increase in pregnancy blood volume, especially the epidural and vertebral blood volume, where venous pool increase is combined with vena cava syndrome. The onset of vertebral injury probably began with the L2 spinal process and then extended to the paravertebral muscles and the epidural space. The epidural puncture drained the epidural abscess and might have, as a mild trauma, exacerbated the tuberculous spinal lesion.8
In summary, we recommend that, however mild the symptoms are, abnormal recovery after an epidural puncture should lead to rapid and effective imaging. Tuberculosis is a latent illness, which is still a topical illness, especially in the nonnative and immunocompromised population.
*Hôpital Arnaud de Villeneuve, Centre Hopitalier Universitaire Montpellier, Montpellier, France. email@example.com