A comparative study to evaluate the feasibility of preoperative percutaneous catheter drainage for the treatment of lumbar spinal tuberculosis with psoas abscess

Lumbar spinal tuberculosis was diagnosed based on a comprehensive medical history, clinical manifestations, imageological diagnosis, as well as staining of pus and microscopy. All the enrolled patients had a history of lumbago, restricted lumbar vertebral function, partial pressing pain, and percussion pain. X-rays and computed tomography (CT) revealed osteoclasia and sequestra, and magnetic resonance imaging (MRI) showed an abscess near the pathological vertebrae and PA. Tuberculosis was confirmed by a staining and microscopic examination of pus drained from the abscess, followed by rapid liquid culture using the BACTEC MGIT 960 system (BD Biosciences, USA) and GeneXpert assay (MTB/RIF, BD, USA).

The inclusion criteria were (1) patients who fulfilled the diagnostic standard for lumbar spinal tuberculosis with PA with an abscess diameter ≥ 3 cm, (2) patients who had indications for spinal tuberculosis surgery (i.e., the presence of spinal compression, a sequestrum, overt bony destruction, and vertebral instability), (3) patients who underwent posterior pedicle screw fixation, anterior debridement, and bone graft fusion, and (4) patients who cooperated with the clinical research and provided signed informed consent.

Patients with one or more of the following were excluded: (1) patients with surgical contraindications, including those with other organ diseases who could not tolerate a long surgery and anesthesia and those with active pulmonary tuberculosis; (2) patients whose surgery had to be postponed because their ESR and CRP levels did not decrease after receiving antituberculosis treatment for 3 weeks; (3) patients with mental illness who could not cooperate with the surgical treatment; (4) patients whose antituberculosis schedules had to be changed during the 3 weeks of treatment; (5) patients who received anterior debridement and bone graft fusion in more than one segment; and (6) patients with a follow-up period less than 6 months.

In this prospective study, 72 consecutive patients with lumbar spine tuberculosis companied with PA who presented to the Hospital of Integrated Traditional Chinese and Western Medicine in Zhejiang Province (Hangzhou, China) from January 2015 to January 2017 were included. Informed consent was obtained from all the patients and this study was approved by the ethics committee of the Hospital of Integrated Traditional Chinese and Western Medicine in Zhejiang Province. All the patients received a standard antituberculosis regimen containing rifampicin (0.45–0.6 g/d), ethambutol (0.75 g/d), isoniazid (0.3 g/d), and pyrazinamide (1.5 g/d).

The participants were randomly divided into two groups: group A (n = 36) and group B (n = 36). Briefly, the randomization was performed using opaque sequentially numbered envelopes labeled with the name of the treatment group (group A and B). Briefly, the randomization was performed using opaque sequentially numbered envelopes labeled with the name of the treatment group (group A and B). These envelopes were sealed, shuffled, and then numbered in sequential order. For each new patient entering the study, an envelope was opened. Thus, this procedure avoided the chance of having more patients in one group than another. All the patients in group A underwent preoperative PCD immediately after hospital admission. Four patients in group B were excluded in accordance with the exclusion criterion (4) after 3 weeks of antituberculosis treatment. There were 20 males and 16 females in group A (age range 24–73 years; mean age 42.5 ± 10.2 years) and 18 males and 14 females in group B (age range 23–75 years; mean age 42.3 ± 9.8 years). There were no statistically significances in the sex, age, ESR, CRP levels, or involved segments of the patients (Table 1).

All patients in group A underwent ultrasound-guided PCD before surgery. Briefly, the location of the maximum diameter of the PA cavity was confirmed by ultrasound guidance after standard skin disinfection procedures and local anesthesia. Puncture needles were then inserted into the abscess cavities. The needle cores were removed when the ultrasound examination confirmed the correct position of the puncture needle. Drainage catheters were inserted into deep abscess to extract pus, followed by a smear test under a microscope to detect acid-fast bacillus. The specimens were detected using the Mycobacterial Growth Indicator Tube 960 (MGIT 960, BD Biosciences, USA) rapid liquid culture method and GeneXpert MTB/RIF assay (Cepheid Inc., USA). Finally, drainage tubes were fixed and connected with drainage bags (Fig. 1). Patients with bilateral PA also underwent bilateral puncture catheter drainage.

All patients in both groups were treated with antituberculosis drugs for 3 weeks, followed by posterior pedicle screw fixation and anterior debridement and fusion, which were performed by the same physicians (Figs. 1 and 2). The drainage bags were replaced each day during the 3-week treatment period. In all patients, anterior fusion was achieved using an autogenous bone graft from the iliac crest. Under general anesthesia, a straight incision was made in the middle of the spinous processes, with the patient in a prone position. After pedicle screw fixation, the length of the posterior incision was determined by the fixed segment. The pedicle screws were inserted into the cephalic and caudal adjacent lumbar spinal vertebral segments. The patients were then moved to a lateral position, and an anterior incision was made according to the site of segmental lesions and the size of PA, followed by blunt dissection of the obliquus externus abdominis, obliquus internus abdominis, transverses abdominis along the direction of the fibers. After exposing the peritoneum, the posterior peritoneum and the intestinal canal were pushed away to expose the psoas, reproductive nerves, and ureter. After exposure of the pathological vertebrae using a longitudinal dissociation of the psoas, a pathological examination was performed. Anterior debridement and fusion were conducted, and streptomycin (2 g) and isoniazid (0.6 g) were then administered.

The intraoperative blood loss, duration of the surgery, and anterior incision length, as well as the postoperative anal exhaust time, visual analogue scale (VAS) (1 month postsurgery), Cobb angle (1 month postsurgery), ESR and CRP levels (1 week, 1 month and 6 months postsurgery), sinus tract formation and spinal cord injury were evaluated.

The amount of anterior blood loss was considered the blood loss mediated by vacuum aspiration and the blood volume absorbed by the gauze packing. Blood mediated by vacuum aspiration was calculated using the following formula: the amount of liquid in aspirator − (the flushing fluid volume + the amount of pus). The blood volume absorbed by the gauze packing was calculated as follows: the amount of absorbed blood in the gauze packing = the total weight of wet gauze − (the weight of a single gauze pad × the total number of gauze pads).

The duration of surgery was determined as the time of starting cutting the skin until suture of the incision.

Due to the use of preoperative PCD in group A, the severity of PA may have differed in group A and group B. Thus, the length of the anterior incision may have been different. After suturing the anterior incision, the anterior incision length was measured.

The pain index was analyzed using the VAS to evaluate the clinical status of the patients before surgery and one month following surgery. Pains scores of 0, 1–3, 4–6, and 7–10 were judged as no pain, mild pain, moderate pain, and severe pain, respectively.

Before and after surgery, the vertebral segment kyphosis Cobb angle was assessed by X-ray images, with the patient in a standing position. An extension cord was drawn along the upper end plate of the centrum prior to the symptomatic vertebrae, and another extension cord was drawn along the lower end plate of the centrum next to the symptomatic vertebrae. The angle between the two lines was defined as the Cobb angle.

The Frankel grade was used to evaluate the status of spinal cord injury before surgery and at the final follow-up. Grade A denoted complete motor and sensory function disorder; grade B denoted complete function and incomplete sensory function; grade C denoted motor function disorder and incomplete sensory function; grade D denoted useful motor function, with or without auxiliary tools; and grade E denoted no motor or sensory function disorder, with a pathological reflex.

The lumbar vertebral function of the patients was estimated using the Japanese Orthopaedic Association (JOA) score.

All patients were postoperatively treated with antituberculosis and liver protection drugs. A compound glycyrrhizin injection was used as the liver protection drug, which was administered via an intravenous injection (20 ml) once a day. Patients with severe anemia and hypoalbuminemia received an erythrocyte suspension and albumin. Pyrazinamide was discontinued 3 months after the operation. After bed rest and immobilization for 12 weeks, all the patients could gradually walk under a protective load.

Sixty-eight patients had a mean follow-up time of 13 months (range 6~ 21 months). The condition of sinus tract formation in group A and group B was analyzed. Six months after surgery, the status of the sequestrum and abscess were assessed through CT examination. Other complications such as mixed infection, recurrence of tuberculosis, pedicle screw loosing, and screw pulling out were evaluated.

Statistical analyses of all data were performed using SPSS, version 16.0 (IBM, Chicago, IL, USA). The measurement data were expressed as the mean ± standard deviation (SD) and compared using Student’s t test. A chi-square test was performed to compare the enumeration data. A value of P < 0.05 was taken as statistically significant.